Bryn Mawr Classical Review 95.02.37


Roland Mayer, Horace: Epistles I. Cambridge: Cambridge University Press, 1994. Pp. viii + 291. $59.95 (hb). $21.95 (pb). ISBN 0-521-25898-7 (hb). ISBN 0-521-27754-X (pb).


Reviewed by John Moles, University of Durham.

Though H(orace)'s Epistles have always inspired some excellent scholarship, this is a particularly fruitful period. Noteworthy recent and forthcoming contributions, from scholars of very different interests and approaches, include: R.S. Kilpatrick, The Poetry of Friendship: Horace, Epistles I (Edmonton 1986); A. Traina, 'Orazio e Aristippo. Le Epistole e l'arte di convivere', R.d.F. 119 (1991), 285-305; N. Horsfall, La villa sabina di Orazio: il galateo della gratitudine. Una rilettura della settima epistola del primo libro (Venosa 1993); W.R. Johnson, Horace and the Dialectic of Freedom: Readings in "Epistles 1" (Ithaca and London 1993); L. Bowditch, 'Horace's Poetics of Political Integrity: Epistle 1.18', AJP 115 (1994), 409-26; R.J. Ball, 'Albi, Ne Doleas: Horace and Tibullus', CW 87 (1994), 409-14; L. Pearcy, 'The Personification of the Text and Augustan Poetics in Epistles 1.20', CW 87 (1994), 457-64; S.J. Harrison, 'Poetry, Philosophy and Letter-Writing in Horace Epistles 1', in D. Innes, H. Hine and C. Pelling, Ethics and Rhetoric: Classical Essays for Donald Russell (Oxford 1995); M. Hubbard, 'Pindarici Fontis Qui Non Expalluit Haustus: Epistle 1.3', in S.J. Harrison (ed.), Homage to Horace: A Bimillenary Celebration (Oxford 1995). But, while there is much of value in existing commentaries (especially Kiessling-Heinze and Préaux), a proper modern commentary has been seriously needed.

M(ayer)'s eloquent 52-page introduction treats: 1. the 'epistle' as a literary form; 2. Horace's career; 3. addressees and date; 4. poetic style; 5. themes; 6. the book's organisation; 7. textual transmission.

1 provides a fair general contextualisation, though it neglects ancient epistolographical theory (cf. Harrison 1995), and, on the question of the Epistles' fictionality, argues: (3) 'if Horace's verse letters were genuine and spontaneous they would cease to be imitations and, in ancient eyes, lose their status as literature'. The Epistles' fictionality is (no doubt) certain, but requires more than circular argument.

2 is an agreeable sketch (more fully developed in M.'s 'Horace's moyen de parvenir' in Harrison 1995), though angled in support of M.'s view that the Epistles outline a dual programme of spiritual and social self-improvement illustrated by H.'s own career and behaviour.

3 claims that, Maecenas excepted, H.'s addressees all seem to have been young men with their careers in full flood, a reasonable general perspective, which, however, hardly accounts for Albius (IV [below]), Fuscus (X), or H.'s bailiff (XIV). Otherwise, this discussion is disappointingly minimalist (considering the work done on the addressees' relevance in the Odes), bland (painting an over-cosy picture of the relationship between H. and Maecenas in I, VII and XIX; cf. now R. Seager in N. Rudd [ed.], Horace 2000: a celebration (Essays for the bimillenium) [Ann Arbor 1993], 34-35) and sometimes untenable in detail (e.g. on Albius in IV [below] and Lollius in XVIII [9]).

4 is an enormously strong section, which will be very useful alike for students and scholars and which I personally found very illuminating.

5 treats the crucial interpretative questions: are the Epistles morally serious and are they in any sense philosophical? To the first, M. answers (correctly) 'yes' ('the central issue of the collection ... could be summed up in the phrase recte vivere' [39]); to the second, a qualified 'no' (cf. M.'s 'Horace's Epistles 1 and philosophy', AJP 107 [1986], 55-73, and, similarly, Rudd, 'Horace as a moralist', in Rudd 64-88). This position seems to me to rest on numerous misconceptions, distortions and omissions (below).

By modern standards of book-analysis, 6 is minimalist and a distinct disimprovement on G. Maurach, 'Der Grundriss von Horazens erstem Epistlebuch', AC 11 (1968), 73-124, and M.J. McGann, Studies in Horace's First Book of Epistles (Brussels 1969), 33-87, whose views have been developed and refined by intervening scholars (e.g. Johnson 66-71). Some useful additional points are made in the brief interpretative summaries appended to the commentaries, although there are discrepancies between these and the introduction (e.g. on the question of H.'s consistency M. is himself inconsistent [218; 51]).

