BMCR 2003.07.20

Martial und das antike Epigramm

, Martial und das antike Epigramm. Darmstadt: Wissenschaftliche Buchgesellschaft, 2002. 168 pages ; 20 cm. ISBN 353415083X EUR 19.90.

I have to declare an interest: Niklas Holzberg, although not a colleague of mine at Cologne, is an acquaintance, and a dear one, from Munich. Our views on Martial differ in quite some ways, and I have always taken our disagreement as a prolific element of pluralism. I had a hard time reviewing his new book, but I myself am to blame, for I volunteered to review it. I wish to stress two things: (1.) None of the arguments below is ad hominem; (2.) the excessive length of this assessment is prompted by the fact that Holzberg’s new book basically deals with ‘all of Martial’ in some 150 pages, and by pointing to its (many, I must say) weaknesses and idiosyncrasies I aim to provide some ground for future research on Martial, who has recently become increasingly popular among classicists.

In the German-speaking world, especially among students of Latin, Niklas Holzberg (henceforth: H.) is well known both for his natural talent to produce very readable introductions to a great variety of authors and genres and for his ability to deliver highly entertaining lectures (not only in Munich). H. has published monographic introductions to Roman elegy, the fable, the ancient novel, Ovid, and fairly recently Catullus. Did I miss anything? Some of his monographs have seen a 2nd or even 3rd edition (if not more) and have been translated into English (e.g. his Ancient Novel, Ancient Fable, and Ovid), which is, not the least, owed to the fact that H. is a brilliant promoter of his books and, along with them, his views. Notwithstanding the at times unfavorable reactions to some Holzbergiana from mostly conservative (German) reviewers, the Bavarian’s natural PR talent surely is to the advantage of Classical Studies as a scholarly discipline in German academe and, at least as importantly, outside it. Conventionalists may feel that the author’s commitment to fun and entertainment is a bit too eccentric (or ‘not scholarly enough’), but H. is intentionally unconventional. Take, e.g., H.’s recent Catullus (Munich 2002), which can, and hopefully will, make us re-think a lot of Catullan questions; and yet, it is clear that the majority of scholars would hardly subscribe to the general picture that book puts forward. The same is true of the Martial, which is written in quite a similar vein.

Idiosyncratic approaches and/or views that challenge the communis opinio without any explicit indication that they do so might, however, very easily leave the readers, esp. the student and the non-Classicist in the lurch, i.e., the uninitiated recipient might take something for granted that is far from certain (and reproduce his/her ‘knowledge’ in exams). While the Catullus is not, I think, intended to be an introduction, the new Martial is. Regarding the 1997 Ovid (2nd ed. 1998, Engl. 2002), a much appraised book full of stimulating detail and acumen, R. Armstrong in her BMCR 2002.11.21 review rightly said that sometimes H.’s “statements can be rather too sweeping”, and that is precisely where the trouble starts for students and non-experts who read the Martial.

H.’s fascination with the idea that the books of the Roman elegists constitute ‘love-story sequences’ with a continuous development is well known from his Die römische Liebeselegie and recurs in the Ovid. Some similar systematization can also be found in the Catullus (e.g. pp. 19-23) and, as I shall discuss below, in the new Martial. It bothers the critical reader that this amazingly persistent bent of H.’s to sense ‘sequential developments’ in seemingly un-sequential units leads to an exceedingly centripetal explication, which seems to be reflecting our modern obsession with coherence and unification.1

In H.’s oeuvre, Martial is well represented. 25 years after O. Seel had published his rather unfortunate ‘appraisal’ of the epigrammatist (“Ansatz zu einer Martial-Interpretation”, A & A 10 [1961], 53-76), which is partly responsible for the snail-like advancement of research on Martial in Germany up until the 1980s,2 H. brought out his “Neuansatz zu einer Martial-Interpretation” ( WüJb 12 [1986], 197-215). In the best tradition of a Forschungsbericht, that assessment succeeded in putting straight most of the ‘facts and fantasies’ about Martial that had made their way into handbooks and minds. Shortly after that, H. published a slender ‘Einführung’, Martial (Heidelberg 1988), that was meant to put into practice some of the new approaches as laid out in his 1986 article. For sure, that was a good thing to do, since at that time there was no (however skeletal) up-to-date introduction to the master of epigram available in any language, and J.P. Sullivan’s milestone was yet to come.3

Now, another 25 years later, we have a second Martial by H., not just a remake or revised and up-dated edition of the 1988 volume but an entirely new book. In its change of approach it is many times more radical than that of H.’s second Liebeselegie (1990 vs. 2001), for which see J.L. Butrica, BMCR 2002.02.20.

In order to comprehend the fundamental shift in H.’s understanding of Martial, one needs to recap the principal line of thinking of the 1988 book. At that time, H. vehemently fought against all previous scholarship on the epigrams for having taken them as nothing but entertainment to amuse their readers (“reine Unterhaltungslektüre”, p. 7). In contrast, H. aimed to explain Martial’s “höheres geistiges Anliegen in Form einer klug getarnten Moral- und Gesellschaftskritik” ( ibid.). Consequently, he analyzes most of the erotic poems as ‘moral satires’ (p. 56), written by an author who has an “ernsthaftes weltanschauliches Anliegen” (p. 58). Therefore, H. interpreted the notorious poems that focus on a corrupt patron-client relationship (pp. 65-73) along similar lines; in that context, the separation of the 1st person speaker ( persona) from the poet himself (p. 67) was surely an important prerequisite in order to prevent Martial from becoming all too schizophrenic.

In short, Martial, in 1988, was a ‘social critic’ (pp. 72-73 and elsewhere), whose goal it was to teach his contemporaries how to behave properly. Following the Horatian ridentem dicere verum (p. 92), the epigrams thus functioned as a literary tool to achieve that goal.

