BMCR 2007.10.48

M. Valerii Martialis: Liber Spectaculorum

, , M. Valerii Martialis Liber spectaculorum. Oxford: Oxford University Press, 2006. lxxxvi, 322 pages : illustrations, maps ; 24 cm. ISBN 0198144814. $110.00.

Table of contents

Professor Coleman’s (hereafter C.) impressive volume has come to see the light after several years of reworking and rethinking (cf. Preface v). I should observe, in the first place, that this commentary supersedes the older commentaries by F. Della Corte (in Italian, reprinted with some additions, but originally Genoa 1946) and by Filomena Fortuny Previ (in Spanish, Murcia 1983): neither of these can be called a critical edition, nor do they confront all the issues of such a difficult work, and on the whole they were both aimed at an undergraduate audience. U. Carratello’s own edition, which is current in Italian libraries, has no commentary, except for some useful discussions of debated manuscript readings in its preface.

Having handed the manuscript to the press in early 2006 (according to OUP catalogues it was first published in the UK in late October of the same year), C. was not able to take into account the second half of the huge and extremely valuable survey of recent Martial scholarship compiled by Sven Lorenz (first part: Lustrum 45 [2003], pp.167-277; second part: Lustrum 48 [2006], pp. 109-223). C.’s text and critical apparatus rest on a new collation of the manuscript evidence made from microfilms and photographs (Preface p. vi). C.’s critical text should be compared with U. Carratello’s edition (Genoa 1980 = Rome 1981) and D.R. Shackleton Bailey’s Teubner (Stuttgart 1990). Lindsay’s text (Oxford 1929) relied on an imperfect knowledge of the manuscript evidence as far as the florilegia were concerned, so that its numbering and ordering of Spect. is now to be considered out of date. The sketch of the manuscript tradition at pp. xxi-xxxiii, which also incorporates a discussion of the title and of all the interesting, if puzzling, manuscript headings, is of admirable clarity.

C. is the first editor to effectively use readings from the W ms., collated by M. D. Reeve. She does not however account for those other mss. that Carratello used (except for a generic mention of their common ancestor and contents), which are supposed to stem from the lost Florilegium Gallicum: it contained four poems, two of them lacking in the rest of our manuscript tradition. So, when we come to items no. 35 and no. 36, we still find in C.’s apparatus the generic indications ς or ‘Flor. Gall.’, where she should have given some separate account of those mss., for one of them in fact offers the right reading. Indeed, her conspectus codicum (p. lxxxvii) lacks the siglum ‘Flor.Gall.’

C. could not make use of the bulky but well-balanced Italian commentary on Book Three by A. Fusi (Hildesheim 2006).1 The latter could have been relevant in that Fusi sketches at p. 74ff. the manuscript tradition of the Epigrams with remarkable clarity, including some detailed notes on the most prominent Renaissance editions. On metrics, one misses references to R. M. Marina Sáez, La métrica de los epigramas de Marcial (Zaragoza 1998), and, on the manuscripts, to the papers by L. Zurli, “Gli epigrammi attribuiti a Seneca I”, GIF 52 (2000), pp. 185-221, and “I codici T ed R di Marziale”, RFIC 129 (2001), pp. 51-56. Here Zurli argues that, as a general rule, the scribe of T meant to supply and complete the imperfect work done some twenty years before by the inexpert copyist of R: this is shown throughout the whole oeuvre until Book 13, especially if we notice that in the ca. 50 instances, where R exhibits a couplet only of the original text, T has the whole epigram, but there are no examples of the opposite. This observation could bear some interesting implications. Nevertheless, none of the epigrams in Spect. is contained in either R or T. On the whole, C.’s text can be called a conservative one: its major virtues lie in the efforts C. has made to explain many difficult passages, rather than in printing a completely new text exhibiting fresh or original conjectures. In fact, in my opinion, C. could have been more selective with regards to some proposals that have found their way into the latest editions. I offer a discussion of some of these below.

