juvenal and hexameter

17
Juvenal and the hexameter (*) This study began as an investigation into the question of style in Juvenal. The originating context was the view that Juvenal’s satire is written in a grand or epic manner. Vocabulary, syntax, word-order, and metre would all come under con- sideration in a full investigation, but an exploratory sketch by Jonathan Powell (1999) suggests strongly that this somewhat standard view is in need of – at the least – qualification. Juvenal’s style, Powell argues, has a rather pedestrian base in comparison with the writers of elevated hexameters, i.e. epic writers, of his own day, a base which is contaminated for satiric purposes with other elements. These are generally elements recognisably drawn from loftier manners, but they do not remain pure in their new context ; rather, they are themselves contami- nated by baser elements natural to that new context. There is much in Powell’s study with which I am in agreement, but there is still room for question and qual- ification. In the course of pursuing the matter, I have both narrowed the sphere of operation (to metre) and broadened it (to include generic issues more gener- ally). Thus, I consider hexameter poets in epic, didactic, bucolic, epyllion, satire, and Statius’ Siluae. Before proceeding to the actual business there are two preliminary observa- tions that need to be made. The first concerns Juvenal’s contemporaries ; the sec- ond, the concept of a stylistic base. The relationship of Juvenal’s style to that of Latin epic remains an important question. Latin epic, however, is not monolithically uniform over its long histo- ry. If one is to say that Juvenal uses (or does not) a grand style, the first com- parands must be his contemporaries (rather than, say, Virgil). Virgil’s epic verse may have numbers of Ennian touches, but his hexameter is an Augustan one and not that of Ennius. Of course, what constitutes elevation is partly communicated by the ‘old masters’, but through a contemporary filter. Immediately we run into a problem. On closer inspection, Juvenal’s contemporaries in the field of epic- writing evaporate. Juvenal parodies Statius’ lost de Bello Germanico of c. AD 90 (Juv. 4), and refers contemptuously to the Thebaid, published around AD 91 (Juv. 7,82-87). He refers also to Lucan (7,79), who died in AD 65. He may allude to the (*) I am very grateful to Drs Stephen HARRISON, Llewelyn MORGAN, and Bruce GIBSON for comments on earlier drafts of this paper.

Upload: xzkbejac

Post on 30-Oct-2014

43 views

Category:

Documents


4 download

TRANSCRIPT

Page 1: Juvenal and Hexameter

Juvenal and the hexameter (*)

This study began as an investigation into the question of style in Juvenal. Theoriginating context was the view that Juvenal’s satire is written in a grand or epicmanner. Vocabulary, syntax, word-order, and metre would all come under con-sideration in a full investigation, but an exploratory sketch by Jonathan Powell(1999) suggests strongly that this somewhat standard view is in need of – at theleast – qualification. Juvenal’s style, Powell argues, has a rather pedestrian basein comparison with the writers of elevated hexameters, i.e. epic writers, of hisown day, a base which is contaminated for satiric purposes with other elements.These are generally elements recognisably drawn from loftier manners, but theydo not remain pure in their new context ; rather, they are themselves contami-nated by baser elements natural to that new context. There is much in Powell’sstudy with which I am in agreement, but there is still room for question and qual-ification. In the course of pursuing the matter, I have both narrowed the sphereof operation (to metre) and broadened it (to include generic issues more gener-ally). Thus, I consider hexameter poets in epic, didactic, bucolic, epyllion, satire,and Statius’ Siluae.

Before proceeding to the actual business there are two preliminary observa-tions that need to be made. The first concerns Juvenal’s contemporaries ; the sec-ond, the concept of a stylistic base.

The relationship of Juvenal’s style to that of Latin epic remains an importantquestion. Latin epic, however, is not monolithically uniform over its long histo-ry. If one is to say that Juvenal uses (or does not) a grand style, the first com-parands must be his contemporaries (rather than, say, Virgil). Virgil’s epic versemay have numbers of Ennian touches, but his hexameter is an Augustan one andnot that of Ennius. Of course, what constitutes elevation is partly communicatedby the ‘old masters’, but through a contemporary filter. Immediately we run intoa problem. On closer inspection, Juvenal’s contemporaries in the field of epic-writing evaporate.

Juvenal parodies Statius’ lost de Bello Germanico of c. AD 90 (Juv. 4), andrefers contemptuously to the Thebaid, published around AD 91 (Juv. 7,82-87).He refers also to Lucan (7,79), who died in AD 65. He may allude to the

(*) I am very grateful to Drs Stephen HARRISON, Llewelyn MORGAN, and Bruce GIBSON

for comments on earlier drafts of this paper.

Page 2: Juvenal and Hexameter

Argonautica (1,7-11) of Valerius Flaccus, who died in AD 92 or 93. These fig-ures (and the epigrammatist Martial) all predate the Satires by at least twentyyears (1). There was some overlap between Juvenal and the prose writers Tacitusand Suetonius, but after this we know little of the developments in Latin litera-ture until Fronto, Aulus Gellius, and Apuleius, all writers of a very different kindand manner (2). By the end of the second century interest in rhetorical displayseems to have increased and an archaising interest in older literature had madeitself felt. In the period between Tacitus and Fronto – the period in which wewould place ‘Juvenal’s contemporaries’ – we know of Hadrian’s small poems invarious metres and work in unusual metres by Florus and Annianus (3). Thesemay hint that the changes visible in the later authors were already beginning. Ofepics like those to which Juvenal refers or which he imitates and parodies thereis little or no sign. The only evidence we can cite is a fragment of an Alexander-poem in hexameters from after c. AD 160 (i.e. too late for Juvenal) byClemens (4). There is a case for suggesting that the literature Juvenal draws on tomake up the texture of his verse is – as are the names he uses in the Satires –from the past, but a past which is treated as a timeless cultural inheritance.

The other preliminary issue is that of a stylistic base. Clearly there areelevated elements sprinkled around the body of the Satires. There are also plain

(1) Juvenal refers or alludes also to ‘typical’ epics (the unknown Cordus’ Theseid (1,2),Telesinus’ unfinished epic (JUV. 7,22-26) and ‘typical’ subject matter – Heracles,Diomedes, Icarus (1,52-54), Aeneas, Achilles, and Hylas (1,162-4). For the dating ofJuvenal’s Satires, see R. SYME, Roman Papers vol. III, ed. A. R. BIRLEY, Oxford, 1984,p. 1135-1157, especially 1156-1157 and 1143.