Another hot issue was (and is) Martial’s relation to Domitian. Following especially J. Garthwaite’s problematic ‘safe criticism attempts’,4 H. in 1988 detected many hidden signposts in the (purportedly?) adulatory poems that ‘clearly’ revealed a number of ambiguities, so that Martial ultimately, turned out to be a regime critic (pp. 74-85). This belief (indeed it is nothing but a belief) met with disapproval in some reviews.5 The overall purpose of the 1988 book is stated unequivocally at its very end: H. wanted to contribute to a “dringend notwendigen Rehabilitierung” of Martial (p. 93). He may or may not have succeeded. Either way, one needs to bear in mind that H.’s first Martial is a palpable product of the Zeitgeist of the 1980s, a decade that was driven to discover ‘hidden criticism’ and subversiveness in so many areas of post-Augustan literature.6

What has become of that rehabilitation of Martial in H.’s 2002 book? It is totally gone, for better or worse. Right on the first page (p. 9), the author basically dissociates himself from himself, as it were, and thus consequentially from Martial, the alleged prosecutor of a moral decline of the society and the (no less alleged) smart critic of depraved Domitian. Today, H. confesses that his ‘first Martial‘ offered a picture of the poet as a “‘Klassiker’ des Altertums” ( ibid.) — a ‘classic’, however, not in Sullivan’s sense of the word but in that of “konservative Altphilologen” ( ibid.). Nothing perhaps changes as quickly as the Zeitgeist. Martial the critic has now undergone a metamorphosis into a Martial ‘who laughs and makes people laugh’ (“der […] lacht und zum Lachen bringt”, ibid.). This new picture most strikingly resembles that of H.’s Catullus. Even if we acknowledge (and we emphatically must) that enlightened literary criticism permits the coexistence of seemingly mutually exclusive results, H.’s 1988 and 2002 Martials cannot possibly coexist, unless we declare authorial intention to be completely devoid of relevance. Any self-adjustment is the result of rethinking and revision, and this process is superbly described by Malcolm Heath in his recent Interpreting Classical Texts (London 2002), 100-107. Here’s the drawback: “If our commitments are too open to revision, enquiry will be thwarted” (Heath, p. 107); in other words, interpretation must not become arbitrary. But how do we protect ourselves?

Let me now come to the details.

The new Martial, in terms of its contents, inevitably overlaps with the old one. There is a brief survey of ancient epigram to contextualize Martial within the history of the genre (ch. 1, pp. 19-32, vs. [1988], pp. 14-23 & 42-47 & 100-109). Some central issues are of course still present (the ‘imperial poems’, pp. 63-74, vs. [1988], pp. 74-85; the ‘client-patron epigrams’, pp. 74-85, vs. [1988], pp. 65-74; ‘sex’ [incl. aischrology, obscenity, etc.] (pp. 109-122 vs. [1988], pp. 48-58); the arrangement of poems within a book (see, e.g., pp. 37-39 and the index, s.v. ‘Martial, Zyklen’, vs. [1988], pp. 34-42 [where the book structure received a whole chapter]).

It may be inevitable that some parts of the new Martial are almost completely copied from the old one, yet the production of a new book offers its author the chance to come up with new material or at least the modification of the old. Take, e.g., the chapter on epigrammatic wit, “Witz”, pp. 86-97, which is for the most part identical with the chapter on Martial’s ‘epigrammatic technique’, [1988], pp. 24-34. Its short interpretations (not meant to be comprehensive) of the following epigrams reappear in 2002 (the changes being immaterial): 1.10; 1.30; 1.79; 1.84; 2.20; 3.17; 3.26; 5.43; 9.15; 7.75; 12.39. Item 1.7 ([1988], pp. 31-32), the famous imitation of Catullus’ Sparrow poems, has been moved to a different chapter, on ‘patronage’ (p. 74-75), where H. briefly discusses Martial’s relation with his patron-friend (and fellow-poet!) Stella.7 Three items, however, have been added: a) under ‘amphibolia / paronomasia’, 3.75 now joins 7.75 and 1.79 (p. 94); b) etymological wordplay in proper names is illustrated not only by 12.39 ( Sabellus — bellus — bellum), but also by the ingenious Palinurus -epigram (p. 95); finally 11.47 as a specimen of sound play through the ‘refrain’ ne futuat (p. 96).

There is also an overlap between the 1988 “Klient und Patron” and the 2002 “Patronat”, but it is confined to the chapters’ titles. In 1988, H. took the patron-client epigrams as epigrammatic critique of social inequality and injustice (pp. 65-73). The poet- persona, acting as a poor client, was viewed as a victim of bigheaded patrons — a self-victimization that H. interpreted along similar lines (p. 73). It is most noteworthy that that analysis formed part of a large section entitled “Zur sozialhistorischen Interpretation”, because this socio-historical approach is now not only gone but even rejected: “das, was wir über ‘non-imperial patronage’ in den Epigrammaton libri XII erfahren, ist sozialhistorisch kaum verwertbar” (p. 74). Consequently, H. questions attempts to detect any serious meaning concerning social matters or those of Lebensphilosophie (p. 80). Contrary to that, the 1988 volume contained an extra chapter on “Lebensphilosophie” (pp. 58-64), where H. reached the following conclusion: “Martials Weltsicht ist von ethischen Wertmaßstäben beeinflußt […]” (p. 63). It goes without saying that that inference was based on the “gesellschaftlichen Voraussetzungen” (p. 64). In contrast, H. now focuses on the fact that the relevant poems are subject to the specific ‘laws’ of epigram as a literary genre (p. 75). No doubt, generic components exist in any piece of literature; and yet, one must not underrate the social and historical context(s). Nobody would, in an overly simplistic fashion, regard the construct of the literary world as a mere equivalent of the non-literary reality, anyway. It suffices to say that the sheer quantity of client-patron epigrams in Martial cannot be explained simply as a result of some generic development or literary fashion.