The variety of topics treated in the General Introduction (pp. xix-lxxxvi) highlights the broad extent of C’s efforts. As is widely known to specialists, research on the book we used to call liber de spectaculis is peculiar for a number of reasons: its thematic singularity, not exactly paralleled by any extant work from Classic Antiquity; its place in the evolution of Martial’s literary activity, which concerns the title as well; and its manuscript tradition, which makes it pose a tough challenge to scholarly ingenuity. Critics agree, indeed, that Martial’s other books, by contrast, are on the whole very well transmitted.2 C., on her part, engages all these, and many other issues, with exemplary insight and a truly impressive command of the scholarship (especially in English and German). Her major achievement, however, consists in having shed light on so many of the Realien that take up the lion’s share of the liber : see for example her comments on the Seven Wonders at Spect. 1 (pp. 3-10). After reading C.’s page on that subject one is no longer compelled to browse through dictionaries and (Real-) encyclopedias. Some purists might view such self-contained, if occasionally verbose, commentary entries with disfavor, but in truth why compel the reader to continually seek other resources? Moreover, nobody is going to pass over absent-mindedly the great number of plates, none of which is merely decorative; they are often invaluable (and sometimes the only) aid to the student keen to fully understand a given passage: see Plates 24, 28, and 29, for instance.

C. discusses (pp. xxxiii-xlv) the characteristics of this unique collection, and the correspondences among the spectacles depicted in M.’s book and other ancient accounts that refer to the inauguration of the Flavian Amphitheater, pointing out relevant similarities and differences. In discussing the identity of the ‘Caesar’ who is the addressee of many poems in this collection C. adopts a line of great circumspection. She is inclined to regard those epigrams where details emerge that are in contrast with a date under Titus as Domitianic in their original composition; basing her argument on a quadrans issued by Domitian in 84 or 85 showing a rhinoceros as emblem of some games displayed by the Emperor, she considers that ‘it seems highly doubtful that Domitian would have minted a series depicting his recently displayed rhinoceros if one had been shown a few years earlier under Titus’ (pp. liv-lv). This issue has obvious implications for the composition of Spect. 11 and 26. Since C. believes that at least these two poems are of Domitianic date, she concludes that it is ‘conceivable’ that Martial may have later assembled a collection intended to honor both Emperors. In fact, some of the alleged hints at a Domitianic date of a given poem might be countered. For instance, epigrams 14.181 and 14.53 are no witness for a later composition or publication of Spect. Quite the contrary: I think the poet would not have repeated himself publishing some nearly identical pieces in the same year (84 or 85), if the argument for a “second” or a comprehensive edition of Spect. under the early years of Domitian are indeed to be accepted. It is true that both pieces of evidence, poem 14.53 and the “rhinoceros” quadrans (reproduced at p. lv), support each other, especially if we postulate those games to have taken place in Autumn 85 and Apophoreta to have been published during the Saturnalia of the same year (‘nuper spectatus’, says Martial 14.53.1), but nothing compels us to believe that Domitian would not have marked a series of ludi with a symbol, the rhinoceros, already used by his brother five years earlier; nor that those two poems on the rhinoceros should be attributed to a later date on the basis that Domitian chose that animal to star in his own games.

As to the title itself ‘we can legitimately postulate Liber spectaculorum as a name for our collection’, on the model of ‘Xeniorum … libellus’ which is the title for Book 13 as attested by the poet himself; C. admits, however, that ‘it may have had no title at all, as Josephus’ Bellum Iudaicum probably did’ (p. xxviii).

Commenting on stylistic issues, C. does not glorify “her” author but honestly notes where his diction is not up to his usual standards, but she obviously refuses to postulate a corruption where previous scholars attempted to “improve” the transmitted text on stylistic grounds only. One may only regret that she makes more than one reference to forthcoming papers, as if the most important remarks on this epigram collection were still to be made.

Here I would like to submit to the attention of the author, and of the reader, some remarks that the study of her critical edition have prompted, in the hope of encouraging further discussion. Occasional points of criticism are in no way meant to detract from my great admiration for C’s achievement.

