I ‘AN AS-YET ENTIRELY UNKNOWN WORK OF ANTIQUITY IN LATIN VERSE’?

Bucolic poetry is hard to date. Its conventions dictate a setting in a timeless world of shepherds — their loves and joys and quarrels and sorrows — and only rare glimpses of the contemporary world are as a rule permitted. Hence, controversies about dating have dogged much bucolic where the external evidence is thin. We can place some bucolic poets, such as Theocritus, Virgil, Nemesianus, Moduin, Petrarch, Boccaccio, and Dante with ease. But the dating of the non-Theocritean idylls, the Einsiedeln Eclogues, Calpurnius Siculus, and Theodolus has occasioned much debate, and the possible range for each author often extends over centuries. Did Calpurnius write in the first century or the third?Footnote ¹ In a recent article, I have argued that the Einsiedeln Eclogues, attributed for the past century to the age of Nero, were in fact written toward the end of the fourth century.Footnote ² Here, I will argue that there is another significant corpus of Latin bucolic poetry from Late Antiquity which has been ignored precisely because it has been misdated.

In 1946, Paul Lehmann brought this corpus to the world's attention for the first time. It contains a prologue and four eclogues, is extant in two manuscripts (one c. 1200, from the Biblioteca amploniana, the other its sixteenth-century copy in Erlangen), and is ascribed to the otherwise unknown Martius Valerius.Footnote ³ A specialist in pseudo-antike Literatur des Mittelalters, Lehmann saw the poems as twelfth-century literary productions in a classicizing mode. He did, however, raise and reject the possibility that they were genuinely ancient:

His fundamental argument — to which I will return in the conclusion — reduces to the unlikelihood of an ancient literary work escaping five centuries of antiquarian enthusiasm to come down to the twentieth century unpublished. In our day, that does not seem quite so strange. Three separate finds — by Johannes Divjak, François Dolbeau, and the Vienna trio of Schiller, Weber and Wiedmann — have uncovered sermons and letters of no less a figure than Augustine, and detailed investigation has uncovered the Gospel commentary of Fortunatian and the poetry of Pacatus Drepanius.Footnote ⁵ Indeed, about the time Lehmann was writing, Raymond Klibansky discovered in the Vatican library a new Latin philosophical text from antiquity, which may well have been written by Apuleius, and was edited for the first time last year.Footnote ⁶ The manuscript containing this text was in two even more famous libraries than that of Amplonius: that of Richard of Fournival in the thirteenth century and that of Queen Christina of Sweden in the seventeenth. Hence it is not prima facie impossible for an ancient text to have slipped past half a millennium's worth of eager humanists, be it housed even in a well-known library. Ultimately the strangely mannered prologue proved decisive for Lehmann — its lexical peculiarities convinced him that the poems which follow must be medieval. But when? Lehmann was constrained by the date of the earlier manuscript (Gotha, Forschungs- und Landesbibliothek mbr. II 125).Footnote ⁷ He also believed that Martius alluded to the Cosmographia of Bernardus Silvestris, written before 1147. Martius would then find a natural home amid the classical revival of the long twelfth century (conventionally, 1050–1215), which saw a number of neglected ancient genres flower again, elegy with Henry of Settimelo, satire with Sextus Amarcius, epic with Walter of Châtillon, epigram with Godfrey of Winchester, comedy with Vitalis of Blois, indeed even Petronian narrative with Elias of Thriplowe.Footnote ⁸ Classicizing bucolic would certainly find a welcoming home in such a milieu.

This idea was further confirmed in Franco Munari's exemplary editions of 1955 and 1970.Footnote ⁹ Munari went through the evidence thoroughly in his second edition, and identified a number of definitely post-classical features in vocabulary, syntax, and prosody. One particularly interesting aspect of his textual work, however, is the manner in which he handled the paradosis. In general, editors treat ancient texts and medieval texts very differently: the former descend through multiple lost generations to our earliest copies, while the latter often exist in manuscripts dating from just after the lifetime of the author. Hence, many editors of ancient texts (though by no means all) have frequent recourse to conjecture and restoration, whereas editors of medieval texts are often hesitant to correct the reconstructed archetype except in cases of simple or gross error. Munari edited Martius as if he were an ancient author (for which he was gently criticized by Herbert BlochFootnote ¹⁰ ), and subsequently some of the finest philologists of the twentieth century worked on the text's problems, including Otto Skutsch, Paul Maas, Sebastiano Timpanaro, Scevola Mariotti, and Giuseppe Giangrande.Footnote ¹¹ None of them seems to have doubted the medieval origin of the text. Skutsch at least recognized the problem:

Indeed, Skutsch's argument would come to be used to justify a more activist approach to medieval texts in general.Footnote ¹³ This critical interest mostly subsided by 1975. Emending Martius fell out of fashion, and the bucolics settled into their modest place as a representative, if uncharacteristic, text of the twelfth-century Renaissance, with only a few dissensions. One scholar assigned them to the late Carolingian period, and another courageously identified Martius Valerius as M. Valerius Messalla Corvinus, the great Augustan literary patron.Footnote ¹⁴

In 1987, François Dolbeau disturbed the placidity of this consensus, asking a simple question: ‘les “Bucoliques” de Marcus Valerius sont-elles une œuvre médiévale?’Footnote ¹⁵ This was no mere hypothetical question, but based on troubling new evidence. There is no specifically medieval content in the poems and all the post-classical features identified by Munari are attested by the fifth or sixth century, he argued. He also brought together for the first time our actual external evidence for the poet (a gloss in a Papias manuscript and a catalogue entry from Thorney near ElyFootnote ¹⁶ ), which seemed to put Martius’ activity in the reign of Justinian (‘Egloge aliquot Marci exquaestoris qui floruit tempore Justiniani’ reads the Thorney entry). Further, turning Skutsch's argument on its head, why is the text so corrupt if the Gotha manuscript was written within decades of its composition?

An answer came five years later from Christine Ratkowitsch, definitively negative.Footnote ¹⁷ She argued that Dolbeau had misunderstood the external evidence, and presented a detailed interpretation of the work in the context of Christian spirituality and pastoral care. She also suggested that extensive textual corruption can happen in a very short time, adducing the example of Joseph of Exeter.Footnote ¹⁸ That is a fair point, although the corruption in Joseph's nearly four thousand lines does not occur with anywhere near the density as that found in Martius’ five hundred. She also questioned Martius’ grasp on prosody, and in particular on the prosody of Greek names, although that question can scarcely be examined without considering the status of the transmitted text. Hence the central question deserves to be reopened: is Martius a medieval poet? Despite Ratkowitsch's best efforts, she still was not able to identify any explicit medieval or Christian elements, only material that was able to withstand a Christian interpretation. In the hands of the most adept commentators of the twelfth century, any pagan poetry was capable of Christian interpretation. Conversely, however, even the most resolutely classicizing work of the eleventh or twelfth century, such as the Satires of Sextus Amarcius, the Alexandreis of Walter of Châtillon, or the Cosmographia of Bernardus Silvestris, was hardly free of Christian or otherwise anachronistic elements. So the fact that the poems are capable of Christian exegesis does not make them medieval, and, if they are medieval, they are more successfully classicizing than virtually any other piece of medieval Latin literature.

But this does not mean we have reached a crux. Dolbeau's question is capable of a definitive answer. In this study, I will demonstrate that the poems of Martius Valerius cannot be medieval (as in from the seventh century or later), and were probably composed in the late fifth or early sixth century. The evidence rests on Martius’ relationship to the Greek world — in the setting of his bucolics, in his language, and in his sources — and strongly indicates that his works must be situated before the medieval divorce between the Greek East and the Latin West.

II TITYRUS A MOLLI SURREXIT IN ARDUA CLIVO

Martius is deeply indebted to the bucolics of Virgil. By way of an introduction to the collection, here I will show how his first three poems ever so closely mimic the first three of Virgil's eclogues, while his fourth imitates Virgil's sixth. This imitation is profound, extending from the highest level of the construction of the corpus to countless individual words. The first presents Ladon comforting and exhorting Cydnus, just as Virgil's first presents Tityrus and Meliboeus. Its geographic setting is not clearly defined, but both speakers are named after rivers, Ladon in Greece and Cydnus in Cilicia.Footnote ¹⁹ The imitation begins with the first line, where Ladon addresses Cydnus (buc. 1.1–4):

The close imitation extends to the very final word ‘et extensas ramorum traxerit umbras’ ~ ‘maioresque cadunt altis de montibus umbrae’. This is Virgil's famed nightfall, the ‘programmatic shadow’ that marks his distinctive approach to the genre. Calpurnius understood this, and ended his imitation of Ecl. 1 with nightfall (buc. 4.169): ‘iam sol contractas pedibus magis admovet umbras’.