The ‘sexuelle Thematik’ of 1988 has now become ‘Obszönität’. In the 1980s, H. took Martial’s view of sexual issues as congruent with the conventional Roman value system ([1988], p. 63), and the respective epigrams, according to H., communicated clearly shaped ethical standards (p. 58). H. now withdraws this position (e.g. p. 110, cf. p. 103).8 The poet’s intention (in case we are interested in it) in 2002 is ‘far away from moral criticism’ (“von Moralkritik weit entfernt”, p. 111). The two-liner 3.71, e.g., addressing poor Naevolus, whose culus so badly hurts, aims to make the readers visualize how “der ‘Schwanz’ des puer den ‘Arsch’ des Naevolus penetriert” (p. 112), that is, Martial wants to stimulate the ‘erotic fantasies’ of his recipients (p. 113). The vetula -poems fall into the same category (p. 114): Ligeia’s repulsiveness at 10.90 functions as a means of literary entertainment rather far from our modern taste. H. rightly points to Martial’s programmatic statements about obscenities in his verse, the first appearing as early as in the prose preface to Book I (cf. p. 37), and refers to poems such as 1.35, 3.68-69, 11.16 and others (pp. 110, 112-113). I agree that this literary-critical standpoint for the erotic epigrams is more suitable than taking them as a stance of the author’s critique of deviations from any sexual norm (pp. 114 and 118).

It is most sad, however, that H. comes to talk about intertextual issues only once in this chapter, namely regarding the well-known Uxor-vade-foras epigram, 11.104 and its relation to Ovid’s Ars amatoria. What he has to say draws heavily on Stephen Hinds’ brilliant Allusion and Intertext (Cambridge 1998, pp. 129ff.), but Hinds’ argument or its connection to H.’s own point (p. 117) remains arcane to me. Even though there is a separate chapter on intertextuality (for which see below), it would have been useful to incorporate more of intertextuality into the sex/obscenity discussion (e.g. Catullus, Horace’s Epodes), because then the reader would be much more able to understand the literary-historical context as well as Martial’s adaptation of generic components of epigram and (more importantly perhaps) other genres.

The section on the ancient epigrammatic tradition, previously containing only a sketch of the pre-Martial epigram, has been augmented by a discussion of Nachleben. H. treats ‘Martial and the Greek epigram’ at two different places, mainly because (as in 1988) he devotes separate space to a comparison between Martial and Lucillius. In scholarship, the link between Martial and pre-Neronian Greek epigram, especially the items collected in Meleager’s and Philip’s Garlands, has so far mostly been considered in very general terms. H. tentatively speculates that Martial’s collections may have been influenced in their structural arrangement by Hellenistic epigram books (p. 23). This is very uncertain.9 The new Posidippus, which H. mentions in this context ( ibid.), does not seem helpful.10 From the Garland of Philip H. briefly mentions Crinagoras’ panegyric poems (pp. 28-29) as precursors of Martial’s imperial epigrams. Agreed, but I would want to stress that the intertextual connection between the two goes well beyond this aspect and calls for further investigation. It may be worth noting that H. follows those who count not only Rufinus but also Strato among the epigrammatists predating Martial (p. 29). I’m not sure whether they have “die besseren Argumente”.11

Lucillius is dealt with, from a literary-critical viewpoint, under the heading “Intertextualität”. The section (pp. 100-109) can be read as a complement to its precursor. The main material (esp. the lists of epigrams cited but not discussed) still comes from W. Burnikel’s instructive book Untersuchungen zur Struktur des Witzepigramms bei Lukillios und Martial (Wiesbaden 1980). H.’s own focus is on narratological aspects and those of the author-reader relationship in Martial and Lucillius. The pairs analyzed in some detail are: Mart. 6.53 — AG 11.257 (as in 1988); 9.27 — 11.155; 11.18 — 11.249; 12.28 — 11.315. As before, H. omits the seminal article on Lucillius by L. Robert: “Les épigrammes satiriques de Lucillius […], in L’épigramme grecque (Vandoeuvres-Geneva 1968 = Entr. Fond. Hardt, 14), pp. 179-291.

The new section on post-Martial Roman epigram (pp. 51-62) comprises sketchy discussions of (the reception of Martial in) the Carmina Priapea, Ps.-Virgil’s Catalepton, the epigrams ascribed to the philosopher Seneca, Ausonius, the Epigrammata Bobiensia, and Luxurius. Two remarks may suffice. I agree that the Priapea postdate Martial (p. 52); however, I don’t think that H.’s reasoning is sound (p. 53). Speculations can be enticing, but is it seriously imaginable that the anonymous author of the Priapea aimed to produce a ‘monothematic’ collection that was meant to be a pendant (“Gegenstück”) to Martial’s Liber spectaculorum, i.e., that the power of the phallic god functions as a counterpart to the divine powers of the emperor?! Regarding the ps.-Senecan epigrams, one has to be careful if one wants to take the order of epigrams in the Vossianus Leidensis Q 86 as reflecting the ‘original sequence’ of items as their author had arranged them (contrast H.’s construct, pp. 56-57).