7 is very brief. In practice M.'s text is founded on those of Wickham and Garrod and Shackleton Bailey. In the commentary textual discussions are infrequent and perfunctory, sometimes unreasonably so (e.g. on the vexed 2.31).

In the 188-page commentary M.'s policy is 'explication from the editor's point of view alone, with only infrequent reference to alternative interpretations' (viii). This saves space, promotes clarity and brings the reader closer to the text. It is therefore justifiable, especially in a series such as this, provided the commentator is reliable. Nevertheless, the near-total absence of bibliographical references entails some loss of utility, and the less satisfactory the commentary, the greater that loss.

As section 4 of the introduction and M.'s general scholarly reputation would suggest, the commentary is excellent on grammar, syntax and metre, in which areas it will be indispensable. There is also much sound and detailed literary criticism, derived both from existing discussions and from M.'s own observation, which can be acute. In my view, however, M. misses an enormous amount in key areas, including: verbal, thematic and structural relationships (both within and between poems); verbal wit of all kinds (including name and etymological plays); analysis of thought; philosophical and literary allusions; and the relationship between poems and addressees. I choose as examples the poems where M.'s treatment seems weakest.

Epistle I

While M., inevitably, notes many and varied philosophical allusions, he persistently minimises the centrality of philosophy to Horace's/'Horace''s project, however wittily and ironically launched. (I write 'Horace''s, because many of those who deny the book's philosophical status impute to their opponents a crude biographical model, to which the latter, nowadays anyway, are not committed.) Thus (according to M.) versus 10 denotes lyrics only (n. ad loc., though the other view is taken on p. 110); decens 11 may have been devised to avoid the Stoic term decorum; H.'s refusal to enlist under any particular philosophical master (14) indicates non-commitment to philosophy as such; hospes 15 only 'may hint' at Aristippus ap. X. Mem. 2.2.13; virtutis verae 17 'is less philosophical than Roman in sense, and refers to a man's physical power in public life'; 19n. merely paraphrases Aristippean doctrine; his elementis 27 refer only to H.'s moral reflections hitherto (11-12); soler 27 evokes only a note on elevated diction; 32n. says nothing about philosophical 'road imagery'; libello (37) refers primarily to 'booklets' of spells, secondarily to books such as Homer; at 68-9 'defiance of ... Fortune ... was a goal of all the philosophical sects'.

On the contrary, 10 rejects all verse for the concerns of 11 (versus can be used of poetic sermo; any verse can be ludus; the polar opposition between 10 and 11 precludes an allusion to non-lyric verse in 11), thus creating multiple ambiguities whereby the Epistles are simultaneously not-poetry and poetry, not-philosophy and philosophy, the latter ambiguity being itself a philosophical ambiguity (words and writings are simultaneously nothing compared to philosophical action and a necessary preparation for it). 11 concerns not just right behaviour in general but philosophical concepts. The meaning of verum is defined by virtutis verae 17 and decens is specifically Panaetian, the participial form actually being closer to Panaetius' PRE/PON than Cicero's decorum. The Panaetian colouring of decens and cognates is confirmed as the book precedes (McGann 10-12; on Panaetian PRE/PON see further P.A. Brunt, PBSR 43 [1975], 13; 32-35). Astonishingly, M. first mentions Panaetius at 7.98n. (one of only two references) and without acknowledgement of McGann's important investigation. H.'s non-commitment is certainly contrary to the philosophical norm but not unparalleled (e.g. Demonax), and his words directly recall the Academic non-commitment of Cic. Tusc. 4.7 sed defendat, quod quisque sentit; sunt enim iudicia libera: nos institutum tenebimus nullisque unius disciplinae legibus adstricti, quibus in philosophia necessario pareamus, quid sit in quaque re maxime necessario probabile, semper requiremus. The specific allusions of 18-19 guarantee the Aristippean allusion of 15. Virtutis verae must be philosophical, because of the Stoic flavour of the rest of the line, of the allusions in 16, of the contrast with Aristippi ... praecepta (18), of Virtus 41 (which even M. concedes to be philosophical) and of veraque virtus 18.8. 19 alludes directly, via the yoking metaphor, to Aristippus' famous dictum E)/XW, A)LL' OU)K E)/XOMAI (D.L. 2.75). His elementis refer to the whole project of writing about philosophy, consoler to philosophy's well-known 'consolations'. The booklet will bring moral recovery if thrice re-read in a pure spirit; H.'s Epistles respond to the promptings of a Socratic inner voice in cleansed ears (7); they are a store-house for future philosophical improvement (12); any of H.'s listeners can be cured if he lends a patient ear (40); it is H. himself who praesens enables 'you' to stand free and erect against capricious Fortune (68/9). The booklet, then, is Epistles I (libellus would be a novel description of Homer!), the liber which will be despatched at the end of the collection (XX). 68-69 have a specifically Stoic colouring.