The second bucolic is the lament of Iarbas for his indifferent lover Euphilis; Virgil's second is Corydon's for Alexis. Martius’ is clearly set in Africa. The singer's name comes from the king of the Gaetuli who was in love with Dido in Aen. 4. He makes references to Father Nile (Pater Nilus) and famous Canopus (traditionally, the boundary of Africa).Footnote ²⁰ The fourth line begins ‘Euphilin ardebat’, an unmistakable nod to the ‘ardebat Alexin’ of the first line of Ecl. 2.

Both laments then conclude with a self-address, ‘Cepit, Iarba, furor quantus te!’ and ‘a, Corydon, Corydon, quae te dementia cepit!’ (ecl. 2.74). Virgil's poem is modelled on an Idyll of Theocritus (11), and I will discuss its relationship to Martius Valerius further below.

The third bucolic is a singing contest between Mopsus and Moeris, with Lycurgus the judge; Virgil's third is the same, with Damoetas and Menalcas the competitors and Palaemon the judge. Martius sets it generally in Arcadia (3.77–8: ‘Me quoque Maenaliis praeponens Pan bonus antris / diligit’), the region of Greece which is one mythical homeland of pastoral. Virgil opens his eclogue with an odd bit of Latinity, ‘Dic mihi, Damoeta, cuium pecus? An Meliboei?’ The adjective cuius, -a, -um is certainly rare, possibly rustic, definitely archaic, probably comic. Virgil's critic Numitorius composed a biting parody: ‘Dic mihi, Damoeta, “cuium pecus” anne Latinum?’Footnote ²¹ In the same way, Martius’ Moeris begins, ‘Discute, Mopse, gregem: caper est erratus et hic est’. Est erratus is not obviously a medieval formulation; a deponent form of errari is only mentioned in the TLL as possible in Commodian's Instructiones, and in two glossaries, and is not found in the medieval Latin lexica.Footnote ²² The point is actually to imitate Virgil: any reader would immediately realize what est erratus means, grammatical or not, just as cuium pecus presented no bar to understanding. The exchange of insults followed by the challenge and the selection of the judge parallels Virgil ever so closely, as does the competition itself, amoebaean in form, with couplets (specific examples are discussed below).

Understanding the relationship of Martius’ third bucolic with Virgil's third helps clarify the preface of the collection. Here Martius is on his own, without a Virgilian precedent — his model is likely the satirist Persius who prefaced his collection of five hexameter poems with a brief metapoetic prologue in a different metre (choliambs). Martius’ prologue makes a direct allusion to Persius’.

Martius addresses his prologue to unspecified patres, and the general opinion, which interprets the whole collection as a spiritual allegory, has identified these patres as members of a religious community. Such an identification is possible, but unlikely. It was only in the late Middle Ages that patres became a standard term to refer to clerics; the preferred term would have been fratres, particularly if the author was also a member of a religious community. The vast majority of the instances of the term patres up to the thirteenth century refer in the biblical sense to ancestors, and particularly the Church Fathers. Instead, patres here means ‘senators’; and this is clear from the third bucolic. Where Virgil mentions Pollio in the third eclogue, Martius mentions Auxentius and Faustus. The widespread belief among grammarians in the Late Empire was that Virgil composed the eclogues at the instigation of, and in honour of, Pollio among others. Donatus’ life of Virgil notes that Virgil ‘moved on to bucolic, principally to celebrate Asinius Pollio, Alfenus, Varus, and Cornelius Gallus’.Footnote ²³ The Servian vita notes that ‘Pollio suggested to him to write bucolic poetry’.Footnote ²⁴ That of Philargyrius notes how ‘Virgil's farm was taken away, but at the command of Caesar Asinius Pollio restored it. In his honour, Virgil wrote the Bucolics’.Footnote ²⁵

Now Pollio is mentioned in ecl. 3.84–9, a passage where Virgil drops his bucolic mask for a moment and refers to a contemporary by name. It was this passage that likely inspired the grammarians’ belief that Pollio was the dedicatee.Footnote ²⁶ At precisely the same point, Martius makes reference to two individuals called by the recognizably late Roman names Faustus and Auxentius.

Faustus is a fairly common Roman name from the third to the sixth century. Auxentius is rarer — not even attested until the third century and modestly represented amongst the Roman élite of the fifth century; the most important people to hold the name were a notorious Arian bishop of the late fourth century, Auxentius of Milan, and two urban prefects of the middle of the fifth, Flavius Olbius Auxentius Draucus and Fonteius Litorius Auxentius.Footnote ²⁷ In other words, these names mean almost nothing to us, and would have meant even less to a twelfth-century audience. The only plausible solution to my mind is that Faustus and Auxentius are real people, the patrons or friends of the poet, just as Pollio was to Virgil.

Hence Martius’ prologue makes explicit what grammarians saw as implicit in Virgil. Auxentius and Faustus are to Martius what Pollio was to Virgil, and hence they are the patres (note the plural) of the dedication. Whoever they were, their names suggest senatorial rank. Some bearers of these names assumed even higher status. Fl. Anicius Probus Faustus the younger, for example, was consul in 490, quaestor sacri palatii from 503 to 505/6, and named patricius afterwards, before 507.Footnote ²⁸ Likewise, we do not have direct evidence for an Auxentius patricius, but Fl. Olbius Auxentius Draucus is called in a dedication v. c. et inl. patriciae familiae viro.Footnote ²⁹ The other prominent fifth-century bearer of the name, Fonteius Litorius Auxentius, likewise appears in a list of patricii and consules. Since he does not appear in the Fasti, it is reasonable to assign him the patriciate, but textual corruption is also possible and the list itself is of dubious origin.Footnote ³⁰

Lycurgus brings the competition to a halt with ‘Parcite iam, satis est; me iudice iurgia cessent’. In Virgil, Palaemon uses a metaphor of a canal: ‘claudite iam rivos, pueri; sat prata biberunt’ (111). The actual phrasing comes from elsewhere in Virgil (aen. 12.693): ‘parcite iam, Rutuli’.

The fourth eclogue breaks this pattern. It is set during the punishment of Apollo, when he was deprived of his divinity (‘humano … labore’) for his killing of the Cyclops, and was forced to herd the cattle of Admetus by the river Amphrysos in Thessaly.Footnote ³¹ Pan and a Napaean nymph, a native of the region,Footnote ³² are love-making nearby, but pause when they hear the music. Apollo then sings a song for Pan on the pyschogony, or the creation of the soul.

He then sings another song for the nymph, recounting a mythological catalogue, with Hero and Leander, the transformation of Cygnus, the loss of his son Phaeton, and Danae and the golden rain.Footnote ³³ He concludes with his own pursuit and loss of Daphne, his sadness at which overwhelms him into silence. In the sense that here Martius begins to ‘sing things a little greater’ (‘paulo maiora canamus’, ecl. 4.1), he is following Virgil's fourth eclogue. But in reality, his model is Virgil's sixth, where Silenus sings a song beginning with the cosmogony and continuing with a mythological catalogue. The parallels between the two poems are extensive: Silenus’ audience in Ecl. 6 is two boys, Chromis and Mnasylus, with the nymph Aegle; Martius’ in Buc. 4 is Pan and the nymph Napaea. Silenus proposes a different recompense for the boys and the nymph, a song for the boys, something obscene for the nymph; Apollo sings one song for Pan and another for Napaea. As Silenus begins, nature itself trembles in anticipation; as Apollo begins, everything becomes still and silent, animals, trees, streams, fields. At the end of Silenus’ song, the poet notes that Silenus sang everything which Apollo had sung on the banks of the Eurotas (ecl. 6.82–3); it is presumably this that inspired Martius’ own song of Apollo, albeit in the more pastoral environment of the Amphrysos, where Apollo is himself a human shepherd, rather than the Eurotas.

Martius’ imitation of Virgil compares well, I believe, with the way that Calpurnius and Nemesianus imitated him, and is not remotely similar to the way Metellus of Tegernsee or Alcuin did. Let it suffice to present the first exchange in Metellus’ first:Footnote ³⁴

Regardless of any subjective characterization (personally I regard Metellus as a master of his very peculiar craft), there can be no doubt that the difference between his bucolics and Martius’ is a distinction of kinds. They are not doing the same thing. Metellus’ use of Virgil is more centonic than imitative; he easily slips into Christian vocabulary; he strives for unrelenting internal rhyme and frequent end rhyme.

One observation suffices to sum up the difference: unlike medieval pastoral poets — but like his ancient forebears — Martius situates his poems more or less precisely in actual locations, a pastoral world firmly set in Greece — not in Theocritus’ Sicily — but stretching out to include Africa and Asia Minor as well. His imitation goes beyond the repetition of the tropes to the most minute level of detail, inviting his readers to precise observation and comparison.