Martial’s attitude towards the emperor Domitian has always been one of the most disputed issues. The disagreement among scholars is in part the result of preconceptions and expectations of the modern reader: the judgment one forms of a poet’s opinion of ‘his’ government and, along with that, of his fellow-citizens’ living conditions both sheds light on one’s own historical context and mirrors the developmental stage of one’s academic discourse. A period of condemnation is thus almost inevitably followed by a reversal of that fashion. As indicated above, H.’s 1988 view ([1988], pp. 74-85) was trendy in this respect. His new assessment (pp. 63-74) is diametrically different, for his focus has now (rightfully, I think) shifted to an examination of the realm of the epigrams’ literary world and the multi-layered interactions between the individual items within the collection. As a starting-point, there is an analysis of the ‘imperial poems’ in the opening section of Book I, items 3-6, in which Domitian appears as a poetic persona fully integrated into the fictional world of the collection. Indeed, 1.5 (as a ‘reply’ to 1.4), with ‘Domitian’ acting as the speaker (which is the only incident of that kind in Martial), gives an important clue for comprehending the playful construct of Martial’s collections (‘I give you a sea fight [in the amphitheater], you give me epigrams. Methinks you want to be in the water with your book, Marcus.’ [tr. Shackleton Bailey]), as H. correctly notes (p. 65). Consequently, the much disputed hare-lion allegory of 1.6 and indeed the entire hare-lion sequence can now no longer be a specimen of the poet’s anxiety and ‘hidden’ critique of Domitian ([1988], pp. 76-79) but one of the transformations of panegyric into the speech-mode of epigram. It is important to note that ambiguous epigrammatic witticism and panegyric are not mutually exclusive: “Es besteht ja kein Anlaß zu der Annahme, die Zeitgenossen hätten die Schmeichelei […] entwertet gesehen […]. Panegyrisches Sprechen ist hier ganz einfach epigrammatischem und damit witzigem Sprechen untergeordnet […]” (pp. 66-67). Well said. As a matter of fact, the hare-poet need not fear the lion-emperor; the poet can safely compose his verse just as the rodent can playfully explore the lion’s gums. H. (p. 67 = [1988], pp. 77-78) may, however, be going too far in taking 1.6.4 to contain a sexual double entendre ( tutus et ingenti ludit in ore lepus, ‘the hare plays safely in the giant maw’). In that case, the equation ‘hare = poet’ would have shifted to ‘hare = poet’s penis’, and lepus as a phallic metaphor is not as well documented as one might think.12 The rest of this chapter gives a neat sketch of Domitian’s role in the remaining books (II-IX). The remarks on Book V are drawn from a piece by J. Garthwaite; those on VIII, esp. poem 21, from another one by L. Watson.13

Dominated by a discussion of Lucillius (see above), the intertextuality section (pp. 97-109) contains only two more examples of Martial’s often ingenious interaction with pre-texts: 1.32 and Catull. 85, and 2.22, where Martial ‘quotes’ from Ov. trist. 2.1-4 (pp. 97-99). Without doubt, there are tons of allusions or references in Martial, especially to the Augustan poets, and this is indeed a vast territory that is hardly explored systematically in any methodologically up-to-date study.14 H. does offer some further examples of intertextuality elsewhere (e.g., pp. 45 and 117), but he misses the chance to organize the material typologically.

While the old Martial was an introduction only to the dozen epigram books proper (Books I-XII) the new one comprises also the so-called liber spectaculorum (pp. 39-43) and the Xenia (Book XIII) and Apophoreta (Book XIV) (pp. 44-48).

H. points out that the generally accepted publication date of the liber spectaculorum (AD 80) is anything but certain (p. 40; contrast [1988], p. 11). That may be true, but in cases such as this, it seems best to subscribe to the most probable belief. If you don’t want to do so, the onus probandi is yours.

Also, the text of the Spectacula, as it has come down to us, is the product of excerption; the ‘original collection’, according to H., was a “Monobiblos” just like Virgil’s Eclogues (p. 40), which strikes me as a crude comparison. The assumption that the excerptor included in his selection only poems in elegiac couplets and that he left the original order of epigrams unaltered, is miles away from being cogent. Again, this needs to be proven, even if only by some probationes artificiales. I see no reason at all not to assume that the collection originally contained elegiacs only. The Apophoreta, as we have them, contain 214 elegiac couplets, 9 hendecasyllabic items; the Xenia consist of 125 elegiac couplets, 1 item in choliambics and 1 in hendecasyllables; that is, those books are metrically much different from Books I-XII. Is it so far from imaginable that Martial produced a book that consisted of elegiac couplets only? Catullus, for example, had reserved that meter for part III (some would say ‘Book III), the epigrammatic section, of his oeuvre. As for the arrangement of items in Spect. : yes, poems 1-3 are suitable as the prooemial sequence and may indeed be identical to the beginning of the original collection, but the excerptor can very well have grouped the rest or parts of it himself. Be all that as it may, the onus probandi is again on H.’s side rather than on anyone else’s. Let us wait for K.M. Coleman’s commentary.

The aesthetic appeal of the Book of Spectacles is surely alien to our current taste. And indeed, we must not measure its ‘quality’ from our 21st century western value system. H. has chosen one poem, Spect. 5 Lindsay = 6 Shackleton Bailey, dealing with the penetration of a woman by a massive bull, to illustrate Martial’s poetic intention. By equating the victim with Pasiphae, H. sweepingly argues, Martial may be playing with the threefold classification of narrative as found, e.g., at Auct. Her. 1.13 (p. 43). What he means is that the fabula of Pasiphae, originally neither true nor likely, now, since the audience witnessed a similar event in the theater, has won the status of historia. Narrative theory is indeed the background, but I am not sure if it also is the poem’s pre-text. Philip of Thessalonica, for example, wrote an epigram on a nightingale rescued by a dolphin ( AG 9.88 = 40 G.-P.); in it, the bird praises the dolphin and compares her encounter with the serviceable sea-creature to the Arion-myth: ‘Dolphins have always done unpaid oar-service to the Muses; the story of Arion is no falsehood [ οὐ ψεύστης μῦθος ]’ (tr. G.-P.). What we have in Martial (the verification of a myth by reality) is a generic element adapted for a specific context, as O. Weinreich ( Studien zu Martial [Stuttgart 1928], 34 and 72) has already observed. It still needs to be explored to what extent epigram exerted influence on other genres/texts, e.g. Apuleius’ Golden Ass.15 At any rate, the reader may be led to believe that Spect. 5 = 6 is unique, which it is not (cf. Weinreich at 29ff., 62ff.; K.M. Coleman, “Fatal Charades […]”, JRS 80 [1990], 44-73 at 62-64). In addition, take for instance the last of the ‘pregnant-sow poems’ in the same book, Spect. 16 Sh.B., where a wounded sow involuntarily gives birth to one single (!) piglet through the gash that has been inflicted on her: What a surprise that the offspring was able to immediately‘run’ as the mother died ( nec iacuit partus, sed matre cadente cucurrit, 16.3) — and how amazing that it’s just one piglet! The reality created by Martial’s fiction in his Spectacles is a construct in defiance of the non-literary reality, and it is important to realize that the sow-poems, too, are not independent of epigrammatic precedents either, among them — alas! — Philip of Thessalonica ( AG 9.311 = 51 G.-P.). See K.M. Coleman’s superb exposition in my Toto notus in orbe [n. 13], at 21-24.

