Scholia Reviews ns 14 (2005) 40.

'The Prince and the Stars: Germanicus' Translation of Aratus'

D. Mark Possanza, Translating the Heavens: Aratus, Germanicus, and the Poetics of Latin Translation. Lang Classical Studies Vol. 14. New York: Peter Lang, 2004. Pp. xiv + 279. ISBN 0-8204-6939-4. Sfr.109.00.

Emma Gee,
Classics and Ancient History, University of Sydney

Possanza's book represents an interesting contribution to studies of Aratus' Phaenomena.[[1]] Focussing on Germanicus' translation, Possanza aims to show that Germanicus re-interpreted the Phaenomena using Greek as well as Latin predecessors as part of a continuous tradition (pp. 1-20 and 112-114). Chapter 1 (pp. 21-77) studies the poetics of translating Greek poetry into Latin. Chapter 2 (pp. 79-99), characterises Aratus' work as both a descendant of oral catalogue poetry, and a masterpiece of Callimachean refined style. Chapter 3 (pp. 105-67) selectively examines Germanicus' method of translation. Chapter 4 (pp. 169-218) shows how Germanicus changes the Phaenomena in translation, Possanza's conclusion being that 'The Greek poet's lofty theme of the constellations as "signs" of the providential deity's immanence in nature is completely subverted and in its place we find no theme of comparable religious and philosophical significance. Instead we discover that it is the poet himself who controls this cosmos, who as a storyteller and self-declared vates (bard) turns the map of heaven into a realm of Ovidian transformations where the revelation of what the constellations once were humanizes and dramatizes the existence of those distant astral bodies' (p. 208).

All of Possanza's arguments are based on the view, argued in Appendix A (pp. 219-43, see also pp. 15f. and 105-109), that Germanicus the son of Drusus is the author and that the poem was composed between AD 4 and 7.

The manuscript evidence for authorship is 'inconclusive' (p. 220), pointing, if anything, more strongly to Tiberius. The name Germanicus is not found in any of the primary mss., and the attribution depends on a later indirect tradition (Lactantius Div. 1.21.38, 5.5.4, 1.11.64, supported by Jerome PL 26.606.706b[Migne], and possibly Priscian[[2]]). Possanza argues that the name found in the O family, T[i] Claudi Caesaris Arati Phaenomena, is an interpolated form of the name Germanicus Caesar, the name Germanicus being re-introduced in the fifteenth century.

In Possanza's view, 'clues' (p. 227) in the first sixteen lines of the poem help confirm Germanicus' authorship. These are that the dedicatee is the emperor, that this person maintained peace on land and sea, and that he had a son. The most natural candidate for the emperor is said to be Augustus (see pp. 231f.). If Augustus is the one being addressed, it might seem to follow that Tiberius is the author. Nonetheless Possanza sees the clues as pointing towards Germanicus. Others have differed; according to Gain, the evidence does not allow one to say whether Germanicus or Tiberius composed the poem.[[3]]

In order to argue for the authorship of Germanicus while retaining Augustus as the dedicatee, Possanza must bridge a generational divide. A lot rests on his interpretation of lines 15f.: haec ego dum Latiis conor praedicere Musis, / pax tua tuque adsis nato numenque secundes ('While I attempt to set forth these things in Latin verse, may you and your peace attend your son and favour him with your divine presence'.) Possanza creates a disjunction between the authorial 'I', the subject of conor (I attempt), and the 'son', arguing that the son and the author of the proem are two different individuals, Tiberius and Germanicus respectively. As Possanza translates (p. 106): 'While I make my attempt to foretell these things, may your peace and you yourself be by the side of your son, and may you make your divine majesty favourable'. In Possanza's version, three separate things are happening in these lines: (a) the poet is writing (temporal clause, related to what follows only in terms of its contemporaneity), (b) the poet is asking the dedicatee to favour his son (not the poet), and (c) the poet is asking for this person to make his numen (divine presence) generally favourable.

In my view, the parallels make it more natural to understand these lines as a prayer for poetic success for the poet/son, the same individual, involved in the poetic task. In Manilius, Caesar is hailed as a deity in the context of his favouring the poet in his poetic undertaking:

hunc mihi tu, Caesar, patriae princeps paterque,

qui regis augustis parentem legibus orbem

concessumque patri mundum deus ipse mereris,

das animum viresque facis ad tanta canenda. (1.7-11)

('You Caesar, Princeps and Father of the Fatherland, you who rule your father's heaven with august laws and, yourself a god, are worthy of the place in the sky given to your father -- it is you who give me this resolve and grant me the power to sing of such great matters'.)