Epigram 4. Although one might have omitted many of the proposals recorded by C. in the apparatus, like ‘Gaetula’ (F. Leo), ‘cuneis’ (J.-M. Pailler), or ‘eculeis’ (P. T. Eden), C. does not discuss the appealing ‘traducta est ferulis’ (ablative, put forward by A. Zingerle, and independently by Th. Birt [ RhM 79, 1930, 309]), which has, at least, some basis in what we know about the informers’ parade: cf. Suet. Tit. 8.13 ‘flagellis et fustibus caesos’; perhaps one could add flagris to the conjectures already proposed. ‘Gyaris’, already proposed by some humanists, was later defended by A. W. Van Buren3 but I have my doubts on the use of the dative in the place of ‘in’ plus accusative. Any conjecture pointing to implausible lands like Gaetulia, Numidia, or Thule, that are not islands, is very likely to miss the mark. Carratello’s own ‘querulos’ is pointless, while Shackleton Bailey obelizes the corrupt word.

Epigram 5. I very much appreciated the note on the double entendre hiding behind the phrase ‘haec licet impensis principis adnumeres’: by exiling informers the Emperor has deprived himself of the revenues from the confiscated wealth of the condemned. The same explanation is found in Elena Merli’s note ad loc.4 Many scholars had thought that 4 and 5 could form a single epigram, but C.’s argument for separation (p. 57) is fully convincing. One may only regret that C. has bothered at all discussing a couple of different proposals of arrangement of this sequence.

Epigram 8. In evaluating Heinsius’ conjecture ‘nobilis Herculeum’, adopted by the most recent editors instead of the transmitted reading (‘nobile et Herculeum Fama canebat opus’), we should consider that in this epigram we have several elements which all seem to point to a humorous belittlement of Hercules’ feat against the Lion: these are ‘Fides taceat’, ‘nam’, and especially ‘iam feminea (…) acta manu’, a phrase that imitates Propertian and Ovidian passages where it bears clearly deprecatory nuances. In my view, therefore, it is the transmitted reading, instead, that provides a much better meaning and epigrammatic sting: ‘Fama used to sing of the lion, laid low in the desolate valley of Nemea, as a noble feat and worth of Hercules. Hush, Credulity of our fathers: we’ve now (even) seen a similar feat made by a woman’s hand’; here ‘nobile (. . .) opus’ has predicative force. I am preparing a brief note on this passage.

Epigram 9. C.’s apparatus is a positive one, so she should have reported that ‘saeuas’ at line10 is the reading of T as well as of W.

Epigram 11. C. rightly supports (p. 110) the interpretation of ‘quantus erat taurus, cui pila taurus erat!’ (v. 4) with a couple of passages where the rhinoceros is called ‘bos Aegyptius’ or ‘bos Aethiopicus’. There is no longer need, then, for a periphrasis to render the first taurus, such as many have used.5

Epigram 13. The rich discussion found in the commentary omits to mention a proverb upon which the pointis probably certainly based, the ‘adynaton of the wrong hunt’, a saying by which one deems an action or a request as impossible: cf. Plaut. asin. 99f. ‘Iubeas una opera me piscari in aere, | uenari autem rete iaculo in medio mari’ (‘You could as well order me to go and fish in the sky, or to fish with a sweep-net on the high sea’; the sweep-net, Italian ‘giacchio’, is a kind of net of conic shape once used in rivers or lakes).6 Professor Parroni has kindly reminded me of a comparable passage in Sen nat. quaest. 3.17.1: ‘Expecto ut aliquis in mari uenetur!’. For the stock phrase, upon which the adynaton plays, see Ov. ars 1.763, with A.S. Hollis’ comment (every kind of fish needs a different kind of net to be captured).