As regards the date of the Xenia and Apophoreta, I agree with H. that Friedländer’s chronology cannot be verified beyond doubt (see BMCR 2002.08.38, par. 7), and that the position of those collections after Book XII is the result of posthumous editorial bustle (p. 44). However, it smacks of idiosyncrasy when H. argues that the redactor placed those books at the very end because of their ‘Saturnalian nature’, so that “wir uns am Schluß der Lektüre des Korpus zu den Saturnalien noch einmal ‘ganz groß eingeladen’ fühlen” ( ibid.). If I had been the redactor, I would have placed XIII and XIV at the very beginning (perhaps even using the same reasoning). Someone else might have dropped them altogether.

The Apophoreta have received a lot of attention from textual critics. Since Martial himself announces that the book consists of alternating pairs, a precious gift followed by one of lesser value (14.1.5-6), one is easily led astray if one uses the material value of an item to check whether the order of poems is sound. H. justly points out (pp. 46-47) that we also have to consider the objects’ non-material value in the context of Martial’s Saturnalian literary world, e.g. at 14.183-184, the Batrachomyomachia vs. Iliad and Odyssey. The same is true of 185/186, Culex vs. Aeneid, and many other pairs. One may also add that the factual value of an item is sometimes hard to determine anyway. I do not think that H. wants to suggest that the Apophoreta contain no lacunae at all, for how would we then explain the fact that the book (except for the proem) consists of an uneven number of items, 221 poems, unless we follow H.’s remark “daß madness herrscht, die method hat” (p. 46)?

Books XIII and XIV have thus far only very rarely been examined as literary compositions rather than simply dull catalogues.16 It’s a pity that H. proffers only one pretty straight-forward specimen (pp. 45-46), the talking cheeses from Trebula (13.33) that so openly interact with Virgil’s more-than-enough known epitaph: Trebula nos genuit, etc. But H. has detected yet another humorous element in the cheese-couplet. In the pentameter ( sive levi flamma sive domamur aqua), he argues, the cheese indicates that fire or water can reduce its rather ghastly stench, which is allegedly generated by the cheese’s = speaker’s bad breath. But there is no backing in the text for this daring assumption. Firstly, we have no idea what kind of cheese we are dealing with (let alone which Trebula is referred to); secondly, and more importantly, domare used of preparing food is well attested (see Leary ad loc. or ThLL V 1.1946.32-42), so it seems murky to regard the poor cheese as ozostomos. But I admit H.’s interpretation is perhaps more epigrammatic than Martial’s couplet.

I come finally to the book’s last part, dealing with the structure and arrangement of Books I-XII as a whole (“Aufbau der Sammlung”, pp. 123-152), which is indeed the culmination of the ‘new Martial‘. At the very outset of the book, H. had indicated that he would view Martial’s dozen epigram collections as a compositional unit. Consequently, the word “Dodekalog” to refer to the epigrammaton libri XII in their entirety appears regularly throughout the volume. Part of this assumed ‘inner logic’ is the idea that Martial intentionally composed 12 books, which the reader would recognize as an allusion to Virgil’s Aeneid (pp. 15, 123, 135); Book XII (in mock-heroic fashion, I suppose) constitutes the narration of the narrator’s (poet’s?) “nostos” (p. 15). Here, a quick reference in the bibliography (p. 152), i.e., to W. Allen’s et al. piece “Martial: Knight, Publisher, and Poet” ( CJ 65 [1970], 345-357 at 351ff.) would have been nice. It may indeed be challenging to consider Martial’s last book the work’s telos serving as the closure of Books I to XII, and the argument might have profited from some theoretic endorsement, e.g. from D. Fowler’s assessment of closure as completion and of intratextual markers of finality, etc.17

At any rate, it is mere speculation that Martial, right at the beginning of his production (!), ‘decided’ to publish 12 books in succession, or that he attached the prose preface of Book I to a revised edition of I-XII to provide a general poetological introduction into his oeuvre (pp. 123-124). But even if we let this hypothesis pass, the fact that Martial in his many self-reflexive ‘epigrams on epigram’ is much more concerned with the composition of entire books than of individual poems (e.g. 7.85) can obviously not function as an argument to support H.’s explication of Books I-XII as an intentionally ‘Virgilian’ or ‘epic’ dozen ( pace p. 124)

There is a shrewd overview of Martial’s “Poetik des Epigrammbuchs” (pp. 124-135), which is meant to further support the construct of the architectonics of I-XII in their entirety. But can it? I don’t think so. I agree that Martial, in the course of his regular book-productions, did aim at connecting each new collection to the existing ones (he numbers his collections [p. 131], and there are cross-references transcending the border of individual books). Also, the epigrammata ad libellum ipsum recur regularly from Book I to XII, but even if that can serve as a connecting rationale, the interplay between the first item of that series, 1.3, and the last, 12.2(3), does not necessarily advance H.’s point (pp. 130-131). Martial may indeed have had 1.3 in mind, when he wrote 12.2(3), but that’s about all we can say.