III RARE SOURCES

The bucolics are extraordinarily allusive to a wide range of Latin poetry beyond Virgil, such as Calpurnius and Nemesianus (both the Eclogues and the Cynegetica), poets hardly known in the Middle Ages.Footnote ³⁵ Martius’ debts to Calpurnius are extensive; many of the former's evocative phrases are lifted directly from the latter. For example, buc. 1.14 ‘rapido sub sole’ and 1.92 ‘sol rapidus’ come from Calp. 1.10: ‘rapidoque soli’; 2.56 ‘Quem crucias? Quem, saeva, foves?’ from Calp. 3.61 ‘quem sequeris? quem, Phylli, fugis?’; 3.40 ‘timeam quam cerne’ from Calp. 6.49 ‘adspice quam timeam’; 3.70 ‘sociatas implicat umbras’ from Calp. 1.12: ‘ramis errantibus implicat umbras’; and 3.116 ‘trepidae multum crepuere cicadae’ from Calp. 5.56 ‘argutae nemus increpuere cicadae’.Footnote ³⁶

The fourth most alluded to poet is Ovid, including the Metamorphoses (e.g. 4.85–6, ‘fronde capillos / pulchraque per fragiles tendebat bracchia ramos’ ~ Met. 1.550 ‘in frondem crines, in ramos bracchia crescent’, both about Daphne), the Amores, the Tristia, the Epistulae ex Ponto, the Ars, the Remedia, the Ibis, and the Heroides, as well as the spurious Nux.

In addition to these, Martius has been alleged to know Seneca's tragedies, Horace's Odes, Satires, Ars and Epistles, Juvenal, Columella (Book 10), Lucan, Statius’ Thebaid, possibly Propertius, and Silius Italicus.Footnote ³⁷ He also has a deep acquaintance with the late antique poets Ausonius and Claudian.Footnote ³⁸ Besides poets, he is probably acquainted with Quintilian and certainly with Hyginus’ Fabulae. Some of these are more certain than others, such as Seneca, e.g. buc. 1.47: ‘Quid, quaeso, repetis curasque revolvis inanes?’ ~ Oed. 764: ‘Curas revolvit animus et repetit metus’ and buc. 4.55: ‘molli … carmine’ ~ Agam. 361: ‘carmine molli’; and Columella, e.g. buc. 4.1–5: ‘Egerat Amphrysi … Huc quoque cum tenera lascivus forte Napea / Pan aderat’ ~ 10.264–5 ‘Maenaliosque choros dryadum nymphasque Napaeas, / Quae colitis nemus Amphrysi’, the only two texts that put Napaeae by the Amphrysos, and buc. 4.36: ‘vivaces non terat herbas’ ~ 10.88 ‘vivacem cespitis herbam’.

At first glance, this list of sources — with all the due caveats proposed intertexts involve — might not seem unusual. Deeper analysis, however, raises questions. No complete manuscript of Calpurnius and Nemesianus survives which predates the fourteenth century (there is an incomplete manuscript of the first three eclogues of Calpurnius and part of the fourth from the twelfth century). Slight traces of both poets have been detected in the twelfth-century library catalogue of Prüfening and Calpurnius by himself possibly in that of Pfäffers from 1155.Footnote ³⁹ Besides that, knowledge of their works was restricted to the bucolic poet Moduin in the ninth century, the compiler of the Florilegium Gallicum and Guido de Grana, the annotator of the lexicographer Papias in Berne 276, from around the beginning of the thirteenth century. (Guido, it should be added, also offers one of the only testimonia to Martius’ work, as is discussed below.) Hence, it is significant that Martius knows their works so thoroughly: if he is medieval, his acquaintance puts him in a highly select group. A similar pattern emerges with other authors: Silius Italicus probably survived the Middle Ages in a single manuscript, and the number of medieval authors who knew him is vanishingly small. Propertius was known to the author of the mysterious De septem septenis in mid-twelfth-century England, and then the earliest copy is French, c. 1200; he was also known to Guido de Grana. Seneca's tragedies scarcely circulated before the thirteenth century, probably from a single manuscript in Northern Europe.Footnote ⁴⁰ Columella was scarcely known in the Middle Ages to anyone outside of St Gall. Nemesianus’ Cynegetica, which survived in two medieval manuscripts, one ninth- and one tenth-century, was apparently only known to Hincmar of Reims.Footnote ⁴¹ Two of the works of Ausonius Martius knows, the Cupido cruciatus and the Professores, each survived the Middle Ages in a single different manuscript.Footnote ⁴² Hyginus’ Fabulae survives today for the most part in an edition printed from a manuscript now reduced to fragments written in southern Italy around 900, and taken to Freising sometime later.Footnote ⁴³

If Martius Valerius were writing around the year 1200, the amount of access he must have enjoyed to rare texts is incredible. Put another way, a sixth-century Martius Valerius would be well-read but not exceptional; a twelfth-century Martius Valerius would vie with John of Salisbury, Guido de Grana or William of Malmesbury for the title of the most well-read scholar in medieval Europe.

IV THE RIDDLES

Martius’ use of Hyginus’ Fabulae sheds considerable light on one of the most widely discussed passages in the collection, the pair of riddles in the third bucolic (3.123–6). Just like in Virgil's third, the two contestants, Moeris and Mopsus, conclude their amoebean duel with an exchange of riddles.

Munari thought the riddles had been solved: the first referred to daisies, which in German are called Marienblumen, or Mary flowers, while the second (the solution credited to Scevio Mariotti) refers to the fact that some medieval ploughs left an M-shaped furrow — hence the Virgin Mary.Footnote ⁴⁵ This solution is beyond unconvincing. Daisies are not the first flower of spring; in so bookish a poet as Martius, we should find a bookish solution.Footnote ⁴⁶ Pliny the Elder is explicit that the violet is the first flower of spring, specifically the viola alba: ‘Florum prima ver nuntiat viola alba, tepidioribus vero locis etiam hieme emicat; post ea, quae ion appellatur et purpurea …’ (nat. 21.64). As for the second riddle, Mopsus is not talking about draught animals, but stock (pecus), and he specifically notes the mark left by their hooves, rather than by a plough or whatever they happen to be dragging. Finally, though ‘virgineum … nomen’ might look like a good Marian tag, I have found no evidence that it is so. In reality, it is Ovid's phrase; he puts it into the mouth of Perimele, lamenting the loss of her ‘virgineum nomen’ at the hands of Achelous.Footnote ⁴⁷ Hyginus provides the actual source, in wording so close to that of Martius that it can hardly be independent, in the story of Autolycus and Sisyphus, which is transmitted nowhere else. Autolycus, the son of Mercury, was so cunning a thief that he could never be caught, until he stole cattle from Sisyphus. Sisyphus responded by putting a mark on the hooves of his cattle (‘pecorum ungulis notam imposuit’), so that he could judge which cattle were his own. The idea is probably that a hoof print looks like a lunate sigma (C), for Sisyphus. Hence the answer to the second riddle ought to be a virgin whose name begins with C. One obvious candidate is Cynthia, or Diana, famed for her virginity. The two answers, then, are violae (ἰά in Greek, ia in Latin in the line from Pliny above) and Cynthia (Κυνθία); take them together and what results is iacinthus or hyacinthus, hyacinth. This is almost certainly the correct answer since it is related to the answer to the second of Virgil's riddles (ecl. 106–7):

The ‘flowers marked with the names of kings’ refer to hyacinths, which in antiquity were known to have marks which looked like AIA(I). This led to two aetiologies: that Hyacinthus was turned into the flower by Apollo, who killed him by accident, and the marks are an exclamation of grief, or that they refer to the name of Ajax in whose blood the flowers first sprang up (see, for example, Pliny, nat. 21.66).Footnote ⁴⁸ As Servius explains, these two are the kings whose names are inscribed on the flower.Footnote ⁴⁹ Another tradition takes it as referring only to Hyacinthus, with AIA as an injection of grief.Footnote ⁵⁰ The version mentioning Ajax, however, identifies the flowers as violae (the scholia Bernensia, for example, and the second recension of Philargyrius).Footnote ⁵¹ On the strength of the appearance of this story in Philargyrius and the Berne scholia (ostensibly derived, one must remember, from Philargyrius, Titus Gallus and Gaudentius), we can safely assume its currency in fifth- and sixth-century literary circles.

Further, we also have some evidence of ancient riddles which depend on this kind of word combination. In Petronius, a guest is promised a muraena, an eel, as a Saturnalia gift; what he actually receives is ‘murem cum rana alligata’, a ‘mouse with a frog tied to it’, since muraena is mus and rana attached together.Footnote ⁵² We can also find these kinds of riddles in the Greek Anthology (e.g. XIV.16):

The answer is Rhodes (‘Ρόδος), imaginatively made up of the cow's ῥο and the lender's cry of ‘hand it over!’ (δός).

In case there is any further doubt, Martius himself seems to confirm the answer in his next poem:

The ‘hyacinthine flowers’ with which the year begins anew cannot be anything but the first flowers of spring, the violet.Footnote ⁵³

V MARTIUS AND THEOCRITUS

The sophistication of the riddles in Buc. 3 should inspire a bit more curiosity about Martius’ sources and the audience for whom he was writing. In particular, it should make us wonder about his relationship to the Greek tradition. If he were writing in the sixth century, and had some familiarity with Greek, should he not have read Theocritus, like Virgil, like Calpurnius, like Nemesianus?