(2) In this gap Greek literature continues to generate didactic, narrative and mytholog-ical epic, epigram, hymns for shrines, festivals and competitions, dramatic competitions,as well as poetry with a more prominent musical element (the citharodia). Some of thisactivity centred on Hadrian – for example, the narrative by Pancrates from Egypt about alion hunt in which Hadrian killed a lion attacking Antinous, and citharodic songs aboutAntinous. See E. L. BOWIE, Greek Poetry in the Antonine Age in D. A. RUSSELL (ed.),Antonine Literature, Oxford, Clarendon Press, 1990, p. 53-90 ; E. L. BOWIE, Literatureand Sophistic in A. K. BOWMAN, P. GARNSEY, and D. RATHBONE (edd), The CambridgeAncient History XI : The High Empire, A.D. 70-192, Cambridge, 2000, p. 898-921 ;B. J. GIBSON, The High Empire : AD 69-200 in S. J. HARRISON (ed.), A Companion to LatinLiterature, Oxord, 2005, p. 69-80.

(3) See E. J. COURTNEY, The Fragmentary Latin Poets ; edited with Commentary,Oxford, 1993, p. 372-90. The fragments and witnesses, very few though they are, attestlargely small-scale occasional or epigrammatic verse, and a ‘strong tendency to affectedsimplicity, to a mingling of colloquialism, even vulgarism, with archaism’ (p. 372). Death,wine, country scenes and ‘Fescennine’ eroticism are attested topics.

(4) E. J. COURTNEY, The Fragmentary Latin Poets [n. 4], p. 372. In the Florida (c. AD160-170) Apuleius attributes to a Clemens a verse account of Alexander the Great’s ‘manysublime deeds’ (Flor. 7) and quotes (6) three lines of distinctly Virgilian hexameters(about the Ganges) which probably belonged to it. See COURTNEY, p. 401.

2 F. JONES

Page 3: Juvenal and Hexameter

passages. Here we may remember the woman caught in adultery. Homo sum(Juv. 6,284), she concludes the speech for her defence, and the simplicity of herconcluding turn is double-edged : elsewhere it might have had a noble reso-nance (5), but here it adds to the woman’s effrontery. If we look for pure andunmixed Juvenal we may come upon the same problem of stylistic differencefrom another angle. The end of the tenth satire presents what is surely a sensibleand indeed incontrovertible position. If it is stupid to pray for anything that canturn out badly for us, we should pray – if for anything – then for that which willgive us the capacity to bear whatever comes our way. A sound mind in a healthybody seems irreducibly good (6). But even here Juvenal contaminates the expres-sion with a sideswipe at the trappings of any sort of prayer : ‘But so that you canpray for something and promise entrails and the holy sausages of a glossy littlepig to the shrines, what one should pray for is a healthy mind in a sound body’(Juv. 10,354-356 ; ut tamen et poscas aliquid uoueasque sacellis / exta et can-diduli diuina tomacula porci, / orandum est ut sit mens sana in corpore sano.).We may believe that Juvenal’s style is, in fact, an anarchic mixture of styles (7).

This problem may arise with other genres (one only has to think of theEclogues or Georgics to be aware of wild fluctuations of vocabulary and style indifferent poems or sections), but one feels that it is more extreme in satire thanelsewhere because of the generic tendency to parody. Nevertheless, if we startfrom the premiss that Juvenal’s satire contains allusions to epic and other genresand the style varies for local effects, we may be able to think in terms of astylistic neutral against which to perceive the differences Juvenal wants us to beaware of. One could hardly perform a statistical analysis on only those parts ofthe Satires which contain no allusion and no stylistic parody – if such partsexisted, disentangling them would be an insurmountable problem – but a broadcomparison of generic patterns in Latin hexameters is still productive. I wish toconsider here a number of metrical features across a broad range (temporal andgeneric) of hexameter poetry. My samples for statistical purposes have in allcases been of 100 lines, and in all cases where this has been possible the open-ing hundred lines. In the case of fragmentary authors I have used the first avail-able whole lines up to the value of one hundred.

The first feature I wish to consider is the separation of noun and epithet.Arguably, this could be considered a stylistic feature rather than a metrical one,

(5) For homo sum, cf. PLAUT., As. 243, 490, 564 ; TER., Hau. 77 ; Ad. 540 ; SEN., Ep.95.53 ; PETR., 130.1 ; PLINY, Ep. 5,3,2 ; [QUINT.], Decl. 4,12. Although this is a miscella-neous collection, Juvenal’s irony is rather pointless unless we also feel the resonance of,for example, homo inter homines (cf. PETR. 39).

(6) M. D. REEVE, Seven Notes in CR 20, 1970, p. 135-136, would delete line 356,arguing (in part) that Juvenal advises praying for virtue rather than health.

(7) In the lines quoted, we may note the different registers of exta and tomacula.

JUVENAL AND THE HEXAMETER 3

Page 4: Juvenal and Hexameter

but unarguably patterns of noun-epithet distribution are part of verse-construc-tion at a basic level, and the phenomenon is, in any case, of such importance thatit would be wrong to pass it over.

In Latin hexameter verse the separation of noun and adjective within a verse-line becomes the most obvious and one of the most important indices of stylisticelevation (8). The standard changes with time, however, so comparisons are ofnecessity somewhat complicated. The early poets Ennius and Lucilius are themost sparing of this pattern. Even here there is a probably a generic element,since in Ennius it is slightly more frequent than in Lucilius. Later – in the gen-eration preceding the Augustan period – Lucretius and Cicero have higher fig-ures than their predecessors Ennius and Lucilius, but they are low by the stan-dards of later verse. In the same period, however, Catullus stands out fromLucretius and Cicero and favours this pattern very strongly indeed (9). A littlelater, Virgil’s Eclogues follow Catullus’ taste for the separation of noun and epi-thet (though not quite favouring it to the same extent). Virgil’s Aeneid, however,is clearly distinct, reverting rather towards the older manner (10). Of course theAeneid is later than the Eclogues, but in this context hardly significantly so, andits author is the same man : we might seem then to be facing some sort of genericfactor. Although this not simply a matter of traditional epic as opposed tomodernist alternative forms such as the epyllion and bucolic, there are bothbackward- and forward-looking elements in the composition of the Aeneid, andthey have an ideological as well as a stylistic importance in this poem which the