Altogether, I broaches a specifically philosophical project, sketches alternatives in terms of recognisable philosophies with orthodox Stoicism as the ideal, introduces H. himself in the Socratic role of philosophical mentor, and claims Epistles I as a philosophical text, which will benefit its readers philosophically.

M. is also insensitive to philosophical detail, such as the witty oxymorons agilis fio 16 and mersor civilibus undis 16 (so much for Stoic EU)/ROIA [Maurach 84 n. 33; Macleod, JRS 69 (1979), 22]; H.'s floundering is intensified by the echo of Cat. 68A.13 merser fortunae fluctibus [cf. Woodman, PCPS 29 (1983), 102]), or the implications of furtim 18, which cannot mean 'unconsciously' (cf. conor 19).

This last example illustrates a general phenomenon throughout the commentary: the narrow demarcation of 'meaning' through reliance on OLD and TLL. Thus M.'s gloss on spectatum 2 as 'proven' annihilates the contrast with latet 5 and the 'inverted' ring structure created by te respicientis 105, as well as the link with VI, where spectare and related words are applied to false values. Other structural/thematic/verbal correspondences requiring discussion include: dura ... custodia 22 ~ custos 17 (with rigidus A)PO\ KOINOU=); agendi 24 ~ agilis 16 and agendi naviter 24 ~ 15-16 (harbour imagery; cf. R. Maltby, A Lexicon of Ancient Latin Etymologies [Leeds 1991], s.v. navus); regam 27 ~ rex eris/si recte facies 59-60 ~ rex ... regum 107; deponere 35 ~ pono 10 ~ compono 12; ne cures 47 ~ curo 11; roget 70 ~ rogo 11 and roges 13; fugiam 72 ~ fugere 41; curatus 94 ~ curo 11; curatoris 102 ~ curatus 94; rerum tutela mearum 103 ~ tuter 13; rides 95 ~ 97 ~ 101 ~ ridendus 9 and ludicra 10 (Maecenas laughs at exactly the wrong things).

M. also misses important points on a general lit. crit. level, e.g. the gladiatorial and old-horse metaphors (2-9) burlesque the lyric themes of militia amoris and the chariot of poetry; in 12 the metaphors for poetic composition remain active, thereby reinforcing the ambiguity (both poetical and philosophical) of the Epistles as texts. Finally, M.'s notes on 1 and 105 ignore the steel behind H.'s rejection of Maecenas' request for lyric and for H.'s continued participation in his life of pleasure (steel implicit e.g. in the relationship between 103 and 13).

Epistle II

M. misses, or underplays, detailed links between I and II, e.g. relegi ~ ter ... lecto 1.37 (which does not make Homer the 'solution' to 1.37; rather, the reading-matter changes); immersabilis 22 certainly recalls, and contrasts with, 1.16; in cute curanda 29 ~ 1.94 (Maecenas ~ Phaeacians); all cura-references (29, 31, 39, 49) ~ 1.102, 94, 47, 11, delineating the different sorts of curae, moral, physical, positive, negative etc.; 52-53 ~ 1.28ff., 1.40, 1.7; puro 67 ~ pure 1.37, ~ purgatam 7; 68 puer ~ 1.59-60.