Theocritus was hardly forgotten among Latin authors and scholars even in the sixth century. In the second edition of his commentary on Aristotle's De interpretatione, Boethius describes Philo's notion of a modal proposition: ‘Philo says that <a proposition> is possible which is capable of truth due to the proper nature of what it says, as for example, when I say that today I will reread the Bucolics of Theocritus.’Footnote ⁵⁴ This offhand comment is quite remarkable, as if the bucolics were a text that a reader of a Latin commentary on a Latin translation of a Greek philosophical text might not just know about, but casually decide to reread at leisure. The Ordo generis Cassiodororum tells us that Boethius wrote a bucolic poem, ‘condidit et carmen bucolicum’; it may well have drawn on Boethius’ reading and rereading of Theocritus.

Nor had grammarians neglected him. Comparing Virgil's eclogues with Theocritus was something of a pastime in literary circles from the first century well into the fifth, and perhaps beyond. Gellius describes how at the dinner table one evening he and his companions read Virgil and Theocritus together, not just to find the obvious similarities, but to find particularly striking instances of clever or charming imitation.Footnote ⁵⁵

It was natural for these loci to make their way into commentaries. Servius’ comments on ecl. 2.21–3 make an excellent example:

Similar notes even occur in commentaries on other texts.Footnote ⁵⁶ Individual manuscripts of Virgil also bear witness to this readerly practice, where lines from Theocritus are noted marginally. A late fifth-century manuscript of Virgil with scholia, the so-called Scholia veronensia, attests to this practice. An example, on ecl. 3.27:

Some of the parallels in the Scholia veronensia are also found in the commentaries we have, but not all of them. Hence, they may reflect lost traditions or even the accumulation of interventions by individual readers.

This is one backdrop against which the literary practice of imitative series, or ‘window allusion’, should be evaluated. A poet can participate in this scholarly tradition, by imitating or referencing a passage of Virgil in which Virgil is drawing on Theocritus, while including elements which draw on Theocritus directly with no Virgilian analogue.Footnote ⁵⁸ In one sense, this kind of poetic nod is the same sort of thing that Gellius’ érudits and the scholiasts and commentators were doing: finding and comparing parallel passages. Instead of comparing directly (Virgil said this better; that is so much better in Greek), the poet embeds his commentary into his allusions. In another sense, however, he is inviting his readers to include him in their literary games, to read his poems against those of his predecessors.

This is how Martius Valerius uses Theocritus. As noted above, he nods to Virgil in the opening of his second bucolic with his ‘Euphilin ardebat’ (2.4). But before that, he speaks of the calamus as granting solacia to the silent flame, or of poetry as the remedy for love (2.1–2):

This is precisely the point Theocritus is making in Id. 11, Virgil's source, from the very first four lines:

The phrase tulerat solacia seems to echo id. 11.17, ἀλλὰ τὸ ϕάρμακον εὗρε, ‘but he discovered the remedy’.

Iarbas then embarks on a series of adynata, or impossibilities to express a lover's incomprehension at his rejection (ll. 30–41).

Martius’ allusions range across the whole tradition. The first adynaton comes from Claudian's De raptu (2. pr. 25): ‘securum blandi leporem fovere molossi’. The second is his own, while the third and fourth imitate Virgil (ecl. 8.52–3): ‘aurea durae / mala ferant quercus, narcisso floreat alnus’. In the second set, the first two are his own, while the third neatly combines Virgil (ecl. 8.52): ‘nunc et ovis ultro fugiat lupus’, with Claudian (rapt. 2.pr. 26): ‘vicinumque lupo praebuit agna latus’. The final adynaton is adapted from Horace (carm. 1.2.9–12): ‘piscium et summa genus haesit ulmo /…/ et superiecto pavidae natarunt / aequore dammae’. What none of these parallel passages contain, however, is a general description of adynata. In Theocritus Id. 1 (1.132–6), the passage Virgil is imitating here, we find intermingled in the adynata a hemistich describing their general nature:

‘All things are turned upside-down.’ The same hemistich is found in Martius Valerius, in both series of adynata: ‘poterunt tunc omnia verti’ and ‘vertentur cuncta, necesse est’.

A little later in the lament, Iarbas wonders whether his Euphilis is too proud to love a shepherd. But in an apostrophe to his absent love, he adduces the counter-example of Phoebe, goddess of the Moon, who loved a shepherd (Endymion, though his name is not mentioned).

The same mython in the same context is adduced in Theocritus(?), Id. 20.37–9, where a herdsman laments his rejection by the town-dwelling Eunica:

These two passages are far closer to one another than to any of the other ancient tellings of this story in context (that is, as an argument that not even goddesses spurn the love of shepherds), in content, recounting her descent (ἀπ’ Οὐλύμπω ~ linquens … polos, μολοῖσα ~ venit, Λάτμιον ~ ad pastoralia lustra) and their concubinage (κάθευδε ~ iacuit), in emphasis (for example, βουκολέοντα ~ pastoralia, agresti), and in omission, that is, neither mention what is usually the most important part of the story, Endymion's enchanted sleep. As if to dispel any lingering doubt, the first two words of the following lines in each are identical, καὶ τύ ~ tu quoque. This is similar to how Virgil will translate a couple of insignificant words of Theocritus to signal a looser adaptation, a winking quotation mark.Footnote ⁵⁹

Finally, at the end of the lament, Iarbas addresses himself directly, just as Corydon does in Virgil, Ecl. 2 and Polyphemus does in Id. 11:

Theocritus’ Polyphemus suggests that he ought to cut down branches for his lambs and milk them. Virgil's Corydon plans on finishing his vine-pruning and weaving something useful out of reeds, not specifying what. A basket for making cheese is probably intended (cf. Nem., ecl. 2.33–4), as in Theocritus. Valerius is clearly imitating Virgil, at points quite closely (cepit ~ cepit, furor ~ dementia, te ~ te, quin potius ~ quin … potius, viridi … iunco ~ molli … iunco). But he includes three elements found only in Theocritus — the needy lambs, the identification of the basket for cheese making, and the milking. The evidence from Buc. 2 is overwhelming: Martius Valerius must have had direct access in some form to Theocritus.

The third bucolic confirms this impression. It is a dialogue between the shepherds Moeris and Mopsus, with Lycurgus as a judge, ever so closely following Virgil's third eclogue. Both Valerius and Virgil use priamels, or linked series of comparisons (3.99–107):

Virgil has in mind a passage in a similar singing contest in Theocritus(?), id. 8.57–8, where Daphnis uses a similar priamel:

One element, or syntagm, is identical, δένδρεσι μὲν χειμὼν ϕοβερὸν κακόν ~ [triste] arboribus venti, as a deliberate nod toward his source. Martius is imitating that same couplet of Virgil, keeping his entire last hemistich almost intact. Yet at the same time, he directly incorporates a different syntagm from the very same passage, ὄρνισιν δ’ ὕσπλαγξ ~ ut viscum contristat aves. The following lines include another syntagm from Theocritus, from id. 9.34–5, γλυκερώτερον, οὔτε μελίσσαις / ἄνθεα.~ gaudet apes calathis (calathus here being the calix of a flower).

At the end of the competition, Lycurgus calls the conclusion to a close: ‘Parcite, iam satis est; me iudice iurgia cessent.’ The satis comes from Virgil, where Palaemon ends with ‘sat prata biberunt’ (3.111). But Palaemon does not seem to interrupt them, whereas Lycurgus cuts the competition short. In the same way, in Id. 5, Morson intervenes to stop the contest: παύσασθαι κέλομαι τὸν ποιμένα (138). In both Virgil and Martius, the competition is inconclusive, whereas in Theocritus, Comatas wins. The reason for the inconclusive result in Martius, however, is explained quite differently than in Virgil, ‘vincere uterque potest, non cedere uterque iuvatur’. This conclusion instead comes from a different singing contest in Theocritus, id. 6.46: νίκη μὲν οὐδάλλος, ἀνήσσατοι δ’ ἐγένοντο (‘neither was victorious; each was undefeated’).