Epigram 16. I prefer here the ancient conjecture ‘pignore’, which used to be printed until Carratello chose to retain the paradosis ‘pignora’. Several reasons make me opt for the emendation,7 among them the awkward syntax and the pleonasm resulting in the form printed by C. At Mart. 4.75.7 the mss. split between ‘pignore’ and ‘pignora’ (see Lindsay’s app.). Moreover, the close imitation of Ov. Met. 9.684 ‘iamque ferendo | uix erat illa grauem maturo pondere ventrem’ (unfortunately not mentioned by C.), seems to me a hint that Martial wrote ‘pignore’.8

Epigram 22. One can reasonably doubt whether Nisbet’s ‘cornuto dente’ is genuine Latin, since the only attested combinations with this adjective are ‘cornuto capite’, ‘cornuto rostro’, and the like: this is not surprising, since ‘teeth’ are less likely to bear horns than a forehead or, say, a snout (the singular number would be awkward here, too). It seems to me the simplest solution (perhaps too simple?) to read here cornu maiore (with W. Gilbert): this is, if nothing else, much closer to the paradosis, and a similar clausula can be found at least in the later Marius Victori(n)us (aleth. 89: ‘cura maiore petitum’, scil. ‘corpus Christi’). One might also think of cornu grauiore (possibly corrupted to ‘grab(i)ore’ or the like: cf. the apparatus at spect. 12.6), that is ‘heavier’, in both physical and metaphorical senses. Some Romans seem to have been unaware, until the Christian era, of the real nature of an elephant’s cornua. One must bear in mind that the point here need not be in a callida iunctura at v. 3, but in the equation straw doll: bull = bull : elephant. In any event the last word (‘petitus’) should not be emended out of the text. Carratello prints ‘et ab ore petitus’, putting in two superfluous particles, while Shackleton Bailey obelizes the whole half-line..

Epigram 25. Here the bear ( ursa) introduced by A.E. Housman in line 2 still seems to be a die-hard presence. The text printed by C. is not good Latin. This is reflected in her translation, which I cannot fully understand. A simple solution, virtually identical with the paradosis (that reads ‘uersa miramur’ or ‘uersa mis amur’) except for the inversion, would be to read: ‘miramur? uersa uenit ab Eurydice’ (‘As for the fact that the ground, through a sudden gap, released Orpheus, shall we wonder? He has came back from Eurydice, once she turned back [to Hades]’). For the sense of ‘wonder’ conveyed by the various spectacula cf. spect. 2, 6, 8, and passim. No one should wonder if Orpheus has come from the nether world: he was coming back to us (and to the spectators in the Colosseum) after the famed, if tragic, episode involving his descent and the error that made him forfeit his wife. This very interpretation is defended by Carratello, pp. 50-53. For the question mark, it’s worth observing that interrogative particles are very often omitted with forms of miror both in prose and in poetry. Cf. e.g. Cic. Verr. 2.1.113 and many other instances; Liv. 34.5.12 ‘ceterum quod in rebus ad omnes pariter uiros feminas pertinentibus fecisse eas nemo miratus est, in causa proprie ad ipsas pertinente miramur fecisse?’; with different forms ( miraris) there are many examples in Cicero; furthermore: Hor. sat. 1.1.86; a dozen examples in both Senecas; for Martial cf.. 5.40.2, 5.73.3, 6.11.2, 6.89.7, 7.18.4, 10.84.1, 11.38.2, 11.57.1, 11.51.2, 13.70.1, and the closest parallel, 13.74: ‘Haec seruauit auis Tarpei templa Tonantis. | miraris? nondum fecerat illa deus’.9 As is demonstrated by poem 24, the most famous aspects of the legend of Orpheus had been enacted with great attention; besides, I cannot believe that the mauling of Orpheus perpetrated by the bear in 24.7 was really unexpected, as Carratello and others seem to imply. If an explanation for the inversion of ‘uersa’ found in the mss. is needed, one can think of the word as wrongly omitted by means of a saut du même au même (‘VErsa VEnit’) and later inserted in the wrong place. Obviously this would be the best solution, granted that ‘uersa’ be equivalent to ‘reuersa’, which I cannot prove. I would personally favour a different attempt. A fast survey of Martial’s métrique verbale in a representative portion of his poems shows that such a sequence, a spondaic word followed by a molossus, is quite rare, but occurs at least 5 times in Spect.,10 and 18 times in Book 1.11 Thus it seems conceivable to me to read the two words in the same order as found in the paradosis, with only a slight correction, thus obtaining a better syntax and word order: ‘uersum miramur? uenit ab Eurydice’ (‘Shall we wonder that Orpheus appeared turned backwards? He has come from Eurydice, i.e., from the nether world’). This conjecture was proposed as early as in 1980 by the Italian scholar Renata Fabbri. Della Corte printed ‘uersum miramur’ without interrogation mark.The corruption of an ‘u’, with a hyphen for the nasal, into an ‘a’, is very easy. The very reading of H ‘uersam is amur’ may hint at the first ‘m’. The participle would allude to the sad end of the best known version of Orpheus’ myth, freezing “Orpheus” (that is, the damnatus) in the moment when he has turned back to look at his spouse, who, however, vanishes immediately leaving him alone. This could be achieved by letting the actor emerge backwards on a stairway, or against a wall or similar furnishings that could hide his face from most of the public. Orpheus stands astonished as if he were petrified (‘non alite stupuit gemina nece coniugis’) in Ov. met. 10.64, a most famous passage.12 However, I must admit that such a device needs further justification, and that there is no evidence in Martial for ‘uersus’ = ‘auersus’. To sum up, while a definitive solution has not yet been found for this corruption in v. 2, if we bear in mind that Martial, in Spect., consistently uses the plural first person in order to ‘describe things from a spectator’s point of view’,13 as an editor I would not alter the verbal form ‘miramur’, especially to substitute Watt’s ‘accituram’, or other conjectures that would make two accusatives depending from one another.