I greatly sympathize with H.’s rejection of P. White’s by now notorious ‘ libellus theory’ (pp. 128-130) in the context of a literary-critical exegesis of Martial’s books, because it works against the design of a book as a published entity and ties us badly to a non-litcrit approach.18 And yet, although I am far from being an adherent of White’s, the alleged incompatibility results from dogmatism on both sides. For it is justifiable and even rewarding to be interested in the investigation of the social background and, along with that, the communicative setting of the epigrams outside the published corpus. Indeed, both viewpoints can complement each other, as R. Nauta’s recent study Poetry for Patrons: Literary Communication in the Age of Domitian (Leiden 2002) shows.19 Fictional texts, their communicative act(s) and societal setups, do not exist completely independent of the extra-fictional reality.

But back to the “Dodekalog” theory. Books ι according to pp. 135-136, exhibit a symmetrical structure, which consists of four units of three books each, a quadruple system of triads, so to speak. As a structural signpost H. uses the ‘imperial poems’. In I-III as well as in χ the role of the emperor is rather marginal, whereas in IV-IX (two triads: IV-VI and VII-IX) the princeps plays an important role, while VII-IX, on account of that triplet’s particularly strong panegyrics, can be called “Kaisertriade” (p. 136). This triadic idea reminds me of the controversial analysis of Ovid’s Amores in H.’s Die römische Liebeselegie (2nd ed., Darmstadt 2001, pp. 110-139). If we consider the mere quantity of ‘imperial poems’ (the term being used as loosely as permissible to denote epigrams in which the emperor appears to have a significant, non-incidental function), then Books VII-IX indeed surpass the rest of the dozen; however, a real change is felt in VIII rather than VII.20

It is correct that in Books IV-VI Domitian is depicted as ‘being in Rome or close by’, whereas in VII-IX he appears predominantly as a successful military leader in the Sarmatan war (p. 138), but it is hard to see from a simple logical standpoint how Martial would have ‘anticipated’ such a change of perspective before the actual events that may or may not have motivated it. The same is obviously true of Book X and the entire fourth ‘triad’, where the notable decrease in ‘imperial epigrams’ is owed to the death of Domitian rather than vice versa. The upsurge in Domitianic poems from Book IV to Book IX mirrors the development of the ‘imperial activities’. Cf., e.g., the assessment by E. Merli, “Ordinamento degli epigrammi e strategie cortegiane negli esordi dei libri I-XII di Marziale”, Maia 45 (1993), 229-251 at 241-245 and (on VIII and IX) 249-251.

What connects Book VII to VIII is the idea of Domitian’s upcoming return from his Sarmatan campaign in AD 93 (cf. 7.1-2, 5-8), which the poet celebrates extensively in VIII, and the political circumstances prompted (if not necessitated) the prominence of such poems in the collection (cf. Coleman [n. 20], pp. 345-352). Taking this into account, it is worth considering what Martial actually says in the prose preface to VIII about imperial vs. non-imperial poems, for which see again Coleman at 339-340.

If you are a triad-fancier, you will perhaps follow H. in viewing Books VI and IX, i.e., the ‘finales’ of the respective Domitianic triptychs, as “Synthese[n] von Erotik und Panegyrik” (p. 138), since the central ‘imperial’ issues of VI and IX are the lex Iulia de adulteriis and the anti-castration edict respectively. One can of course find similar parallels between Books V and VIII ( ibid.), etc. This makes me feel quite uneasy. And yet, the chapter as a whole contains quite a few interesting and thought-provoking observations, e.g. on some connections between Books I and VII (pp. 139-140), but again: I am not sure if that alone proves that Martial intended VII to be regarded as the beginning of the dozen’s second half (p. 139) — by analogy with Aeneid VII, I assume.21

As for the original editions of Martial’s collections, a rather vexed issue is surely the second edition of Book X and its ‘disturbed chronology’ (Book X, as we have it, appeared sometime in AD 98, whereas Book XI came out in December 96), which disrupts the process of a sequential (chronologically unbroken) reception of Books I-XII. Since H.’s thesis as a whole depends on such a successive reading process, he has to deny the very existence of a second edition of X (pp. 147-148). No matter how coherent H.’s train of thought, it is very taxing to argue away what the poet himself so clearly states at 10.2.1-4: festinata prior, decimi mihi cura libelli / elapsum manibus nunc revocavit opus. / nota leges quaedam sed lima rasa recenti; / pars nova maior erit: lector, utrique fave (‘In composing my tenth little book, too hastily issued earlier, I have now recalled the work that then slipped my hands. Some of the pieces you will read are already known, but polished with a recent file, the greater part will be new.’ [tr. Shackleton Bailey]). According to H. (p. 147), Martial is not talking here about a second edition, and the nota quaedam does not refer to elements of the first edition but to elements that the reader is already familiar with from Books I-IX. That is not at all convincing. But H. (pp. 140-141) gives one important interpretive detail regarding Book X, namely that the opening poem with its witty reference to the book’s coronis contains a closural allusion, while the book’s last epigram (10.104), addressing the libellus, features elements that Martial usually employs in opening items. It sounds not unreasonable that 10.104 thus is intended to make the reader go on to Book XI (p. 142). In a forthcoming publication, in which I shall discuss the phenomenon of inversion in Martial in a broader context, I will return to this issue.

The liber Hispanus, as Martial himself calls his last collection (cf. 12 praef.), is equally troublesome because it includes praise poems on both Trajan and the already deceased Nerva. I shall not discuss the question whether or not XII contains epigrams from a special epitome of Books X (1st ed.) and XI, put together for Nerva after the publication of XI, since H.’s account is too compressed. However, since he follows closely (pp. 146-147) the very careful analysis by S. Lorenz, Erotik und Panegyrik: Martials epigrammatische Kaiser (Tübingen 2002), pp. 232-238, I suggest that one read Lorenz on the problematic transmission of XII and the deductions that can be drawn from it. At any rate, H.’s claim that the emperor addressed as Caesar in 12.4(5) (and cf. 12.11) is not Nerva but Trajan and that the ‘last triad’, χ was in its entirety published under Trajan (p. 151) is over-bold.22 One would have to utilize a lot more methodological care if one wanted to prove that Book XI was not issued in AD 96 but after Nerva’s death (pp. 148-150). For sure, Book XI is (more) Saturnalian (than any other collection in ι but it is more than illegitimate to ‘kill’ Nerva because he purportedly acts as the Saturnalicius princeps of the book (cf. 11.2) — or, I had better say, in order to save the triadic, Trajanic pattern. Back in 1869, Th. Mommsen expressed his outrage at Martial’s disrespectful attitude towards Nerva ( Hermes 3 [1869], 121) because he misunderstood the book’s Saturnalian license; some things obviously never change.