(8) T. E. V. PEARCE, The enclosing Word Order in the Latin Hexameter in CQ 16, 1966,p. 140-171, 298-320. When the figure is split across a line boundary its effect seems to beneutralised, perhaps because the coherence of the line as unit (i.e. verse) is blurred.However, in contexts where there is a noticeable rate of separation of noun and epithetwithin the line, split-line separations may perhaps reinforce the general effect. There are,of course, degrees of elevation : a noun-epithet pair split by a preposition is probably notsignificant at all, separated by a verb it is more significant, but less so (the format is ableto move increasingly into prose ; J. N. ADAMS, A Type of Hyperbaton in Latin Prose inPCPhS 17, 1971, p. 1-16) than distributions where, say, another noun-epithet pair is locat-ed within the first pair, or the noun and epithet are placed at the beginning and end of theverse-line.

(9) The anonymous authors of the Moretum and the Ciris generally resemble neotericstylists, and are also both high in separation of noun and epithet, the author of theMoretum especially so. Also very high are various later didactic verse writers, Columella,Nemesianus (in the Cynegetica), the anonymous author of the Aetna, and (especially)Grattius. Lucretius is not uniform : in regard to separation of noun and epithet in the‘pathetic’ style versus the ‘expository’ style in the De Rerum Natura see E. J. KENNEY

(ed.), Lucretius De Rerum Natura Book III, Cambridge, 1971, p. 26-29. (10) Cicero still uses for metrical convenience older endings (-ai) and occasionally

omits final ‘s’ before a consonant, in this coming some way between Lucretius andCatullus (see CIC., Orator 161).

4 F. JONES

Page 5: Juvenal and Hexameter

intersection of the past, present, and is thematically so important. The later epicwriters are interesting. Ovid’s Metamorphoses is comparable with Virgil’sAeneid. Later again, Valerius Flaccus and Lucan are by comparison as markedlyin favour of the separation of noun and epithet as Catullan epyllion (so too theSiluae of Statius). Statius’ Thebaid, if we take into account interlaced pairs ofseparated nouns and epithets, belongs with Valerius Flaccus and Lucan. By con-trast, Silius Italicus’ epic resembles the Virgil of the Aeneid in this and in otherfeatures (as we shall see below). It looks as though to some extent the separationof noun and epithet is a modernising element as well as a stylistically elevatingone. It is very prominent in Virgil’s Georgics (as in the Eclogues, but not theAeneid), and very much more prominent there than in the didactic poems of thepreceding period, Cicero’s Aratea and Lucretius’ De Rerum Natura. The laterbucolic writer Calpurnius Siculus (11) favours the separation of noun and epithetvery strongly indeed (counting double as well as single separations). Perhapshere we have a case of inflation as the author seeks to mark his work with a fea-ture of the modernising Virgilian Eclogues (12), but one which was tending astime went on to become a more frequent mannerism in general.

We might expect the avowedly pedestrian satirists to be different from otherwriters of hexameters. If we take the outlines given so far as crude markers wemay insert the satirists. Lucilius is very sparing with the separation of noun andepithet indeed. He is, admittedly early, but his figure is lower than that for theepic of his contemporary Ennius (13). Horace’s avowedly pedestrian Satires aremore sparing – in the case of Book 1 much more sparing – than hexameters inthe other genres of his own day In Book 1, indeed, the figure is lower than anyof the sample except Lucilius. Horace’s remaining books of hexameters (allsatiric, or related to satire) rise to the level of the didactic poets of an earlier time,Cicero and Lucretius.

The next extant satirist, Persius, also favours separation less than his contem-poraries in other genres. The figure for Persius is a rough match for the laterHoratian books rather than Book 1, but this does not mean that he starts at the

(11) The dating is still subject to debate : E. CHAMPLIN, The Life and Times ofCalpurnius Siculus in JRS 68, 1978, p. 95-110 ; G. B. TOWNEND, Calpurnius Siculus andthe Munus Neronis in JRS 70, 1980, p. 166-174 ; R. MAYER, Calpurnius Siculus :Technique and Date in JRS 70, 1980, p. 175-176 ; N. HORSFALL, Criteria for the Dating ofCalpurnius Siculus in RFIC 125, 1997, p. 166-196 ; T. P. WISEMAN, Calpurnius Siculusand the Claudian Civil War in JRS 72, 1982, p. 57-67.

(12) Nemesianus does not seem to do this in his Eclogues : the figures for both singleand double hyperbaton are higher in his didactic work, the Cynegetica.

(13) Ennius himself, it should be noted, is very sparing by the standards of later hexa-meter writers : the nearest match among later writers is provided by a satirist in the firstbook of Horace’s Satires.

JUVENAL AND THE HEXAMETER 5

Page 6: Juvenal and Hexameter

level of stylistic elevation Horace had reached. In comparison with his contem-poraries, Persius is less elevated, at least in this feature of the hexameter, thanlate Horace.

Like the earlier satirists, Juvenal too favours separation of noun and epithetless than his contemporaries, but there are still differences. Juvenal is more freewith this stylism than Persius. The difference between Juvenalian satire and con-temporary hexameters in other genres is still present, but Juvenal seems closer tohis contemporaries (if one can use the term so crudely) than Horace or Persiusto theirs (14). Juvenal, on this score, is less pedestrian than the other satirists (15).

When the noun-epithet pair encloses or is interwoven with another separatednoun-epithet pair the effect is more elevated, and one can consider this phenom-enon in its own right as well as merely letting such double separations contributetwofold to the figures for separation of noun and epithet. The picture thatemerges is not straightforward. For patterned double separations we get extraor-dinarily high figures in the quasi-neoteric authors of the Culex and the Ciris, andManilius. After these, the next highest figures occur in Catullus, CalpurniusSiculus, and – perhaps surprisingly – Lucan. Another apparently unhomoge-neous group still have quite high figures : Virgil’s Eclogues (again, as with sep-aration of noun and epithet generally, distinct from the other Virgilian poems),Cicero’s Aratea, Statius’ Thebaid (but not his Silvae), Ovid, and Juvenal. Thelowest figures are for both Lucilius and Ennius, and Lucretius, Virgil’s Aeneidand Georgics, the epic of Valerius Flaccus, the later hexameters of Horace, andStatius’ Siluae.