More important, he does not register the fundamental structural parallelism between the two poems, which maps out the geography of the whole book. I deploys a basic philosophical polarity between Stoicism/Virtue/constancy and Aristippus/adaptability/pleasure (15-19), and concludes by advocating a sort of compromise between the two: viz. an ironic description of the Stoic sapiens (106-8). II deploys a basic polarity between Virtue/wisdom/Odysseus (17-26) and pleasure/folly/Phaeacians (27-31) and concludes with a sort of compromise between the two (cessas 70 ~ cessatum 31 [or whatever form is read]). Again, M. minimises the formal philosophical element by failing to note that the categories of 3 are clearly Panaetian (cf. Cic. De off. 1.9; 3.7-10), as mediated by Lucilius 1329-30 (also unremarked), and that H.'s personal compromise (nec tardum opperior nec praecedentibus insto 71) pointedly subverts a dictum attributed to Aristotle (D.L. 5.20 E)RWTHQEI\S PW=S A)\N PROKO/PTOIEN OI( MAQHTAI/, E)/FH, E)A\N TOU\S PROE/XONTAS DIW/KONTES TOU\S U(STEROU=NTAS MH\ A)NAME/NWSI) and highlights the question of philosophical PROKOPH/profectus.

II is a corner-stone of M.'s claim that H. ultimately finds moral wisdom in poetry rather than philosophy. But philosophers themselves can argue this, particularly in protreptics (Zeno, Chrysippus and Plutarch wrote such works). It is also misleading to emphasise that the case is put 'suggestively early ... in the collection' (41): rather, (a) as in I, H. is still 'setting out his stall'; (b) if the case were decisive, all subsequent epistles would be redundant. Nor is it true that 'nowhere ... does H. press a philosophical text upon his addressees' (124): Epistles I is itself such a text (above) and cf. 18.96-103 below. And the formulation 'the epistle is a protreptic to poetry' (124) is odd; it is a protreptic to v. 3: that is, to ethical goals expressed in formal philosophical language.

Epistle III

At the Horace Bimillenary Conference held in Corpus Christi College, Oxford, 14-15 September 1992, M. and the reviewer both heard Margaret Hubbard's important paper (to appear in Harrison 1995) on this dense and difficult poem. The paper has left no trace on M.'s treatment, except for an afterthought in the Addenda (275). The following comments are inevitably influenced by Hubbard's insights, but observe some decorum in pillaging them and offer substantial independent observations.

M. misses a great deal and does not achieve a coherent interpretation. Given the bee/flower poetic imagery of 21ff., Flore 1 must be a significant name. The first six lines make ironic play with the notion that the studiosa cohors is seriously engaged in military affairs, hence M.'s note on militet 1 ('need not entail fighting, cf. militiae 'abroad') is exactly wrong, operum 6 covers both military and poetical 'works', and diffundit 8 has double reference, as the verb can be used of removing a river barrier by diverting it into several streams (Hubbard). To translate curo 6 as 'am interested in', curarum 26 as 'cares' and curae 30 as 'a subject for concern', 'dear to' obscures the logical connexions between the three occurrences, as well as the connexions with the various sorts of curae of I and II (above). Titius' Pindaric efforts (9-11) are implicitly criticised (pace M.), since (a) the contrast between longum in aevum 8 and brevi 9 works against him; (b) his lofty ambitions find a parallel in those of Celsus (below), who is explicitly admonished (15); and (c) in this very poem H. himself is demonstrating what can be done in Latin with Pindar (Hubbard, who adduces several key Pindaric allusions hitherto unspotted). Celsus, another significant name (as in 8.17), is clearly aiming too high (tangere 16 of touching the untouchable). The bee/flower imagery of 21ff. echoes V. Geo. 4.96ff. (D. West, Reading Horace [Edinburgh 1967], 34). Whether in law or in poetic composition, Florus (by contrast with the others) will win the victor's prize, though if he could only relinquish frigida curarum fomenta 26, he would be able to follow where caelestis sapientia led, a task which all, small and great alike, should speed, si patriae volumus, si nobis vivere cari 29.

Exegesis of 26-27 is extremely difficult and failure to solve these lines is hardly a sin. Nevertheless: (a) M.'s interpretation of 26 is: 'the chilling compresses your cares apply' - 'Florus' cares - public business, perhaps love ... and the pursuit of gain - chill his ingenium' (22); on this reading the 'cares' are Florus' activities (apparently not including poetry) and his ingenium is 'his full potential both as a poet and as a man in society'. (b) M. takes caelestis sapientia 27 as 'understanding', 'divine' because reason is divine. (c) In 29 he sees a disjunction between the spheres of interest of the ampli and the parvi (M.: 'the ampli will attend to the state, we parvi must look after ourselves').

These interpretations are untenable. (a) It is hard to see why public business etc. should be 'cares' while poetry is not (seu ... seu ... seu 23-24). True, the victorious ivy is proper only to poetry, but it can be applied metaphorically to other activities, especially civic, and it logically must be here: Florus' ingenium is such that he will be victor in whatever sphere he chooses. Moreover, since he will win (feres 25 - simple future) the victor's crown in poetry (thereby surpassing all his friends, whose poetic activities are variously flawed), how can his poetic ingenium be regarded as chilled?