One final comment on yet another oddity: Saturn (Kronos), in whose honor the Saturnalia were celebrated, is called “Regent über die Gefilde der Seligen und insofern ein princeps in der Unterwelt” (p. 150). This sounds like the invention of a new myth. According to H., it might not be merely incidental that the Nerva-book is # XI, because Odysseus, too, on his way back home, went to see the underworld in Odyssey XI. Sic! In a similar vein: poem 64 is the by far longest item in Book VI, and, as Catullus is Martial’s favorite model, someone might sooner or later argue that the length and position of item 64 in that book is meant to allude to Catullus’ Peleus-and-Thetis poem. (I didn’t do so in my commentary.)

The fourfold triadic structure of the “Dodekalog” is, I fear, a modern construct for which, with different interpretive tools and starting-points, one could surely substitute a threefold tetradic system, etc.

I wish to stress that I do take it to be important if not imperative that one consider Martial’s Books I-XII as a unit that the poet himself, once it was finished, wanted to be read as such; and I am convinced that a reading of the ‘thing as a whole’ is structurally as justified as in Augustan poetry and will yield results that go well beyond the mere interpretation of individual poems or cycles.

But let me call a halt to my carping criticism by a quote from M. Heath’s Unity in Greek Poetics of 1989: “If one’s techniques of exegesis are sufficiently powerful, it will always be possible to reconcile any text with any set of aesthetic norms that one chooses” (p. 149). It should have become clear that the exegetical techniques employed in H.’s new ‘Martial’ are not “sufficiently powerful”; nor do I think that its “aesthetic norms” (nowhere defined anyway) are applicable, although the book contains many insightful and sensitive observations. Why? Mainly because H.’s centripetal approach to the “Dodekalog” led the author to a teleological construct of unity that is destructive rather than beneficial. My suspicion is that it is mostly this construct from which the rest of the book suffers (at times painfully).

The book contains useful indices nominum/rerum and locorum (pp. 164-168) and a general bibliography (pp. 153-163). It’s a big plus that each chapter at its end has individual references to items in the bibliography, and this offers some guidance, even though direct references to secondary reading suo loco are extremely scarce. Hence, the reader needs to be already quite solidly acquainted with the scholarship, if s/he wants to grasp the intellectual background and to differentiate between H.’s own position and someone else’s. It mirrors the development of research that some pieces that were part of H.’s 1988 bibliography are now absent from the new one (C.J. Classen’s, W. Hofmann’s, J. Kruuse’s, and H. Szelest’s articles).

Niklas Holzberg’s new Martial is much too rash in its many idiosyncrasies and simplifications of tricky problems, most of which cannot possibly be explained away so easily. He acknowledges his great debt to H.P. Obermayer and especially S. Lorenz, who both wrote their doctoral dissertations on Martial under his supervision. That collaboration prompted the book under review here (p. 10). I hope to have sufficiently demonstrated that the book, although it cannot serve as an introduction, contains many challenging ideas and hypotheses that may stimulate further work on Martial.


1. Yet another ‘system’ of H.’s, ‘Ovid’s elegiac system’, appears to be likewise disputable, as K. Galinsky (and others) emphasized (see, e.g., Gnomon 72 [2000], 214).

2. In fact, Seel did little more than summarize and reinforce the biased dogma of 19th and 20th century scholarship that Martial was a rather mediocre poet whose views of Roman society (in particular of Domitian) and whose ‘innumerable sexually perverted’ epigrams were seemingly the product of a morally depraved mind. Seel’s “Ansatz” got disseminated even among non-Classicists by being reprinted in G. Pfohl’s Das Epigramm: Zur Geschichte einer inschriftlichen und literarischen Gattung (Darmstadt 1969), 153-186. An abbreviated English version, meant to illustrate the then stereotyped status quo of scholarship, appeared in J.P. Sullivan’s anthology, Martial (New York-London 1993 = The Classical Heritage, vol. 3), 180-202.

3. Martial: the unexpected classic (Cambridge, etc. 1991). Also, as P. Howell rightly remarked ( CR 43 [1993], 277), the book ought “to be used with great caution” and, due to its length as well as its idiosyncrasies (among which esp. Sullivan’s psychoanalytical bent), cannot function as an introduction proper.

4. Originally put forward in Garthwaite’s PhD thesis ( Domitian and the court poets Martial and Statius, Cornell U., Ithaca, N.Y. 1978), parts of which, in revised form, appeared as articles in Prudentia 22 (1990), 13-22 (on Domitian’s enactment of the marriage/adultery law and Martial, Book VI) and Ramus 22 (1993), 78-102 (on the castration edict in Book IX).

5. See, e.g., I. Opelt, Gymnasium 96 (1989), 87, or P. Howell, CR 40 (1990), 36. H. Szelest ( Gnomon 61 [1989], 358-360) is sadly uncritical. Sullivan [n. 3] 128 with n. 17 is vague and unfortunately offers no discussion.

6. This tendency was (and is) of course partly (or entirely?) influenced by Augustan scholarship. Many publications could be cited. One of the leading figures was F.M. Ahl, e.g. in ANRW II 32.1 (1984), 40-124 and ibid. 32.5 (1986), 2803-2912, or AJPh 105 (1984), 174-208. Most of the relevant literature is critically examined by F. Römer, “Mode und Methode in der Deutung panegyrischer Dichtung der nachaugusteischen Zeit”, Hermes 122 (1994), 95-113.