In each of these groups we find authors of different periods. The earliestauthors are among those with the lowest figures, but Catullus – the next earliest– is among those with the highest. The conspicuous difference between theEclogues and the other works of Virgil may suggest that there is an element ofmannerism or flagrant stylism about the double separation, an idea supported bythe difference between Catullus (high figures) and both his predecessors (low)

(14) J. G. F. POWELL, Stylistic Registers in Juvenal in J. N. ADAMS and R. G. MAYER

(edd.), Aspects of the Language of Latin Poetry, Oxford, 1999, p. 324 gives figures (witha due note of caution about arbitrary assumptions about what constitutes a hyperbaton)for the first 50 lines of Horace’s Satires 1 (17), the Aeneid (20), Statius’ Thebaid (36), andJuvenal 1 (26).

(15) Individual instances of hyperbaton cannot automatically be registered as ironic orparodic. Concentrations are a different matter, as at the beginning of the sixth satire. Herethere is an unusual concentration of the more elevated stylistic features. Even here, how-ever, it is too simple to talk of parody. Modern civilisation is shown as adulterated at thesame time as the poeticised Golden Age is revealed as uncouth. This is what the standardpoetic Golden Age looks like viewed through the contaminated eyes of the post-Goldenera, the Ninth Age (JUV. 13,28). The complex mixture of styles mirrors the complexity ofthe perspective we are invited to share.

6 F. JONES

Page 7: Juvenal and Hexameter

and his contemporary, Cicero (quite high, but not as high as Catullus), and alsoby the low or lowish figures for double separation in at least some of the pedes-trian satirists (Horace, especially the later Horace, and Persius). Silius Italicus isnot at the lowest end of the scale in absolute terms, but in comparison with mostof the later authors he does not seem to favour double separation very much. Inthis – and in other features as we shall see below – Silius may intentionally befollowing the model of the Aeneid. Statius’ Siluae, although stylistically man-nered, perhaps need to maintain the pose of near-improvisation too much toindulge in such a mannered device.

Given the generic anomalies here, it is difficult to claim that high figures fordouble hyperbaton represent a grander level as such, but the pairing of Catullusand Virgil’s Eclogues in contrast to that of Horace and Persius suggests that dou-ble hyperbaton is a mannerism of conscious stylistic polish. There is probably achronological or fashion element in which Catullus (very different fromLucretius in this and other respects) is a seminal figure. We may be inclined tosee this level of stylistic decoration as a feature especially of neoteric verse andits descendants (including bucolic). It is striking that Juvenal is, though perhapsnot by much and apart from the sample from Satire 10 (Book 4), the freest of thesatirists with this figure, and matches some of the epic writers. This may help tostrengthen the idea that Juvenal goes some way towards elevating satire’s rela-tive stylistic position.

Most hexameters end in words of two or three syllables. The use of singlemonosyllables at line-ends is a distinctive variation in hexameter verse. Its dis-tribution seems at first to put Juvenal in an unlofty stylistic bracket with the othersatirists. The phenomenon is important, but need not, in fact, conflict with theidea of a degree of elevation on Juvenal’s part. Monosyllabic line endings (evenlines ending less objectionably in two monosyllables) (16), become very rare afterLucretius (already there are none in the sample from Catullus) (17), but they are

(16) The single monosyllable, but not double monosyllables, produces conflict ofaccent and ictus at the end of the verse-line. Figures for the two types are distinguished inthe table below.

(17) In fact there is only one real case in CAT. 64, dens at 315 (final est is elided at 147and 301). J. HELLEGOUARC’H, Le monosyllabe dans l’hexamètre latin : essai de métriqueverbale, Paris, 1964, 15-17 finds in the global figures for monosyllables at any point inthe hexameter both a tendency towards the gradual elimination of the monosyallable, anda tendency to favour the monosyllable less in loftier genres. It is striking – and coherentwith the results elsewhere in this paper than if we take a low figure for monosyllables asevidence of elegance or loftiness, we find Juvenal’s place in the sequence Catullus (64),Lucan, Statius, Virgil (Aen. and Georg.), Lucretius, Horace, Persius unexpected. He fallsin with Virgil’s Eclogues, within the range shown in the books of Lucretius, and signifi-cantly more polished than both Horace and Persius. The extreme position of Persius andthe very low figure for Lucan are also in line with other evidence given in this paper.

JUVENAL AND THE HEXAMETER 7

Page 8: Juvenal and Hexameter

found : they are present in Horace’s hexameter poetry, especially the Satires andfirst book of the Epistles (18). Double monosyllables are a significant presence inPersius (though with a smaller profile than Horace’s Satires), and both single anddouble monosyllabic line-ends are found in Juvenal (with a smaller profile thanHorace’s Satires). Although both are found in the Aeneid (where Homer andEnnius often provide precedents), and – so far as one can see – both seem to havethe same sort of frequency in the satirist Lucilius as in the epicist Ennius, theyvirtually disappear except in the satirists after the Republican period. Althoughit looks as though this feature becomes an element of satiric character, it mustalso be noticed that many cases of single monosyllables at line ends in Juvenalhave special point, and may therefore be exceptional rather than part of a normalstylistic level as such (19).

Other non-standard patterns of word shape at line-ends (typically final wordsof four or occasionally more syllables, or lines ending with double-spondees)have high figures above all in Ennius and Lucilius, then in Cicero, Lucretius andCatullus (in Catullus they are almost all associated with double-spondaic lineends), to a lesser extent in Persius and Juvenal, and to a lesser extent againHorace and Ovid. Quite often these cases arise from Greek names, and quiteoften (whether this is the case or not) there is in the satirists a special point.Nonetheless, after the earlier periods there appears a bias towards the satirists inthese figures.