(b) Caelestis sapientia is something which everybody should pursue, not just Florus. hoc opus, hoc studium glosses 26-27. Now hoc opus, hoc studium ~ operum ... studiosa 6, and similarly the image of caelestis sapientia as a general to be followed ~ militet ... / Claudius Augusti privignus 1-2. Does this mean that not only H.'s young friends but absolutely everybody, small and great, should abandon all opera mentioned (military opera, service under Tiberius, poetry, and the law), and presumably all conceivable opera, for caelestis sapientia, whatever precisely it is? Surely not, for all normal life would come to a halt. Thus caelestis sapientia is not something that supplants the other opera; rather, it is something that can be deployed in all the various opera, which it redefines by lifting to a higher plane. Thus e.g. Celsus, who is aiming 'high' in the wrong way, would benefit from caelestis sapientia: celsus in a bad sense is redefined as caelestis (the words are etymologically linked: Maltby s.v. celsus).

Hence the usual rendering, 'heavenly philosophy', can stand, provided it is taken not to refer to full-time intellectual study but to the wider phenomenon of moral wisdom. M.'s objection that '"philosophy" ... cannot be expected of young men' has no force: cf. 1.24-26 id quod/aeque pauperibus prodest, locupletibus aeque,/aeque neglectum pueris senibusque nocebit, a passage which looks directly parallel. Further support for an allusion to 'philosophy' comes from indomita cervice feros 34 ~ tenera ... cervice 2.64 ~ nemo adeo ferus est, ut non mitescere possit,/si modo culturae patientem commodet aurem (1.39-40), in both of which 'philosophy' is the cure.

(c) Caelestis sapientia promotes a state of affairs where all, small and great, can be dear both to themselves and to the patria. It is wrong to disjoin the spheres of the parvi and ampli: the parvus H. has been exhibiting cura for state (1-2, 7-8) and ampli (6), as he will also for ruptured friendships between individuals (30ff.), and as the ampli Florus and Munatius should for each other (30ff.). 29, then, emerges as the key proposition of the whole Epistle, for which it is tempting to seek a philosophical grounding, especially given the reference of caelestis sapientia itself. There is one: the philosophical doctrine that true virtue involves concord/friendship both with oneself (nobis ... cari does not imply, banally, 'we naturally love ourselves' [M.], but that proper self-love is a philosophical goal) and with one's fellow men. This doctrine, while widely disseminated, is found particularly in Cynicism and Stoicism (cf. e.g. R. H_istad, Cynic Hero and Cynic King [Uppsala 1948], 107-115; M. Schofield, The Stoic Idea of the City [Cambridge 1991], index s.vv. 'friendship' and 'concord'; the material in M.'s note on 18.101 is relevant). For H. Panaetius is an obvious source (cf. esp. Cic. De off. 1.50-58). Perhaps, then, caelestis has a further resonance: bees, exemplars par excellence of the cooperative virtues, have a divine nature (Geo. 4.220ff.).

Given all this, the frigida fomenta are something like 'the cold/ineffective alleviations/consolations for your concerns'; the 'concerns' are what Florus does (law, poetry), the 'alleviations' false glory arising from success. Victors wear crowns of ivy (25), things applied from the outside. 'Cold compresses' are also external applications and ivy itself is 'cold' (Serv. auct. Ecl. 8.12 haec herba nimium frigida est). Extremes of temperature (frigida 26, calidus 33, even perhaps 26 itself, if felt as an oxymoron) promote disharmony. Only concord unites, and it unites: individuals within themselves and with their fellow men, whatever their different concerns and abilities; the great and the small; the public and the private. Note the parallel with II: 2.6-16 concern discord in Greek and Trojan camps; III is set (however facetiously) in a military camp.

Epistle IV

M.'s treatment of the addressee and his place in the poem is unconvincing. Agnosticism about the identification of Albius with Tibullus is frivolous: see NH on C. 1.33, P. Murgatroyd, Tibullus I (Pietermaritzburg 1980), 2-3 and Ball 1994 (above). The etymological pun (Albi ... candide), which is a certainty, not a 'perhaps', has a double function: compliment of Albius' 'white' behaviour as iudex of H.'s sermones (which must be the Satires, not 'all H.'s hexameter writings on moral matters' [M.]), and preparation for the idea that, despite his intrinsic 'whiteness' and all his advantages, his present behaviour is rather gloomy, in contrast to H.'s own, also described by a colour/light word: nitidum 15. (Note the piquant reversal of the colour contrast in X, where the 'dark man' [Fuscus] is 'light'-hearted.)