Some of these parallels are more muscular than others, but the two or three which are absolutely certain are sufficient to secure the others and demonstrate beyond question or doubt, that Martius Valerius knew Theocritus. And if he knew Theocritus, he is not a French or German poet of the second half of the twelfth century. To argue that would overturn almost everything we know about Greek studies in the West in the Middle Ages.Footnote ⁶⁰

VI QUI FLORUIT TEMPORE JUSTINIANI

The evidence I have brought together has given us warrant to answer Dolbeau's original question in the negative; the bucolics of Martius Valerius are not a production of the Middle Ages. What we have found, however, is consistent with the external evidence Dolbeau brought to bear. We should now examine these two testimonia in detail. Both come from renowned bibliophiles and book hunters, John Leland in the fifteenth century, describing a manuscript he saw at Thorney, and Guido de Grana in the thirteenth, illustrating the word mens in Papias’ Elementarium:

First, Leland. Ratkowitsch's attempt to dismiss the entry is unconvincing. She argues that the entry refers to some otherwise unknown legal text by some otherwise unknown Marcus who took part in Justinian's legal reforms.Footnote ⁶¹ Interpreting a testimonium which plausibly refers to a known text and author such that it refers to a person for whom we have no other evidence is in itself dubious. But her argument is problematic in another way: she bases it on the fact that ecloga does not necessarily refer to an individual bucolic poem until much later. A simple survey of the evidence, however, contradicts this assertion. The vast majority of instances of the term ecloga in Latin texts from both antiquity and the Middle Ages refer to poetic works, and the majority from Late Antiquity on refer to Virgil's bucolics. But in the end, that does not even matter, since the entry was composed in the sixteenth century when ecloga only had the meaning of bucolic poem. The wording of the entry confirms that it was composed by Leland himself and not simply copied from the Thorney manuscript he saw. Leland is often concerned with dating authors: so, for example, for Llanthony (Collectanea III, p. 159), he lists: ‘Clemens prior tertius inter Lantonenses super evangelia. Hinc coniectura est floruisse illum tempore Richardi primi vel Joannis.’ Or, for Rievaulx, he includes a work by Walter Daniel ‘qui floruit circa tempore Stephani & Henrici 2^i’ (Collectanea III, p. 38). Numerous other examples of such formulations can be found throughout the Collectanea. When Leland did not want to actually count the number of discrete works in a collection, he would simply add aliquot; compare the listing for Cirencester, where he lists the commentary of Gervase of Chichester's commentary on Malachi, and then follows it with ‘eiusdem aliquot omeliae’ (Coll. III, p. 159). Finally, for Leland, eglogae are individual poems in a bucolic collection; witness an entry from Balliol College, ‘Bucolica Boccatia continentia 16. Eglogas’ (Coll. III, p. 66). Nearly every word of the Thorney entry betrays Leland's hand, and must indicate an unspecified number of bucolic poems by somebody named Marcus (or Martius) who was an exquaestor, and who was alive in the time of Justinian.

This curiously exact information must have been deduced from paratextual materials accompanying the poems, a long inscription at least, or perhaps a short vita. Sometimes, as in the examples above, Leland would use chronology and registers to establish a floruit for an author. Obviously that would not be possible for a Justinianic author. In other cases, Leland would deduce a notice from a text's prologue. For example, for Malmesbury, Leland lists, ‘Sententiae Xysti interprete Rufino qui contendit hunc fuisse Xystum pontificem Romanum’ (Coll. III, p. 157). These are the Sentences of Sextus translated by Rufinus, a neo-Pythagorean ethical collection in a Christian frame. In his prologue, Rufinus notes that ‘some people relate that Sextus himself is called Xystus, among you, that is in the city of Rome, honoured with the glory of bishop and martyr’.Footnote ⁶² Jerome showers scorn on Rufinus for this identification (cf. ep. 133.3), which at any rate did not originate with him, as it is found in the Armenian version of the text as well. Leland, looking for information to identify this rare text (the editio princeps, Lyon 1507, came out in Leland's own lifetime), strengthens this note to suggest that Rufinus definitively identifies the author of the text with Pope Xystus.

Now for Guido: despite the grammatical irregularity, he undoubtedly is suggesting that Martius Valerius was a consul. Again Ratkowitsch attempts to dismiss this evidence by pointing to the hazy status of the title consul in the Middle Ages, suggesting that it could be applied to all sorts of people.Footnote ⁶³ While that fact is indeed true, the reason why consul is such an ambiguous term is because it was used analogically to describe a person of some importance, whose rôle was parallel to what they imagined a Roman consul's rôle would have been. In other words, the Roman sense of consul remained the primary meaning of the word throughout the Middle Ages. I am aware of a few occurrences of literary figures mentioned with consul attached to their name (I list them below) — in all cases it is because they actually held that office. The other problem with Ratkowitsch's argument is that her explanation requires independent convergent error: as she admits, the two testimonia fit perfectly together. Dismissing them separately entails an extremely heavy burden of proof since they are mutually reinforcing.

The only place he could have gotten such information is from the title in a manuscript. Sometimes manuscripts of late antique authors give titles indicating the office they held. For example, one manuscript of Ausonius, Voss. Lat. F. 111 (c. 800), f. 20v, calls him ‘Ausonius consul’ in the title to the Ludus septem sapientum, while another manuscript, Vat. lat. 1611 (fifteenth century), calls him ‘Ausonii poetae viri consularis’. Likewise, the catalogue of Saint-Oyan (eleventh century) lists a codex containing ‘libri carminum Ausonii consulis’.Footnote ⁶⁴ Even better is the title of the lost manuscript of Helpidius Rusticus’ Historiarum testamenti veteris et novi tristicha which had been in the hands of Johannes Hartung, who passed it on to Georgius Fabricius to be printed:Footnote ⁶⁵

The fact that this title contains an absurd error — it should be v(ir) c(larissimus) et inlustris, as Seeck realizedFootnote ⁶⁶ — guarantees its authenticity, despite the fact that Hartung's manuscript has never been found.

Dozens of examples of even longer inscriptions can be found in Boethius manuscripts. Two representative specimens:

Of course, many authors, Boethius included, were honoured with a vita preceding their works. But there is also an intermediate stage between vita and inscription, represented by the anecdoton Holderi, or the Ordo generis Cassiodororum, a very short piece transmitted with the Institutiones of Cassiodorus which tells us about his career, as well as those of Boethius and Symmachus. It begins with what seems to be a long inscription from the text from which it was derived (Karlsruhe, Aug, 106, f. 53v):

A longer paratext like the anecdoton may well have included something about Justinian which would have led to both Leland's deduction and Guido's title. Since the Thorney entry bears all the marks of Leland's own composition, all we can safely conclude is that the manuscript contained some mention of Justinian, be it consular dating of the year 521, Justinian's first consulate, in a subscription (which often included consular dating) or a mention of the emperor under whom he obtained preferment. From Guido, we can surmise that it also included some other title as well, consul, exconsul, or something else which could have given rise to it.

Unfortunately, Martius Valerius cannot be securely identified with any known fifth- or sixth-century figure. The first potential piece of evidence comes from Sidonius, who mentions an otherwise unknown poet Martius (carm. 9.306): ‘et nulli modo Martium secundum.’Footnote ⁶⁷ But that poem was probably written no later than 470, which makes it difficult to identify Sidonius’ Martius with someone who flourished in the time of Justinian.

One might hope that the titles Martius received would assist us, and indeed in the year 521 there is an ordinary Western consul named Valerius, and his colleague in office is none other than Justinian. But that Valerius’ full name might be Iobius Philippus Ymelcho Valerius, although the identification is not entirely certain.Footnote ⁶⁸ It is certainly not impossible that this Valerius had been a quaestor sacri palatii sometime earlier. Prosopography gives us all sorts of combinations of quaestorships and consulates. (It bears reminding that both these offices could be held in either East or West.) Cassiodorus was first quaestor sacri palatii and then ordinary consul,Footnote ⁶⁹ while one Proculus was first quaestor sacri palatii and then an honorary consul.Footnote ⁷⁰ But it can go the other way: Anicius Probus Faustus was ordinary consul in 490 and then made quaestor in 503.Footnote ⁷¹ But one should be cautious: Priscian's Institutiones is addressed to a certain ‘Iulianus consul ac patricius’. This Julian, despite his eminence, is entirely unknown.Footnote ⁷² He was evidently not an ordinary consul — not listed in the Fasti — but an honorary one. In point of fact, the Prosopography of the Later Roman Empire, as useful as it is, gives a slightly deceptive illusion of solidity and authority. Had the editors known about the Leland note, they would probably have included the exquaestor Marcus, and then of course, we would have an entry in which to slot our poet. The fact is that many individuals in the early sixth century are attested by a single piece of evidence; if anything, our Martius is unusual in that he is attested with two. If the consul of 521 is not Martius Valerius — he probably was not — then his putative consulship was almost certainly honorary, since our consular lists tend to be reliable. Further, if Martius was quaestor sacri palatii, he was much more likely to have been in office under Anastasius (491–518) or perhaps even earlier, when our information is much more fragmentary than for the reigns of Justin and Justinian.Footnote ⁷³

But there is another solution to this problem: the title consuli, in its bizarre ungrammatical form, may actually be due to confusion on Guido's part. Medieval and Renaissance scribes and scholars were generally baffled by late antique nomenclature and titulature — witness the exinlustris of Hartung's Helpidius. They found obscure even the most common honorific of all VC, for vir clarissimus, and often expanded it to vir consularis.Footnote ⁷⁴ One early medieval scribe, copying some excerpts of Macrobius’ Saturnalia known as the Disputatio Chori et Praetextati in Padua, Antonin. 27, f. 66r, even expanded the abbreviation in the author's name to Vncies [i.e. quinquies] consulis, or ‘five-time consul’.Footnote ⁷⁵ So the Marcus Valerius consuli[!] may well have been meant by Guido as consularis and not consul. His source would have read Marti Valeri VC, or perhaps, like the Helpidius inscription above, Marti Valeri VC exquaestoris. If our Martius was not actually a consul, honorary or otherwise, it would be much less surprising that we do not find him attested elsewhere. Regardless of whether one accepts this reconstruction, the point remains that Guido's title points unmistakably toward late Roman titulature, and would make no sense for a twelfth-century poet.