There is also the evidence of the third foot word-break (caesura) in the hexa-meter, though this is rather ambiguous (20). Usually the caesura is found after thelong syllable of the third foot in the hexameter (strong caesura). Alternatively, itmay come after the first short syllable of the third foot (weak caesura), in whichcase it is often supported by strong caesurae in the second and fourth feet.Juvenal and – especially – Persius seem to be low in weak third foot caesuraecompared to their contemporaries, but Horace is in a more complicated position.Most of his hexameter books have roughly the same order of frequency asCatullus, Lucretius, Cicero, Virgil’s Aeneid and Georgics, and Ovid’sMetamorphoses, but much less than Virgil’s Eclogues and a number of laterauthors and bucolic generally (perhaps oddly, Statius is very high in both the

(18) For these features in Horace’s Satires and first book of Epistles see R. MAYER

(ed.), Horace Epistles Book I, Cambridge, 1994, p. 13-21.(19) See J. HELLEGOUARC’H, Le monosyllabe [n. 18], p. 50-69 for more detail over a

range of authors and various categories of final monosyllable.(20) See J. G. F. POWELL, Stylistic Registers in Juvenal [n. 15], p. 314. Perhaps it is sig-

nificant that a noticeable number of lines in Persius (especially) and Juvenal lack a thirdfoot caesura (15 in the first 300 lines of Juvenal’s sixth satire), whereas there are none inthe sample from the Silvae and 7 in 400 lines of the Thebaid.

8 F. JONES

Page 9: Juvenal and Hexameter

crafted Thebaid and the professedly impromptu Silvae) (21). Here it is theEclogues that seem to stand out. In this case (as opposed to hyperbaton) theystand apart from Catullan epyllion, where we might have expected to see com-parability rather – it is surprising that the modernising Catullus is so close hereto the in some ways rather old-fashioned Lucretius. If, however, the weakcaesura really is an elegance – as its high profile in the Eclogues suggests – it isnoteworthy that the figure is about as high as for the Eclogues in the last two ofHorace’s hexameter books, Epistles 2 and the Ars Poetica. In other respectsHorace’s hexameters become more elegant with time and perhaps this is how weshould see the increase in weak caesurae here too. We might note in this contextthat the figure for weak third foot caesura in Ennius’ epic is higher than that inthe satirist Lucilius (22).

Within the verse of individual authors the weak third foot caesura seems to bedistributed rather randomly. The low concentration at Juv. 1,1-34 does not seemto have a special purpose, and the posture of indignant exclamation in that pas-sage is not too dissimilar from that of Juv. 6,217-22 where there is an unusuallyhigh concentration. If, then, individual weak third foot caesurae, or even con-centrations of them, are not significant for the tone at the level of the passage, itmay be that they do not stand out as such. If, then, their general levels of fre-quency have any generic correlation it may be that this is by-product of othergeneric metrical characteristics. A high or low level may well in that case stillcontribute to an overall impression of a generic metrical or stylistic norm.

Lines in which there is no third foot caesura – although there are exceptions– are most frequent in early authors and in the satirists, and amongst those mostfrequent in Lucilius and Persius. Here (as perhaps also in his comparatively lowincidence of weak caesurae) Persius demonstrates a roughness which appears ina number of other phenomena considered in this paper. Amongst the satirists, theabsence of a third foot caesura is by contrast rarest in Horace’s earlier hexame-ter books and Juvenal’s first two books. If we can take a regular presence of thirdfoot caesurae as orthodox, we might see the low figures in the earliest Horatianworks as part of the programmatic tightening up of Lucilian slackness. This,however, will not explain the low figures in Juvenal’s first two books. We shouldsee him, rather, as conforming in this regard with the values of the less lowlyhexameter genres.

Figures for elision show mixed, but highly interesting – and possibly morepalpable – results. It is important to distinguish between degrees of elision. Theelision of long vowels and diphthongs – especially before short vowels, and

(21) Unusually in this respect, Silius stands closer to the Eclogues and his own con-temporaries than to the Virgil of the Aeneid and the Georgics.

(22) Lucilius stands roughly equal to Catullus, Lucretius, Cicero, Virgil (Aeneid andGeorgics), Ovid, and the satires of Horace (and Epistles 1), Persius and Juvenal.

JUVENAL AND THE HEXAMETER 9

Page 10: Juvenal and Hexameter

especially elision of -ae – is more obtrusive than elision of short vowels (23).Nevertheless, clusters or concentrations of elision, whether largely or wholly ofshort vowels, can still draw attention to themselves. The beginning of Persius’first satire has two elisions in line 1 (of short vowels), two more in line 2 (oneshort, one long) (24), two more in line 6 (of short vowels), and three in line 9 (ofshort vowels). The cumulative effect in this short initial passage is striking, butthe broader picture is also noteworthy : in a sample of 100 lines Persius hasseventeen lines containing two or more elisions. From the opposite angle, theextreme infrequence of any form of elision in Calpurnius Siculus (on whom, seebelow) also suggests that light elision cannot be ignored

An extremely high figure for elision is found in Lucilius, but a very low onein Ennius. Both are at the early end of the period considered, and the large dif-ference clearly has to do with generic aesthetics. It is of some note that the nexthighest figure after Lucilius is the satirist Persius. High figures are also found inLucretius, but not in his contemporary Catullus. Here we may detect another fac-tor at work. Catullus and Lucretius also differ in their treatment of final ‘s’ beforea consonant, something that Cicero indicates (Orator 161) was an issue of fash-ion (i.e. not genre) in this period (25). Thus Catullus is, and (in this respect atleast) Lucretius is not, a modernist.

In the next generation a high rate is still found in the Aeneid and the Georgics(the Eclogues are again distinct from the other Virgilian poems). The figure isless than half that for Lucilius (from whom Virgil differs in period and genre),and over twice that for Ennius (from whom he differs in period, but not genre).The only other really comparable figures are for later writers, the epic-writerSilius (who resembles Virgil in other metrical respects) and the satiristPersius (26). Among both his contemporaries in other genres and also writers ofepic and didactic, Virgil seems to be anomalous. On the other hand, any propo-sition one might care to make about the figures for elision proves problematic. Itis not the case that earlier authors are systematically freer with elision than later ;nor is it the case that satirists are systematically freer than writers of other kinds

(23) See N.-O. NILSSON, Metrische Stildifferenzen in den Satiren des Horaz, Uppsala,1952 ; J. SOUBIRAN, L’élision dans la poésie latine, Paris, 1966.