As in 6.68 candidus also approaches the English 'candid', with the implication that, just as Albius has been H.'s liber amicus in relation to his poetry, so H. has the duty, and the right, to be Albius' in relation to his way of life. (I suspect also an implicit etymological allusion to Flaccus in porcum 16 [cf. Epod. 15.12; Sat. 1.9.20; 2.1.19], emphasising that by contrast with Albius H. is true to his name and nature). Candidus can perhaps also be etymologically connected with dare (7; Maltby s.v. candidus) and is certainly often applied to gods (6-7). All this intense etymologising and verbal playing suits a poem to Tibullus, master of that art (F. Cairns, Tibullus: a Hellenistic Poet at Rome [Cambridge 1979], 90-99).

Though the tone is teasing (Tibullus can surely better the poetry of a dead assassin of Julius Caesar), there are philosophical resonances. That (pace M.) sapiente bonoque 5 is philosophical language is proved by the echoes of 2.17, 1.41 and 1.11 (Albius is doing the serious philosophical thing, as H. himself was when he renounced poetry) and by the contrast with H.'s own porcine and Epicurean behaviour (15-16). Thus the poem reworks the central philosophical contrasts of I and II: virtue/Stoicism -- pleasure/Aristippeanism/Epicureanism. The Epicurean element is integral: M. misleadingly describes the sentiment of 13-14 as 'a moralist's commonplace'; in context and with the explicit 16, it must be read as a direct translation of a famous dictum of Epicurus (fr. 490 Usener O( TH=S AU)/RION H(/KISTA DEO/MENOS H(/DISTA PRO/SEISI PRO\S TH\N AU)/RION). Similarly, M.'s contention that the emotions of 12ff. 'need not describe Albius, they are common to all men' ignores the verbal links between curam 12 and curantem 5 (Albius' strong philosophical preoccupation is itself a 'bad' cura) and between grata 14 and gratia 10 (Albius' gratia does not seem to give him much pleasure). Nor is it unduly 'biographical' to detect a link between the emotions and state of mind that H. attributes to Albius and the somewhat plaintive notes of Tibullan elegy (cf. Ball 1994).

M. also misses numerous detailed connexions with other Epistles: Cassi Parmensis 3 ~ III (reminding us of the dangers of political dissension); curantem 5 ~ III, II, I; sapiente 5 ~ sapientia 3.27 and di tibi formam,/di tibi divitias dederunt 6-7 (with probable additional etymological plays on di/ divitias, dederunt [Maltby s.vv. deus, dives]) ~ caelestis 3.27 (Albius has divine attributes but does not use them aright; he even has the skill [artem] to do so, but not the will); divitias 7 ~ mitte ... certamina divitiarum 5.8; spem 12 ~ mitte levis spes 5.8; 13-14 ~ 5.9-10 (Caesar's birthday allows a busy man to implement Epicurus' advice) and 11.22-23; pinguem 15 ~ pingues Asiae campi 3.5 and pinguis ... Phaeax 15.24; ridere 16 ~ 1.101, 97, 95 (laughter begins to find a proper context); porcum 16 ~ sus 2.26 (the young men risk turning into Phaeacians in Asia; the Phaeacians and their like in II are now revealed as Epicureans); Epicuri ... porcum also ~ V (where H. turns out not after all to be a vulgar Epicurean).

Epistle X

M. seems to me to miss not only numerous felicities of detail but fundamental points about the structure and argument of this delightful and ironic but toughly-written poem:

2 hac in re scilicet una/multum dissimiles, at cetera ~ excepto quod non simul esses, cetera laetus 50. The poem is ring-structured; the initial dissimiles implicitly challenges ancient philosophical assumptions about friendship and about those with whom the virtuous can associate; cetera pointedly changes reference at the end, stressing H.'s need of Aristius' friendship and of his presence to complete his laetitia and the final inadequacy of his own self-sufficiency. dissimiles / simul also interact with simul 8. dissimiles ~ simul 8 builds up the picture of H. the instant and apparently self-sufficient philosophical king once he reaches the country; 'instant' because philosophical conversion is instantaneous; 'king' because regno contrasts with the worldly 'kings' of 33, and the whole phrase vivo et regno echoes and contrasts with H.'s earlier failure at 8.4 (vivere nec recte nec suaviter). simul 50 ~ simul 8 contributes to the final demolition of that self-sufficiency.