7. In 1988, H. was startlingly hesitant to take Martial’s poem as a tool to interpret Catull. 2 and 3 (p. 31), and he misses this chance again in his new book, where he could have included 1.7 into the (new) chapter on ‘intertextuality’ (pp. 97ff.) to fruitfully discuss it there. See R.F. Thomas, Helios 20 (1993), 131-142 = Reading Virgil and His Texts (Ann Arbor 1999), 52-67 (esp. 55-65), and cf. Holzberg’s Catullus, pp. 61-67.

8. It was itself heavily dependent on J.P. Sullivan’s “Martial’s Sexual Attitudes”, Philologus 123 (1979), 288-302 (and cf. his Unexpected classic, pp. 185-210).

9. The fact that (according to K. Gutzwiller’s luminous Poetic Garlands of 1998) Hellenistic collections, like Martial’s, were most likely framed by programmatic and/or metapoetic items helps us only in terms of Martial’s proems and closures.

10. The 2001 edition of P. Mil. Vogl. viii 309 by Bastianini and Gallazzi may have appeared after completion of H.’s book.

11. Some references are given on p. 32. I am inclined to agree that D. Page’s results ( The Epigrams of Rufinus [Cambridge 1978], 23-49) cannot stand, and A. Cameron’s assessment ( The Greek Anthology from Meleager to Planudes [Oxford 1993], 65-69 and 78-84) is strong. Finally, add L. Robert, “La date de l’epigrammatiste Rufinus: philologie et réalité”, CRAI 1982, 50-63.

12. For sure, sexual animal metaphors are pretty productive in Latin (see J.N. Adams, The Latin Sexual Vocabulary [London 1982], 29-34), but many of them are ad-hoc formations that are easily comprehensible in their context. Consequently, Adams (at 34) notes only one ‘penis-hare’ image, Petr. 131.7 ( leporem excitavi). Funnily enough, in Martial we find a picture that would make a sexualized reading of 1.6.4 so absurd that it even transcends the limits of epigrammatic possibilities, namely 10.90.1/10, where the leo corresponds to the cunnus of a vetula. However, this did not prevent Sullivan [n. 3] 207 n. 35, from referring in the same footnote to the sexual ambiguity of both 1.6 and 10.90.

13. Garthwaite in F. Grewing (ed.), Toto notus in orbe (Stuttgart 1998), 157-172; Watson in ARCA 38 (1998), 359-372.

14. A point of departure would be the material as it is collected in quite a few late 19th or early 20th cent. dissertations and/or Schulprogrammen.

15. See, for instance, Apul. met. 6.29 ( iam credemus exemplo tuae veritatis et Frixum arieti supernatasse et Arionem delphinum gubernasse et Europam tauro supercubasse), which reminds us of both Philip’s item and Martial, spect. 5 and 19(16b).

16. It is good to have T.J. Leary’s commentaries of 1996 ( Apophoreta) and 2001 ( Xenia), but his primary interest lies not on the literary-critical side. Thus he even went so far as to doubt that one can assess Books XIII and XIV “primarily as literature” ([1996], 22), because their “subject matter [is] generally unpoetic” ([2001], 15); cf. Grewing, BMCR 2002.08.38, par. 8.

17. See his “First / Second Thoughts on Closure”, recently reprinted in D. Fowler, Roman Constructions: Readings in Postmodern Latin (Oxford 2000), 239-307, esp. at 242-246 and 284-285.

18. H. rightly refers to D. Fowler’s seminal piece “Martial and the Book” ( Ramus 24 [1995], 31-58).

19. See Nauta’s discussion of “Patronage in Martial’s Epigrams” (pp. 37-90) and especially “Modes of Reception of Martial’s Epigrams” (pp. 91-147), where he examines White’s theory and underscores the prime importance of the published book (108-118). A valuable assessment of Nauta’s book is available in BMCR 2002.11.22 (B. Gibson). S. Lorenz, in his review ( Plekos 5 [2003], 75-86), objects to Nauta’s methods (like H. to White’s), but is a little too unwilling to consider their potential (esp. at 77-79). Nauta’s examination of “You and I” in Martial (pp. 39-58), on which much of his argument depends, can easily be attacked by literary critics (e.g. Lorenz at 79, but also Gibson), and indeed, Nauta’s statement (at 48) that “[t]he idea that the speaker of a poem is always fictional is a modernist doctrine”, is as dogmatic as the so-called doctrine itself. Of course the speaker is fictional, but we can still fruitfully apply Nauta’s approach. H. has a lot to say about ‘fictional I’s and you’s’ (e.g., pp. 13-18, 74-75, 129, et saepius), but he does not utilize any coherent system; see only pp. 114-118 or 137.

20. Figures vary greatly, of course. Suffice it to refer to K.M. Coleman’s statistics, “Martial Book 8 and the Politics of AD 93”, Papers of the Leeds Intern. Latin Seminar 10 (1998), 337-357 at 339 with nn. 10-11. Mathematically (however inappropriate this may sound), Books V and VI do not differ much from VII. The percentages of ‘imperial items’ (following Coleman’s reckoning) are: 10.7% in V, 9.6% in VI, 12.1% in VII. VII and IX stand out (28% and 26% respectively).

21. I shall pass over the idea found quite regularly in scholarship since Friedländer (1886) that Martial issued a ‘collected edition’ of Books I-VII (not VI !), but, if that were the case (I don’t think it is), it would weaken the ‘two-halves’ construct.

22. The German rendering of 12.4(5).1-2 (“Die längere Mühe, die ich für das elfte und zehnte Büchlein aufbrachte, habe ich eingeengt”, p. 145) sounds odd. What does “eingeengt” mean in this context?