(24) The following vowel is long ; that is to say, this is not in itself the most flagrantform of elision possible (elision of long vowel before short vowel).

(25) Common in Lucretius, still used by Cicero, but only once in Catullus (116.8).Cicero sees it as newly unfashionable (Orator 161). Note VIRG., Aen. 4,628-9, litoralitoribus contraria, fluctibus undas / imprecor, arma armis, where Virgil breaks thesequence of polyptota by writing undas instead of fluctus, which would have been possi-ble with elision of final ‘s’.

(26) The anonymous author of the Ciris and Grattius are perhaps rather surprisinglyfavourable to elision, given their apparently high stylistic aspirations.

10 F. JONES

Page 11: Juvenal and Hexameter

Table. — Verse figures per 100 line samples

A monosyllabic line endingsB other non-standard line endingsC lines ending with two monosyllablesD lines with weak caesura in foot 3E lines with no caesura in foot 3F elisionsG Hyperbata inside linesH Double-hyperbata inside lines

Author A B C D E F G H

Ennius 10 16 1 14 6 16 14 1V., Aen. 1 1 2 8 5 47 33 1Ovid, Met. 0 2 0 8 0 25 31 4Val. Flacc. 0 0 0 24 5 33 45 2Lucan 0 0 0 18 1 14 43 11Silius Italicus 0 0 0 16 3 44 34 3Stat., Theb. 0 0 0 22 5 32 35 6

Catullus 64 0 17* 0 7 2 25 45 12Moretum 0 0 0 17 1 16 51 8Dirae 0 0 0 6 1 24 38 4Culex 0 0 0 13 2 10 32 20Ciris 0 1 0 4 1 33 42 21

Lucretius 2 8 0 6 6 35 21 0Cicero, Aratea 2 3 0 6 7 20 25 5V., Georg. 0 0 0 8 3 43 50 1Manilius 0 0 0 15 3 23 25 18Columella 0 2 0 10 2 11 40 5Nem., Cyneg. 0 0 0 22 3 10 49 7Grattius 1 0 0 11 3 42 53 11Aetna 0 1 1 7 0 26 46 8

V., Ecl. 0 0 0 17 2 22 43 6Calp. Sic. 0 0 0 15 0 1 51 12Nem., Ecl. 0 1 0 19 3 14 34 2

Lucilius 6 15 1 6 9 104 9 0Hor., Sat. 1 5 3 7 8 2 33 14 3Hor., Sat. 2 1 1 7 7 0 34 22 3Hor., Epp. 1 4 1 7 5 2 26 26 1Hor., Epp. 2 0 2 1 14 4 19 23 2Hor., AP 2 0 2 18 4 14 24 0Persius 0 4 6 4 8 63 21 2Juv., Book 1 3 5 2 8 1 22 28 4Juv., Book 2 2 1 4 7 1 24 32 4Juv., Book 3 2 6 3 9 5 33 25 3Juv., Book 4 3 1 3 5 5 46 25 1Juv., Book 5 4 1 3 11 3 32 34 3

Stat., Silu. 0 0 0 25 0 18 47 1* Sixteen of these are double-spondaic line ends.

JUVENAL AND THE HEXAMETER 11

Page 12: Juvenal and Hexameter

12 F. JONES

of poetry. It is the case that among epic writers Virgil and Silius are exceptional.It is also the case that two satirists stand out as very free with elision comparedto any other writers – except to some extent Virgil of the Aeneid and Georgics –and it is true that Horace begins his satiric career being quite free with elision(but only a little more than some epic writers and Lucretius), but increasinglyless so after the Satires. By contrast Juvenal’s later books are freer than his ear-lier (he peaks in the penultimate book). Clearly this is an aspect of versificationwhich means something to hexameter writers, but does not fall into a simple his-torical or generic pattern.

If we look specifically at patterns of strong elision (elision of long syllables)and elision clusters in Augustan and later writers, the general picture alters onlya little. Horace’s first book of Satires assumes a higher profile in this respect withten strong elisions in the sample (of which two are followed by short syllables,and five are of monosyllables) (27). There are twelve cases in the sample of SiliusItalicus (of which seven are followed by a short vowel – one involving -ae – butnone are of monosyllables), eight cases in Persius’ Satires (of which four are fol-lowed by a short vowel, and two are monosyllables, and one is elision of -ae) (28),eight in the Aeneid (all are followed by long syllables, but none are of monosyl-lables), six in the Georgics (all are followed by long syllables, none are of mono-syllables), four in Horace’s first book of Epistles (of which all are of long sylla-bles and three are of monosyllables) (29), three in the Eclogues (all followed bylong syllables ; one case is elision of a monosyllable) (30), three in Ovid’sMetamorphoses (all followed by long syllables, and none involving monosylla-bles), two in Statius’ Siluae (both followed by long syllables, neither with mono-syllables), and one each in Juvenal’s first book of satires (not involving a mono-syllable), and Statius’ Thebaid (followed by a short vowel, but not involving amonosyllable).

In this evidence we see – as before – a distinction between the Aeneid andGeorgics as against the Eclogues, the comparability of Silius and the epic Virgil,the increasing refinement of Horace’s hexameter poetry (31), and the fallingtogether of the satirists Horace and Persius in contradistinction to Juvenal. In thisrespect there seems to be a clear element of elevated polish in Juvenal’s hexa-meters, although the evidence of the later books shows that a slight qualificationis needed. Figures for elision of a long syllable are as follows : two in the

(27) HOR., Sat. 1,1,41, te ; 44, ni ; 81, te ; 86, tu. Three other monosyllables are elided,nam (33), dum (52), and quam (56). PERS. 1,62 elides -ae before occurrite.

(28) PERS. 1,66, si ; 89, te. Two other monosyllables are elided, namely cum (9), andquem (44).

(29) HOR., Epp. 1,1, 3, me ; 27, me ; 75, te.(30) V., Ecl. 1,40, me.(31) Cf. R. MAYER (ed.), Horace Epistles Book I [n. 19], 13-21.