We then have one possibility remaining. Orlandi edited for the first time a late fifth- or early sixth-century inscription from the Colosseum, of which all that could be read was […] et Martius […].Footnote ⁷⁶ She points out that the only person named Martius we have among the Roman élite of the period is Sidonius’ poet. An exquaestor is the sort of person who we might expect to have a seat at the amphitheatre, and the relative rarity of the name makes the identification attractive. An intertext we have not yet discussed may well bind our poems very closely to this milieu. In celebration of his first consulship in 521, the future emperor Justinian distributed diptychs inscribed with a couplet:Footnote ⁷⁷

Compare the opening of Martius’ prologue (buc. prol. 1–2):

Neither the words nor the sentiment are particularly uncommon, but the conjunction of so many lexical points is unlikely to be coincidence. We do know that these diptychs were widely distributed (we still have three surviving), so Martius may well have been nodding to some lines well known in senatorial circles. But it is possible that Justinian (or whoever wrote the couplet) was reworking a bit of popular poetry. The fact that Justinian's colleague was a Valerius is interesting, although hardly indicative of anything in particular. At the very least, the language of the dedication in Martius is redolent of the language of senatorial dedications in the sixth century.

If the political context of Martius is unable to be reconstructed with precision, the literary milieu is slightly clearer. Boethius’ carmen bucolicum has already been mentioned; as I noted above, it very likely drew on the works of Theocritus. We cannot tell whether it was a single bucolic poem, like Endelechius’ De mortibus boum, which is called simply a carmen in the manuscript, or a bucolic series divided into individual eclogae, as the Servian vita refers to Virgil's Eclogues, and as Petrarch and Boccaccio would later compose.Footnote ⁷⁸ But we have another ‘Christian Virgil’ of the sixth century: Severus Episcopus, to whom the Lorsch catalogue attributes twelve books of verse on the Gospels, ten eclogues and four books of georgics (Metrum Severi episcopi in evangelia, libri XII, eiusdem eglogas X, eiusdem Georgicon, libri IIII).Footnote ⁷⁹ The twelve books of epic, ten eclogues, and four books of georgics proved too much for earlier scholars such as Becker and Manitius to handle: they assumed that there must have been some confusion with Virgil, since after all the works of Severus did not survive. But then in 1967 the astounding Bischoff managed to conjure up some unnumbered parchment sheets from Trier with several hundred lines of hexameter verses on the Gospels, including an explicit for book nine of Severus Episcopus, and an incipit for book ten.Footnote ⁸⁰ Since this Severus did indeed, it seems, write a twelve-book epic, it seems eminently plausible that he had completed an entire rota Virgiliana. Martius offers some unexpected support for this, since even though we only have four of his eclogues, each of them follows a specific Virgilian model so closely that a collection imitating all ten is easy to imagine. (Such a collection in fact exists in the much later and very different Quirinalia of Metellus of Tegernsee.) Further, Martius hints in his prologue about continuing his poetic career unto loftier genres (prol. 1–10):

Whether or not Martius himself ever mounted a loftier hill, and composed poetry in other genres, it is certainly possible that a contemporary of his like Severus well may have. Hence, the total bucolic production of the sixth century which we can be fairly certain existed amounts to at least fifteen poems (possibly a deal more) by three different authors, as rich an output as any century saw from the time of Virgil on.

VII SPECIFIC PROBLEMS

The foregoing has been sufficient, I hope, to demonstrate that Martius Valerius was writing no later than the sixth century, and we have no reason whatsoever to doubt the external testimonia. Some scholars, nonetheless, have identified particular features which, they argue, prove that the poems must have a medieval origin.

At the end of the first bucolic, Ladon urges lovelorn Cydnus to accompany him home:

No one is quite sure what these lines mean, that is whether they describe noon or late afternoon, and textual corruption is likely. Nonetheless, there is no justification for assuming that apricum conceals a French abri. Apricus meaning ‘cover’ is already attested in Late Antiquity, and that is not even the only possible interpretation of these lines.Footnote ⁸²

Zicari has argued that an anecdote in 3.10–11, in which Moeris notes that Mopsus has suffered the punishment due a thief, demonstrates that the poems must be medieval:

The analogue of this passage is in Calpurnius, ecl. 3.73–4:

But the punishment of Martius’ Mopsus goes beyond that of Calpurnius’. He is not only tied up, but shaved and branded as well. Zicari notes that there seems to be some precedent for judicial shaving in the Middle Ages, and concludes that this is a definitely medieval element.Footnote ⁸³ The examples he brings forth (from Du Cange) are not convincing: they relate primarily to the punishment of slaves, which clearly does not fit the context here.

In fact, this punishment does not correspond precisely to known ancient or medieval punishments for theft. It is, however, the punishment decreed in the Lombard kingdom by King Liutprand in 726 (no. 80) for recidivist thieves:

Where this part of Liutprand's law came from is unknown, but Julia Hillner has argued that the language of rehabilitation embedded in the longer passage harks to a late Roman context.Footnote ⁸⁵ This anecdote makes sense in a post-classical, pre-Carolingian context, but very little either before or after. It cannot be considered a specifically medieval feature.

Another passage which has been seen as definitively medieval is the prologue, where Martius presents four elegiac couplets all consisting of lines of two or three words. Janet Martin has argued that this kind of mannerism is characteristically medieval, in contrast to the classicism of the poems that follow.Footnote ⁸⁶

There is nothing like this extended passage in Latin literature, ancient or medieval. There are occasional isolated hexameter lines of three words, e.g. Lucretius 3.907: ‘insatiabiliter deflevimus aeternumque’, Horace, sat. 1.2.1: ‘ambubaiarum collegia pharmacopolae’, and Ovid, fast. 2.43: ‘Amphiareiades Naupactoo Acheloo’. The device is also found in Late Antiquity (e.g. Claudian, de iv. cons. Hon. 8.560), and was a particular favourite of Sidonius (e.g. 2.104 and 2.507). Sidonius also indulges in polysyllabism in prose, such as when he closes ep. 1.1 with ‘volumina numerosiora percopiosis scaturrientia sermocinationibus multiplicabuntur’. Rutilius Namatianus offers a striking example of a two-word pentameter line in an elegiac couplet (1.450): ‘Bellerophonteis sollicitudinibus’. In the Middle Ages, such lines are occasionally found as curiosities (later sometimes called versus macrocoli Footnote ⁸⁷ ); Peter of Isosella in the mid-thirteenth century mentions verses of this sort in his popular Compendium grammatice:

A similar couplet is found in the proverbs of Serlo of Wilton Paris lat. 6765, f. 54v (thirteenth century):

Serlo composed his proverbs freely pillaging other texts. The line that follows in the Paris manuscript, for example, ‘saepius obliti quidam dixere tacenda’, is taken from Vital of Blois's Aulularia 441. Serlo's lines are too close to Martius, pr. 15–16 to be independent; contra Munari, however, the presumption should certainly be that they are extracted from Martius (with ‘improvements’) and not vice versa, considering they represent a remnant of a longer passage of such verses. After all, Martius’ couplet occurs in a longer passage in which it makes sense; Serlo's stand-alone version is literally senseless: who are they, what did they say, and what did they neglect to mention? It is almost unthinkable that Martius found this odd couplet in a florilegium and built a coherent passage around it.

We also get a fair number of such verses in the De trinitate of Bernard of Cluny (e.g. 616: ‘Inconprehensibilis inconmutabilitatis’), but that could as well result from the polysyllabic tendencies of Trinitarian theology (in prose compare Aelred, serm. 127, p. 266 Raciti, on the Holy Ghost: ‘incircumscriptus incomprehensibilis’ and Bonaventure, itinerarium 2.9: ‘lux veritatis in qua cuncta relucent infallibiliter indelebiliter indubitanter irrefragabiliter indiiudicabiliter incommutabiliter incoarctabiliter interminabiliter indivisibiliter et intellectualiter’).