M.'s note on vivere naturae si convenienter oportet 12 misses (i) the interaction with 8; (ii) that in Cynicising Stoicism and Cynicism 'living in accordance with nature' was held to entail a primitivist life and rejection of the city; (iii) that, consequently, the principle is here interpreted in absolutist terms. By contrast, the re-statement of the conveniens-criterion at 42 is (a) morally relativist (because it allows different interpretations according to the individual); (b) clearly Panaetian; (c) a pointed echo of 7.98 (where even M. concedes Panaetian color).

Epistle XIII

M. fails to note the ring-structure Augusto 2 (second line from the beginning) ~ Caesaris 18 (second line from the end), which (a) centres the poem on H.'s relationship with Augustus; (b) does for Augustus something of what H. does for Maecenas (prima dicta mihi, summa dicenda Camena 1.1; I ~ XIX), though Augustus is significantly 'second'. He also misses the nicely meta-literary quality of fabula fias 9 and of 16-18. In the vexed problem of 17 he is (I believe) right to insist that carmina = the Odes, but for the wrong reason. The argument (4) that carmina cannot refer to the Epistles because of the literary 'fiction that the letters are not poems' has no force, because (a) the Epistles are both not-poetry and poetry (see above on 1.12); (b) at 19.27 and 31 carmina is used of non-lyric poetry. Rather, a reference to the Epistles would be incompatible both with the internal indications of XIII (2, 4, 6, 13) and with the dramatic economy of the book (cf. XX and libello 1.37 [above]). M. also fails to contextualise XIII adequately within the collection: it is one of a group (XII-XIV) addressed to men variously subject to others; in evoking the crisis year of 23 (3), it contrasts with the triumphs of 20 and 19 (XII); it offers a sort of anticipatory defence (praemunitio) of the literary claims of XIX (especially ingenuis oculisque legi manibusque teneri 34).

Epistle XVII

M. seems to me to misinterpret this poem fundamentally. Against the notion that 'Scaeva' conveys 'gauche' (vel sim.) and that the addressee therefore temperamentally resembles Diogenes, he argues (a) that the name 'also means a favourable omen ... what is more H. uses the proper name for nobody in particular at S. 2.1.53'; (b) that Scaeva must be imagined to have written to H. asking for advice on quo ... pacto deceat maioribus uti 2. But: (a) the fact that a word can be used positively in certain contexts does not preclude its being able to be used negatively in others. (b) The hypothesis that Scaeva has written to H. depends solely on tandem 2 ('"please" is retained from Scaeva's implied direct question'), but tandem can function as well within the indirect question after scis. The decisive fact is that, if Scaeva is already convinced of the need maioribus uti and merely wants advice about how best to do it, then vv. 1, 3-5, and 6-42, especially utrius horum/verba probes et facta doce, vel iunior audi/cur sit Aristippi potior sententia 15-17, make no sense. Consequently, vv. 1-5 are ironic (an irony punctured by the brusque 15-17), Scaeva is actually an admirer of Diogenes, and he is appropriately 'gauche', just as Diogenes himself is ineptus 32. The debate between Aristippus and Diogenes is therefore a real debate, as indeed H. himself emphasises (17, 39), not a redundant demonstration of what Scaeva already accepts. There is also a link with XVIII: Scaeva is on the Cynic side of the debate between (alleged) complete self-sufficiency and social dependency, just as Lollius (liberrime Lolli) is on the Cynic side of the debate between libertas and scurrari within the context of social dependency.