Page 13: Juvenal and Hexameter

JUVENAL AND THE HEXAMETER 13

hundred-line sample of Book 2 (one followed by a long, and one by a shortvowel ; both elision of -ae), two in Satire 7 (Book 3), both followed by shortvowels, four in Satire 10 (Book 4), two followed by a short vowel (one elision of-ae) and two by long vowels (one elision of a monosyllable ; si, 75), two in Satire13 (Book 5), both followed by long syllables. The figure is never high, but therefinement – unlike that of Horace’s hexameters – diminishes somewhat towardsthe fourth book.

As regards elision clusters, a cursory review of some Augustan and laterauthors shows only two modifications of any substance : Persius the satirist rais-es his profile, and Juvenal lowers his. Lines with two or more elisions (and insuch cases it is usually a matter of short vowels) are found (in samples of 100lines) seventeen times in Persius, ten times in the Aeneid, nine times in theGeorgics, seven times in Horace’s Satires 1, six times in Silius Italicus, fourtimes in Horace’s Epistles 1, three times in Ovid’s Metamorphoses, once inVirgil’s Eclogues, and not at all Lucan, or either Statius’ Thebaid or Siluae. InJuvenal Satires, there are likewise none in the first book, but two in Book 2, threein Book 3, six in Book 4, and four in the last book. Again we see a differencebetween Juvenal and other satirists, most marked in the first book, but somewhatdiminishing in later books, especially the fourth.

The global figures for all kinds of elision is middling rather than high inCicero, Catullus (32), Virgil’s Eclogues, Ovid, Juvenal, and Statius’ Silvae (wherethere are only two elisions of long vowels in a hundred lines, both followed withlong vowels). This does very little to take the edge off the impression received ofJuvenal’s refinement in respect of elision gained from looking at heavy eli-sion (33). On the other hand, the presence of the Silvae – however slapdash theypretend to be, they are not – should not surprise us (especially if we also believethat a high rate of weak third foot caesura was a sign of elegance, since they arevery frequent in the Siluae) (34). The smoothness of low rates of elision is furthersupported by observation of the distinctly low rate in Lucan (whose sample hun-dred lines have no elision of long vowels, not elision clusters) (35), since he is also

(32) There are seven elisions of long vowels in CAT. 64, one of which is followed by ashort vowel (64.70), and two of which ore on monosyllables. Of clusters there are onlythree. In Lucretius there are three elisions of long vowels (two followed by short vowels),but ten clusters.

(33) See also J. SOUBIRAN, L’élision dans la poésie latine [n. 24], p. 597-610 on thegeneric factor in general, and 608-9 on Juvenal in this context.

(34) Perhaps it is rather surprising at first sight that the impromptu (for so Statius says)Siluae are much more restrained with elision than the carefully wrought Thebaid. Thespontaneity and casualness of the Silvae is, however, part of an elaborate game of displaywhich is half-concealed at the same time as being ostentatious.

(35) Such elision as one finds is chiefly on -que.

Page 14: Juvenal and Hexameter

14 F. JONES

very distinctly high in the elegant hyperbaton and double hyperbaton (and theweak third foot caesura). Further corroboration comes from the fact that in thisfeature as in others Horace’s hexameter books move away from roughness (wehave already seen this confirmed in the matter of heavy elision and elisions clus-ters), and also the fact that the Eclogues of Calpurnius Siculus have a highish rateof weak caesura, very high figures for hyperbaton and double hyperbaton, and(in the sample taken) an astonishingly low – almost non-existent – figure for eli-sion.

Even in his early hexameter books Horace is much more sparing of elision (36)than the Virgil of the Aeneid and the Georgics (but considerably more free withit than the Virgil of the Eclogues). Calpurnius Siculus’ Eclogues have an almostcomplete absence of elision. If we consider this in the light of the figures forweak caesura, hyperbaton, and double hyperbaton, it seems possible thatCalpurnius is taking to an extreme the distinction of Virgil’s Eclogues from hisother works, and thereby defining it as a genre with its own metrical character inaddition to its other individuating characteristics (37). On the other hand we areleft with the question of Virgil’s Aeneid.

The coincidence of Catullus and bucolic at one end of the scale, and Lucilius,the Virgil of the Aeneid and Georgics, Silius, and Persius at the other cannot bewithout significance. The resemblance of Silius to Virgil in this feature falls intoline with his resemblance in other aspects of metre. We might wonder whetherwhatever effect this or that feature of verse technique has in Virgil is lost in Siliusbecause he is simply being ‘Virgilian’, but we do not have to consider him inde-pendently. This allows us very clearly to see an opposition between the aesthetes,if one may so call them, on the one hand, and, on the other, Lucilius, the non-bucolic Virgil, and Persius, an opposition which in turn suggests that frequentelision marks a rough ‘manly’ style (38). This suits each in various ways. We cansee how this suits Lucilius’ independent outspokenness, and the character attrib-uted to him by later writers (39), Virgil’s Aeneid with the complex mix of forwardand backward looking in its composition, allusiveness, and revaluation of pastliterature and ideologies, and the Georgics with its return, after the Theocritean

(36) Elision is a clear difference between Horace and Lucilius, one which we shouldsee in the light of Horace’s strictures on Lucilius’ verse technique (HOR., Sat. 1,4,8-13 ;1,10,1-3 and 50-1.

(37) Nemesianus’ Bucolics seem to adhere to this reasonably closely, but his didacticCynegetica seem to break the pattern by being even more polished.

(38) See SOUBIRAN, L’élision [n. 24], 613-45 for the variety of local expressive effectsof elision and elision clusters.

(39) For Lucilius’ occupation of a special place in the Romans’ sense of themselves seeLl. MORGAN, Satire in S. J. HARRISON (ed.), A Companion to Latin Literature [n. 3], p. 177-178 ; ‘adoption of the Lucilian mode conveys that all is right with the Roman world’.

Page 15: Juvenal and Hexameter

JUVENAL AND THE HEXAMETER 15

Eclogues, to Hesiodic antiquity and the roughness of the earth. Finally, the toughmanliness I am attributing to a high rate of elision suits Persius’ abrasivelyStoical postures.

With elision, then, we appear to have a stylistic feature where the differencesbetween authors are very striking, but where, if the satirists mark themselves offas a group, it is a group from which Juvenal distances himself. It is also a featurein which the early books of Juvenal seem as polished as the approximately con-temporary epic author of the Thebaid. It is, indeed, very striking how the sam-ples show an increasing freedom with elision as Juvenal’s books of satires pro-ceed, along with a diminution of elegant or grand features such as hyperbata anddouble hyperbata (40).