In Greek, three-word hexameter lines are attested early, in both Hesiod (op. 383: πληιάδων Ἀτλαγενέων ἐπιτελλομενάων) and Homer (il. 11.427: αὐτοκασίγνητον εὐηϕενέος Σώκοιο). The feature was pointed out in the old scholia to the line in Homer (p. 206 Erbse): ἐκ τριῶν μερῶν τοῦ λόγου ὅλος ὁ στίχος.Footnote ⁹⁰ If anything, Latin scholiasts were less interested in the phenomenon; the ancient scholia on Horace, for example, make no comment about the feature in sat. 1.2.1. In fact, the Greek tradition supplies the only real analogue to this passage, an old Greek epigram about the Sophists quoted by Hegesander of Delphi, according to Athenaeus (deipn. 4.53 162ab)

Hence, the polysyllabic exuberance of the preface does not answer the question of how to date the poems, nor does the intertext with Serlo provide a terminus post quem but rather a terminus ante quem. They could be ancient or medieval; they could be relying solely on native Latin traditions, or they could be imitating Greek mannerism. It is quite possible that the real inspiration was Persius. As I argue above, Persius’ choliambs are one of the inspirations of Martius’ elegiac prologue, and one line of the prologue does indeed consist of only three words (4): ‘Heliconidasque pallidamque Pirenen’.

Finally, there has been a deal of discussion regarding a vexed passage in the first bucolic which seems to refer to a real place and which has been regarded (on no particular grounds) as giving some information about Martius’ identity (buc. 1.50–4)

No consensus has been reached about the meaning of these lines: some have suggested Lucerne in Switzerland, Laurana in Istria (Lovran in Croatian), Loreto in the Marche, the abbey of Lorsch (Lauriacum), and Laurensberg, a town outside of Aachen.Footnote ⁹¹ It was even suggested that the referent was Rome itself.Footnote ⁹² Everyone has been looking in the wrong part of the world. Cydnus is the speaker of these lines, which suggests in finding his home we might want to begin with Asia Minor. What he tells us in particular is that it is near to a city, and that from it (that is from Lauri altae) he can see a forum and a citadel. The obvious answer is Daphne, on a hill, near a large city, Antioch, with a prominent forum and an arx, and which happens to be called Laurel, or Daphne.

Martius intended the reader to see the Greek name behind the Latin rendering. These lines are addressed to a shepherd named Ladon, named after the river who in some accounts was the nymph Daphne's father.Footnote ⁹³ Hence there is particular force that the line ‘Scis, reor, hunc collem, Lauros ubi dicimus altas’, ‘You know, I believe, this hill which we call lofty Daphne’, is addressed to Ladon. Ladon, of all people, would be the one familiar with it.

There are other possible reasons for transferring Arcadia to Daphne. First, the name: the programmatic allusion in pastoral is to the Daphnis who invented bucolic song. One explanation of his name connects it with laurel, a confusion encouraged by ecl. 8.83: ‘Daphnis me malus urit, ego hanc in Daphnide laurum’. Servius notes that Virgil perhaps chose laurel ‘propter nominis similitudinem’ (ad loc. 106 Thilo/Hagen). In Philargyrius, the confusion between Daphnis and Daphne, the nymph beloved of Apollo who was turned into a laurel, is almost complete.Footnote ⁹⁴ Martius implies this same identification by referencing the story of Apollo and Daphne in the final poem (buc. 4.83–90), neatly closing the circuit.

Second, the region. Virgil introduced the character Mopsus into bucolic, for reasons that no one has quite been able to grasp.Footnote ⁹⁵ He is also a star of Martius’ collection, as one of the singers of the third bucolic. Before Virgil, Mopsus was the name of two different seers, one of whom was regarded as the founder of a number of cities in Cilicia, foremost Mopsuestia (Pliny the Elder calls the city simply Mopsos, nat. 5.22.91), just ten miles from modern Adana. By the sixth century, we can see the simple identification of Virgil's character with the seer.Footnote ⁹⁶ If Mopsus the shepherd and Mopsus the seer are the same, there is no reason why bucolic should not be set in Cilicia and bordering Syria. The possible connection between the family of Auxentius and this region also makes the possibility attractive.

Third, the place itself. It was well known as a natural paradise; Ammianus speaks of it as ‘amoenum illud et ambitiosum Antiochiae suburbanum’ (19.12.19), and the author of the Historia Augusta speaks of its delicia (Marcus Aurelius 8.12). There was a shrine to Apollo there, and an oracle. In the fourth century, evidently, there was some belief that the Castalian spring was located in Daphne.Footnote ⁹⁷ Indeed, one late version of the story of Apollo and Daphne transferred the setting of the event itself to Daphne.Footnote ⁹⁸

Finally, there may be some connection between the place and Martius’ patron Auxentius. Martius’ collection is not the only place where the river Cydnus and someone named Auxentius are both mentioned. In Adana, next to Tarsus through which the Cydnus flows, there is preserved probably from the fourth century a stone with a poem of twelve lines in elegiac, addressed to an Auxentius, describing a building project (either a bridge or an aqueduct), and mentioning the Cydnus.Footnote ⁹⁹ Whatever the epigram is referring to, we do know that an Auxentius was a local notable of Tarsus, which makes any connection with the Cydnus more than incidental.Footnote ¹⁰⁰ It is more than strange that the only two places a personage named Auxentius is mentioned along with the river Cydnus are in a Greek poem of eight elegiac couplets found inscribed on a stone being used as an altar at a church in Adana and in a collection of Latin bucolics transmitted in a manuscript written around 1200 in France. Such are the byways of the transmission of the legacy of the late antique world.

In short, there are many possible reasons why the first bucolic refers to Lauri altae. Earlier scholarship was perhaps wrong to assume that it would tell us something about the identity of the poet, as if Martius were identifying his own home. Instead, Lauri altae is a definite place in the bucolics’ literary world, and one that we must presume his audience would have been able to identify.

Hence none of the individual features of the collection which have been identified as definitive evidence of medieval composition withstand close scrutiny, and present no bar to following the internal and external evidence in assigning their date of composition to the sixth century.

VIII THE TRANSMISSION OF THE BUCOLICS

The history of the transmission of Martius must begin in the sixth century. The first sure glimpse we get is in Venantius Fortunatus, who seems to have known the poems: ‘aut hyacintheo sudat honore manus’.Footnote ¹⁰¹ The word hyacintheus is significant, listed in the TLL as only occurring here and in one other passage of Venantius. In both cases it refers to stones and not flowers. But it occurs in Martius as we have seen above (buc. 4.75–6):

The shared use of an otherwise unknown word links Martius and Venantius very closely; that both of them also use honore in what are otherwise two very different passages confirms it. But the primary sense of all hyacinth- root words is the flower, not the stone. This suggests what our chronology requires anyway — that Venantius, who died around the beginning of the seventh century, is the imitator.Footnote ¹⁰²

To pursue the history of Martius’ text further we need next to consider the lost manuscript(s). The Abbey of Thorney must have had a library with unusual holdings. Of the five manuscripts Leland mentions, two contain astounding rarities: our Martius and Isagoge Porphirii Victorino interprete. The latter work does not survive anywhere independently, although Boethius made extensive use of it in his translation and commentary on Porphyry. Caution is in order, however: the fact that Victorinus had translated the text was widely known from Isidore (etym. 2.25.9), and some manuscripts of the first edition of Boethius’ commentary specify Victorinus’ name (such as in Isagogen Porphirii … a Mario Victorino translatum).Footnote ¹⁰³ Surer are two Carolingian gems from Thorney in Edinburgh, NLS, Adv. 18.6.12 (A.6.4), the third book of Abbo's Bella Parisiaca (ff. 36r–v) and an anonymous epic Gesta Ludovici imperatoris (ff. 32r–33r), preserved nowhere else.Footnote ¹⁰⁴ Abbo's third book did achieve a modest circulation mostly in England, but the first two books are exceptionally rare, surviving to this day in a single manuscript, which may well never have left the city of Paris in its twelve centuries. The Thorney manuscript is uniquely linked to the Paris manuscript by the preface which calls it the third book.Footnote ¹⁰⁵ Around 980, Aethelwold donated a library to Peterborough Abbey, not eight miles from where Thorney would be founded, containing a Descidia Parisiacae polis.Footnote ¹⁰⁶ How much of Abbo's text it contained is unknown, but the fact that he calls it polis and not urbis, strongly suggests that he had seen the introductory letter or the explicit of the second book, which both use the phrase Parisiacae polis.