Epigram 31. In my opinion, the plate C. includes at p. 229 wholly sweeps aside the lucky conjecture ‘parma’ that has been received into most critical texts since Schneidewin. It is possible, indeed, that a palm was set up in the arena with a precise function, that of signalling a formality concerning the end of a gladiatorial duel. As C. explains, the fighters cannot receive missio and must fight until complete surrender: the loser raises his finger. As described in the epigram, the emperor has to find an alternative to the missio demanded by the spectators, that cannot be granted them due to the leges established by himself. This would be the meaning of the palm in the arena. Carratello and Shackleton Bailey both have ‘parma’, which may now be dispensed with.

Epigram 34.10 ‘diues Caesarea praestitit unda tuba’ is a quite attractive proposal by U. Dubielzig (cf. C., “Launching into History”, JRS 83 [1993], p. 66). Unfortunately, at least in my view, the very presence of ‘tuba’ at v. 2 (‘Augusti labor hic fuerat committere classes | et freta nauali sollicitare tuba’) makes it unmethodical to insert a repetition here. The sense provided is not beyond criticism too: if the ‘unda’ “grants” the shows I would expect also the receiver to be mentioned; therefore I would not alter ‘tibi’ at the end of the couplet. Moreover, whoever defends this reading should maintain ‘Caesar’ too. Perhaps I might add to the large number of conjectures put forward something like ‘diues, Caesar, et hoc praestitit unda tibi’, with ‘et hoc’ to pick up ‘quicquid’. Thus the phrase would give a reasonable syntax and word order, if nothing else: ‘Whatever can be seen in both the circus and the amphitheatre, this too has been offered to you by the rich waves’. N. Heinsius’s (and Carratello’s) ‘Caesar, io’ introduces an exclamation out of its place, while Housman’s (and Shackleton Bailey’s) ‘id diues Caesar’ results in a broken rhythm.

Epigram 34.11. C. impeccably comments that ‘stagna Neronis’ must refer to real naumachiae, not to the pool banquet recounted by Cassius Dio (61.20.5). Here we find another notorious crux, but again I would not have weighed down the apparatus with a number of untenable conjectures. C. chooses to print ‘Teucri’, emending the transmitted ‘tigri’. This is a proposal put forward by Housman, who supported it with a passage from Valerius Flaccus (1.7-9), a phrase which is not pertinent. C’s own proposal domini (in the app.) is very lame. In emending this passage it seems clear to me that we start from tigri : the honest copyist of the archetype gave us an almost meaningless reading, which K emended to pigri since in later Latin both ‘taceantur’ and ‘stagna’ are likely to convey an idea of immobility. Hence the possible development: ‘tetri’ > ‘tigri’ > ‘pigri’. The emendation ‘taetri […] Neronis’ (where ‘taeter’ = ‘immundus’) was proposed by W. S. Watt,14 but earlier by Albert W. Van Buren, too.15 Indeed this conjecture is very close to the paradosis tigri, it gives us a relatively rare adjective (what may account for a corruption) and Nero had already been labelled as ‘ferus rex’ at Spect. 2.3: ‘inuidiosa feri radiabant atria regis’, a passage where the ‘stagna Neronis’ are mentioned as well. ‘Taeter, saeuus, dirus, ferus’ obviously are no synonyms but can be found substituted each other in poetry as well as in prose.