There is another body of material, which shows some similarities and somedissimilarities with the material rehearsed above. In 1991 Stephen Harrison gavefigures for a particular kind of hexameter end – the format in which a noun ispreceded by an adjective with a similar short ending, as discordia taetra or flu-mina nota (41). According to the figures provided by Harrison there is, in broadterms, a chronological evolution towards greater refinement marked by anincreasing rarity in hexameter ends of this format. However, Harrison also notessome anomalies. Silius, for example, stands out from his contemporaries and (asabove too) shows an anachronistic resemblance to the Virgil of the Aeneid. Lucantoo shows less polish than might have been expected from a post Ovidian author.Harrison reminds us that Lucan used a lower and more realistic vocabulary thanVirgil : ‘a (mildly) less refined vocabulary is matched by a (mildly less refinedmetrical practice, and both are deliberate authorial choices’ (42). However, thepicture is complicated. As Harrison observes, other authors (Manilius andGermanicus) revert to a pre-Augustan level. Furthermore, as emerges in thematerial reviewed earlier in this paper, Virgil’s level of polish is itself anomalousin some respects – in a number of respects other than that considered byHarrison, it is the earlier Eclogues that show remarkable polish, whereas theGeorgics and Aeneid show consciously less refined features, and Lucan himselfhas some unexpectedly refined features. As regards the satirists, in the use of thiskind of hexameter end Horace stands with Virgil, Persius (no examples in thecorpus) is perhaps unexpectedly refined for a satirist, and Juvenal shows a level

(40) It is interesting that the ‘rough’ feature of elision increases at the same time asJuvenal’s posture of direct anger decreases.

(41) HARRISON (1991), with an appendix in HARRISON (1995). S. J. HARRISON, DiscordiaTaetra : The history of a hexameter ending in CQ 41, 1991, p. 138-149 ; see alsoS. J. HARRISON, Discordia Taetra : Appendix in CQ 45, 1995, p. 504.

(42) S. J. HARRISON, Discordia Taetra, 1991 [n. 42], p. 146.

Page 16: Juvenal and Hexameter

16 F. JONES

like that of Virgil and Horace – i.e. noticeably less refined than most epic writ-ers close to his own time (43).

We may now ask to what extent there is anything like a distinctively satiricversion of the hexameter, and part of this question must be to consider what sucha version would be distinct from. There are, perhaps, indications that Virgil andCalpurnius sought to distinguish their bucolic from epic in metrical terms. A dis-tinction between epic and didactic is less easy to believe in (44). Certainly in thecase of Virgil there is much more of a difference between the bucolic and the restof his poetry than there is between the epic and the didactic. Within epic itselfthere are variations. Virgil and Silius, for example, are much less averse to eli-sion than other epic writers. Lucan, although he shows unelevated signs invocabulary and in the level of the discordia taetra hexameter end, nonethelesshas also a number of features that seem symptomatic of stylistic elevation : a lowrate of elision, and high rates of hyperbaton and double hyperbaton. The earlypoet, Ennius, is quite different from later verse-writers. However, despite suchvariations we find that the satirists tend in comparison with their contemporaries– and in various degrees – to have lower figures for elegant or elevated featuresand higher figures for unelevated features. Persius very strongly favours elision,a distinctive, but not a particularly satiric, characteristic, but he is also rough inall the ways satirists tend to be, except in his apparent distaste for the discordiataetra type hexameter end ; Horace becomes less unrefined as his work proceeds,Juvenal less refined. Of the three Juvenal is the closest in most respects to themanner of serious, non-satirical poets of his own time.

Various features, for Powell, erode the validity of describing Juvenal’s style asgrand. He points to a persistent stylistic contamination, the mingling with moreelevated features of various prosaic elements such as diminutives, vocabulary,and certain turns of phrase, and concludes that in general Juvenal belongs withthe pedestrian satirists and that epic features in his poetry stand out against thisbackground as different – i.e. not as part of his own style as such, so much assomething alien, or perhaps a target. Powell goes on to say that ‘epic’ passages

(43) Juvenal also seems to stand with the satirists rather than the loftier authors in thematter of trochaic punctuation, although here too there are anomalies, such as the high fig-ure in and Statius’ epic Thebaid (J. GÉRARD, La ponctuation trochaïque dans l’hexamètrelatin d’Ennius à Juvénal, Paris, 1980, p. 206-207). It might be noted that the figures seemalso to distinguish Virgil’s Eclogues from the Georgics and Aeneid.

(44) On the relationship between didactic and epic, see K. VOLK, The Poetics of LatinDidactic, Oxford, 2002, p. 25-43 (didactic is separate from epic) ; P. TOOHEY, EpicLessons : an Introduction to Ancient Didactic Poetry, London, 1996, p. 5-6 (didactic is aform of epic) ; see further M. GALE, Myth and Poetry in Lucretius, Cambridge, 1994,p. 99-104 ; M. GALE (ed.), Latin Epic and Didactic Poetry : Genre, Tradition andOriginality, Swansea, 2004).

Page 17: Juvenal and Hexameter

JUVENAL AND THE HEXAMETER 17

in Juvenal may indeed be infiltrated by the lower style of the linguistic matrix inwhich they are set. At one point he takes a rather more advanced position : ‘somepassages ... appear to shift about so quickly from the everyday world to the worldof epic and back again that the unwary reader might get an impression of a chaot-ic mixture of stylistic levels.’ (45). This advanced position has a considerabledegree of plausibility, and it is not just a matter of linguistic register ; Juvenalalso blends allusions to genres of widely different stylistic and thematic colour-ing in the allusive texture of his satire. However, the evidence of a range of met-rical phenomena adduced above suggests the possibility of a slightly differentreading. Although there is a progression towards a somewhat more pedestriantone (peaking in the fourth), the verse technique of Juvenal’s earlier books iscloser to his grander contemporaries than earlier satirists were to theirs, and it isplausible to think of this as part of a palpably more literary general tenor.

University of Liverpool, UK. Frederick JONES.

(45) J. G. F. POWELL, Stylistic Registers in Juvenal [n.15], p. 327.