Interestingly, we can also connect Guido of Grana to England. One manuscript he corrected and annotated with his own hand is now in Edinburgh.Footnote ¹⁰⁷ He wrote:

Debate continues about John of London;Footnote ¹⁰⁸ absent definitive evidence to the contrary, however, there is no reason not to put him in England. Further, one of our only external testimonia to Guido's writings are two books in the library of Syon Abbey not far from London two centuries later.Footnote ¹⁰⁹ In fact, many of the rare texts Guido knows we know had been in England — Petronius, Calpurnius and Propertius. In 1423, Poggio wrote to Niccoli, ‘Please send me the bucolic of Calpurnius and the section of Petronius, which I sent to you from England’ (ep. 2.3: ‘Mittas ad me oro Bucolicam Calpurnii et particulam Petronii quas misi tibi ex Brittania’, vol. 1. p. 91 Tonelli). Reeve has suggested that Poggio's manuscript is the source of one branch of the tradition of Calpurnius, which does not distinguish between the seven eclogues of Calpurnius and the four of Nemesianus.Footnote ¹¹⁰ Guido likewise makes no distinction.Footnote ¹¹¹ Propertius is first mentioned in the medieval period by the author of the De septem septenis who may have been active around Peterborough in the 1160s. In the same way, Ernesto Stagni has shown that Guido had seen the first two books of Abbo's Bella.Footnote ¹¹² Is this because he saw the Paris manuscript, or because he had seen one in England? The clincher is the text: Guido's text of Martius is (from the slim evidence we have) not derived from the Gotha manuscript (G). In the space of the three lines he cites, Guido gives three correct readings not found in G. More importantly, G does not give Martius any title, whereas both Guido and the Thorney manuscript do give him late Roman titulature. Both also give him the incorrect name Marcus, where G gives us the correct Martius. This strongly suggests that they are both a witness to a tradition of the text that transmitted the author's name with a string of titles, which wrote out the genitive Marti (leading to Leland's Marci and Guido's Marcus) instead of Martii.

The outline of Martius’ poems grows very clear, descending in two branches, one represented by the Gotha manuscript, the other by Guido and the Thorney manuscript. The second cannot be derived from the first, although there is no particular reason why the Gotha manuscript could not have been a copy of the other.

The Gotha manuscript's origins are unknown: the first place it is mentioned is in Amplonius Ratink's library catalogue of 1410–12, listed as liber 5 bucolicorum Marcii Valerii Maximi. (The final word was probably added as a conjecture or a mental lapse, through confusion with the more famous M. Valerius, the Tiberian paradoxographer.) There is no doubt that this refers to the same manuscript, and that it had made its way to Erfurt. A century later, someone went through and annotated the manuscript; around the same time it was copied, producing manuscript E, now in Erlangen.

In his 1791 catalogue of Altdorf's manuscripts, Christoph von Murr lists E among the books of the bibliotheca vetus academica, and so presumably including the books held before the academy became a university in 1622. Von Murr calls the text a Liber spurius: understandably he considered an unknown Latin literary work by an author with a classicizing name in a fifteenth-century manuscript (so he dates it) to be yet another specimen of the unimaginably vast roster of neo-Latin literature. Around the same time, the Benedictine monk Jean-Baptiste Maugérard had acquired Amplonius’ manuscript from Erfurt, and evidently read it, since his hand can still be seen in the volume.Footnote ¹¹³ In 1795, he sold it to Duke Ernest II of Gotha-Altenburg.

Forty years later, in 1838, as Friedrich Jacobs was cataloguing the Gotha manuscripts, he could not have recourse to the same theory, since the Gotha manuscript was plainly medieval.Footnote ¹¹⁴ Jacobs does not attempt a specific dating, nor does he evaluate the poems beyond suggesting that the dedication to patres indicates that the poet was young. Nonetheless, the fact that he catalogued it in the category of Lateinischer Dichter und Redner des Mittelalters is argument enough. Likewise, in 1904, Traube and Ehwald in their study of the Maugérard manuscripts had nothing new to say on the possible origin of the text.Footnote ¹¹⁵ The trail had gone cold enough by the second decade of the twentieth century that even Sabbadini could be baffled by the entry in Amplonius’ catalogue, and assume it referred to a lost manuscript.Footnote ¹¹⁶

What this survey of the pre-Lehmann evidence for Martius reception shows is that at precisely the period in which the canon of minor Latin literature was being crystallized — the first volume of Riese's Anthologia latina sive Poesis latinae supplementum was published in 1869 — Martius Valerius remained virtually unknown. The Einsiedeln Eclogues were discovered in an era that still believed there was more ancient literature to be found in manuscripts — indeed, Cardinal Mai was dead not fifteen years before Hagen produced the editio princeps. Hence there was astonishingly little scepticism attending their swift canonization as Neronian. But Martius’ bucolics were born into a very different world, one which set the bar very high indeed for any work to be assigned to antiquity. In other words, there seems to have been a presumption in the nineteenth century and before, that if a work could be either early or late, it was probably early, while today we operate under the Winterbottom rule, that is, if a text could be either early or late, it is probably late.Footnote ¹¹⁷ As a scholarly instinct, this is undoubtedly correct. And yet as it turns out, in the case of Martius Valerius, further digging should have been undertaken.

IX A DEBATE CLOSED

Dolbeau was right to characterize the question of the dating of Martius Valerius as ‘un débat prématurément clos’.Footnote ¹¹⁸ Let us return at last to the facts as they stood five years before Dolbeau's article. Four pieces of evidence — all of them available in print at the time, and three of them in print from the date of the first edition — point to a single possible solution. The name Martius Valerius is a Roman and not a medieval name. The only ‘real’ people mentioned in the eclogues are Auxentius and Faustus, and they can only be late antique Romans. The only external testimonia for the poems assign Martius offices of the late Roman state, and put his floruit under Justinian. To try to dismantle each piece of evidence separately — to claim that Martius Valerius is a pseudonym, that Auxentius is a pun, and that the two testimonia mean something other than what they plainly state — is nothing other than special pleading. The additional evidence adduced here, that Martius Valerius knows Theocritus, is writing for an audience at least conversant in Greek as well as Latin, and that he knew ancient material which was transmitted through the slenderest of lines from the Carolingian period, merely corroborates what Dolbeau already argued in 1987. His question should now be considered ready for timely closure: the bucolics of Martius Valerius are not a medieval production, but a witness to the literary florescence of the fifth and sixth centuries.

Virtually everyone who mentions Martius gives him an epithet denoting bafflement: ‘misterioso poeta’ (Orlandi), ‘l'enigmatico Marco Valerio’ (Salemme), ‘the mysterious M. Valerius’ (Walsh), ‘le mystérieux Marcus Valerius’ (Dolbeau, Tilliette, Meyers), ‘l’énigmatique Marcus Valerius’ (Dolbeau).Footnote ¹¹⁹ But it is not as if his works are massively obscure; the mystery, rather, arises from the disjunction between his poetry and the supposed context in which he was writing. The reason Martius has absolutely nothing in common with an actually twelfth-century bucolic poet, Metellus of Tegernsee, is because they do not share a similar literary world. Instead, the relation between Martius and Metellus is roughly that of Maximian to Henry of Settimello. Once Martius is restored to the sixth century his milieu is startlingly clear. He is a revivalist poet in a revivalist era. Along with Boethius, he took to bucolic, just as Maximian took to elegy, Dracontius to tragedy (in a fashion), Corippus to historical epic, and Luxorius to epigram.Footnote ¹²⁰ Like Maximian, Priscian, and others, Martius interacted directly with his Greek models.Footnote ¹²¹ Indeed, he shares much in common with the author(?) of the Appendix Maximiani writing under Theodahad (534–36): both of them mention the story of Hero and Leander together with that of Apollo and Daphne, which is not the most natural collocation (App. 2.5–6 and 9–10 and Martius, buc. 4.56–67 and 83–90), and the Appendix poet artfully toys with bucolic conventions in his description of a villa (e.g. 5.5 ‘hic gelidi fontes hic dulces arboris umbrae’ with Virg., ecl. 10.42: ‘hic gelidi fontes, hic mollia prata, Lycori and Martius’, buc. 1.85 ‘nunc dulces umbrae’).Footnote ¹²²

More work will undoubtedly shed light on this poet and dispel the lingering enigmas. I hope I have demonstrated that Martius is a sophisticated and sensitive poet, whose works will amply repay close study and attention. A full commentary would uncover further riches, and help situate Martius in the literary world of the sixth century. For now, a few minor suggestions of the implications. If the first bucolic does indeed reference Antioch and its arx, it is perverse to think that the poems post-date 540 when the city was sacked. The striking lexical innovations of the bucolics ought to be included in the Thesaurus linguae latinae: certainly indelimatus and blandisonus, and probably sublangueo as well (the other two hapax, imposco and subpenetro, are likely textual corruptions).Footnote ¹²³ If I am correct as to the inspiration provided by Persius, Martius also provides some indication that the choliambs were transmitted as a preface to the collection, at least in the sixth century. Latin bucolic should now be considered to have a virtually continuous history after Virgil from the third century to the sixth, with three significant bodies of material, Virgil's ten eclogues, the eleven of Calpurnius and Nemesianus, and the combined eight of Olybrius, Endelichius, Pomponius and Martius. The line that runs between them — as the slow work of dating and contextualizing continues — is the outline of the history of one of the most characteristic classical genres in the post-classical world, a history that has yet to be told.