Epigram 36. C. has once and for all shown that #36 is just a fragment (p. 263), and that represents the closing sententia of a poem on a duel between gladiators (that is, monomakhoi).

The commentary is followed by a very intriguing sketch of the ancient quarrel about the source of the Nile (pp. 267-269, with a map). Martial writes ( Spect. 3.5) ‘deprensi flumina Nili’: C., who rightly translates ‘discovered at last’, does not actually believe that the poet is telling us news, or showing erudition commenting on a topic that was then en vogue, but that he is merely expressing an adynaton to say “people living at the end of the world”: ‘it will no longer be a mystery once the people who imbibe it reach the civilized world’; in her opinion, it seems to be implied by references by Seneca and Claudian that the source of the river had remained a mystery until the 4th century (see p. 269).

Among the very few misprints I would note the following: p. xi for ‘Bene e le Attività’ read ‘Beni e le Attività’; p. xii for ‘Goertz’ read ‘Goetz’; p. lxii for ‘interpersed’ read ‘interspersed’; p. 183 (mid-page) for ‘uersa is amur’ : read ‘uersam is amur’; p. 245 for ‘nun tigris’ read ‘nunc tigris’.

Even if some readers may regret that such an accomplished work cannot be considered definitive from the standpoint of the Latin text, everyone will agree that it has already gained the status of a cornerstone in 21st century Martialian studies.


1. See now the review by R. Moreno Soldevila at BMCR 2007.04.51.

2. For a review of emendations and other textual proposals put forward in these years see the survey by J. Fernández Valverde in J.J. Iso and A. Encuentra (eds.), Hominem pagina nostra sapit, Zaragoza 2004, pp. 247-270.

3. Cf. AJA 41, 1937, 650. He later defended this correctionon two more occasions.

4. Cf. M. Scandola and E. Merli, Marco Valerio Marziale Epigrammi, Milan 1996, i p. 120, fn. 7.

5. Cf. Scandola and Merli (i p. 125): ‘Che forza taurina in questa bestia’ eqs.

6. Here the iunctura ‘rete iaculo’ is not corrupt, as it can be seen by a comparison with the wording in Plaut. truc. 35. See F. Bertini (ed.), Plauti Asinaria, Genoa 1968, II p. 163.

7. Schmieder (1837) was certainly not the first to print ‘pignore’, since this reading is discussed in many XVIII century commentaries. I have tracked it back to V. Collesson (1680), but it may be more ancient.

8. Here ‘uixerat’ or ‘uinxerat’ read all the mss. but one.

9. I think, with A. Fusi (in his comm., mentioned in the text, p. 256), that adding an interrogation mark after 3.28.1 would make that epigram clearer (so ‘Auriculam Mario grauiter miraris olere?’).

10. Namely 1.8, 12.2, 19.2, 23.4, 26.12 (here I include some cases of ‘metrical words’).

11. Namely 12.4, 18.4, 20.2, 21.7, 25.7, 26.5, 43.3, 43.14, 51.4, 53.8, 62.4, 67.2, 68.3, 78.7, 100.2, 103.11, 111.4, 112.1.

12. The Georgics passage compared by Fabbri loc. cit. is not pertinent.

13. Cf. Spect. 2.7; 6.2; 8.4.

14. LCM 9, 1984, 120; quoted by Coleman.

15. ‘Osservazioni su alcuni testi letterari ed epigrafici’, RPAA 19, 1942-1943, 183-185; he thinks of taetri left out by a scribe by a homoiocatarcton, then reinserted in its place, though mangled as tigri. This is not an easy explanation.