M.'s documentation of the poem's philosophical motifs and his analysis of the argument are again lacunose. The decet-criterion (2, 23, 26) and persona-doctrine (29) are Panaetian. 10 certainly (not 'perhaps') recalls the Epicurean motto, LA/QE BIW/SAS (M. obscures this by omitting vixit from the lemma). neque divitibus contingunt gaudia solis 9 ~ non cuivis homini contingit adire Corinthum (Epicureanism offers a legitimate alternative to virtuous social dependency). rectius 19 interacts with regibus 13/14 and with recte petit 42 (and also with 1.59-60) to suggest that Aristippus offers one route to a sort of philosophical kingship. praesentibus 24 evokes the philosophical doctrine of contentment with 'things present' (characteristically Cynic, but applicable to Aristippus, inasmuch as he can play either role). 36 cannot simultaneously (a) provide 'a fancied objection to the stand H. has just taken' and (b) suggest that 'not everyone can succeed in associating with principes viri; pleasing them requires its own virtus' (it is at 37ff. that H. argues this). It is highly relevant that both Aristippus and Diogenes frequented Corinth, the former 'frequenting' Lais. In 38-42 M. fails entirely to see that the argument systematically subverts Cynic claims and terminology (Moles, PLLS 5 [1986], 44-45; to accusations of 'wrong perspective' [Rudd 1993, 69], the answer is that in a debate between Cynic and anti-Cynic positions it cannot possibly be insignificant that the latter deploys against the former words spoken by the dying Heracles when he had failed to maintain Cynic standards and succumbed to luxury, words which were part of the Cynics' own debate about Heracles' moral decline, and perhaps even written by Diogenes himself). Finally, the beggar (58-62) obviously recalls the Cynic type.

Epistle XVIII

M. misses, or under-interprets, many of the detailed links with XVII and with earlier poems: e.g. scurrantis 2 does not just 'forge a programmatic link between the letters', it redefines/corrects, the Aristippean role of scurror ego ipse mihi 17.19 (as 3-4 make explicit); the same applies to discolor 4 ~ color 17.21; tonsa cute 7 ~ 4.15 and 2.29 (there is a mean between Cynic-style maltreatment of the body and vulgar Epicurean self-pampering); veraque virtus ~ 6.15-16 and 1.17 (H. rejects the exaggerated pursuit of virtue); 13 ~ 1.3, 10 etc. (school imagery); 14 ~ 17.29 (dramatic imagery); pretium 18 ~ 17.42; expertis/expertus 86/7 ~ experiens vir 17.42; semper inops ~ 2.56; cupido 98, pavor ... spes 99 ~ 4.12, 6.5-14, 16.65f.; fallentis 103 ~ 17.10.

On philosophical matters, 18 pretium aetas altera sordet evokes the Cynic doctrine LO/GOS H)\ BRO/XOS and 103 fallentis has clear Epicurean implications (above). However, the key interpretative question (which has implications for interpretation of Epistles I as a whole), concerns the section 96-103 (inter cuncta leges et percontabere doctos,/qua ratione queas traducere leniter aevum etc.). Who are the docti? M. argues that the 'sense is unspecific, and points generally to the cultivated and well-read, not necessarily to the philosophical (though they are not excluded either)'; that Homer is particularly appropriate to Lollius, addressee also of II; and that anyway the advice to read is 'not typical of the philosophically committed'.

These arguments are forced. It is difficult to see Homer as a useful guide to the precise questions asked, which are standard philosophical questions. Since the questions resume themes treated in earlier poems, one of the docti is (ironically but implicitly) H. himself, qua philosophical teacher (in this, as in other respects, the end of the book recalls the beginning [~ 1.36-7]). Moreover, 17 lines later XIX begins Prisco si credis, Maecenas docte, Cratino. Maecenas is doctus in the sphere of poetry, not (surely) in the doctrina required in 18.96-103 (or if the latter, only because he has learned from Epistles I-XVIII). We are again back at the beginning of the collection. There Maecenas was associated with poetry as opposed to philosophy; here he is associated with poetic doctrina, not the doctrina of 18.96ff. One does not have to be expert in algebra to work out that the latter must = 'philosophy'. As for the atypicality of the advice to read, it is well-known that philosophers speak with two voices on this matter, as they must (see on 1.10-12).

104ff. revert to H. himself and his own choice of life, not permanent (cf. quotiens 104 and XIX and XX), but clearly preferred. That choice is implicitly Epicurean (103 ~ 104; 110 ~ 4.12-14), and for H. himself the final solution to the problems raised by 1.11-19. It is quite impossible to trace his fluctuating progress throughout the collection without recourse to philosophical concepts and terminology.

M. (I stress) is a brilliant Latinist, whose linguistic and metrical exegeses are excellent but whose literary interpretations are seriously under-powered. The phenomenon is not uncommon, especially, perhaps, in the UK, but seems to have two specific causes. M. simply does not analyse the poems hard enough, a failure due less to incapacity than to ideology. Also, a sort of exaggerated aestheticism ('frigidum P., mod. Palestrina, whence the name of the pleasure garden at Rousham in Oxfordshire' [111]) enervates his critical thought. Considering the varied greatness of the Epistles, this commentary is a major missed opportunity.