There is no separate prayer for the imperial family or for generalised favour. In the proem to Ovid's Fasti (1.5f.), Germanicus is the numen, who favours the work dedicated to him (officioque, . . . / en tibi devoto numine dexter ades 'come, favour with your godhead the work dedicated to you'.) Later in the proem (Fasti 1.15) he is asked to approve the author (adnue conanti per laudes ire tuorum, 'approve me as I attempt to sing your praises'). At no stage in the Fasti proem is there a prayer for the imperial family or a generalised prayer for him to make his divine majesty favourable, as there is in Possanza's translation of Germanicus Phaenomena 15f, quoted above.

In fact, the proem to Ovid's Fasti would have helped a lot in Possanza's argument. As it is, he omits evidence which may point to Germanicus' composition of an astronomical poem, despite arguing strongly for his authorship of the Phaenomena. In the Fasti proem, Germanicus is hailed as a poet (vates, Fasti 1.25), most likely as author of the Phaenomena and as predecessor of Ovid in the astronomical part of his task. This supports the hypothesis that Germanicus' Phaenomena had been written by the time the Fasti was revised, some time after Ovid's exile (between AD 14 and 17?).[[4]]

Analysis of the relationship between the Fasti and the Phaenomena would be useful in settling the date of the latter. Given Possanza's programme of demonstrating that Germanicus' Phaenomena is closely related to Ovid, it would be worth asking which version of the Fasti -- pre-exilic or revised -- Germanicus was working with. Are we to see the composition of Germanicus' Phaenomena and Ovid's Fasti as proceeding hand-in-hand, both written around AD 4 (the Fasti a little earlier), both revised after AD 14? Or are we to see the entire Phaenomena as written after AD 14, with full knowledge of the Fasti?

The most serious obstacle to identifying the reigning Augustus as dedicatee of Germanicus' poem is the reference to the apotheosed Augustus at lines 558-60. Possanza therefore argues that the poem must have been revised and these lines inserted after Augustus' death, its original composition taking place shortly after AD 4, when Tiberius became Augustus' son. Possanza puts the terminus ante quem for the original version at AD 7, when Germanicus took up imperial responsibilities. In addition, he argues that Ovid's exile in AD 8 would have prevented Germanicus from following in that poet's footsteps by giving prominence to the theme of illicit love.[[5]] But could not the numen of Augustus which appears in the proem be that of the apotheosed emperor, as it is in 558-60? In this case, all of the poem could have been composed after AD 14.[[6]] The use of Ovid's exile to support an early terminus ante quem is hypothetical.

Are acrobatics with transmission, nomenclature, and date ultimately useful? Surely the point is that although Germanicus' Phaenomena could not be earlier than Ovid's Fasti or Metamorphoses, its Ovidian nature need not determine its authorship. Or does some sort of Tacitean characterisation ('If it's Ovidian, it has to be by Germanicus, rather than crusty Tiberius') underlie Possanza's arguments? The important question is what the ascription to Germanicus does to our reading of the poem. Will this reading differ depending on whether the poem was written by a young Germanicus under Augustus, a more mature Germanicus under Tiberius, a young Tiberius under Augustus, or a not-so young Tiberius after Augustus' death? Possanza enagages in little consistent argument about the political context. Although he states that 'the political ideology of the Augustan age exerted a powerful influence on the way in which [Germanicus] read and interpreted the Phaenomena' (p. 36), he remains throughout more interested in poetics (with the exception of the discussion of the proem, 105ff.). Yet the date of Augustus' banishment of Ovid was given by Possanza as a terminus ante quem for the composition of Germanicus' Phaenomena. Possanza cannot use Augustan politics to date the poem externally without considering the role of Augustan politics in reading the poem as a whole, and forming a view of the 'Augustanism' or otherwise of the poem in a more than purely literary sense. This is not to argue for a mindless return to the old 'subversion' theme of Ovidian scholarship[[7]] -- merely for more overt recognition of the co-extension of the literary and political dimensions of the piece.

Another element 'lost in translation' is Aratus' Stoicism. Possanza refers to Aratus' Phaenomena as 'theistic' (for example, p. 114). Stoicism should be mentioned as the driving force behind Aratus' teleology, and recognition of it would be helpful in clearing up a number of details. For example, it is stated that, 'just as Germanicus's omission of Aratus' hymn to Zeus signalled his abandonment of the Phaenomena's theological and philosophical perspective on the order of the universe, so his omission of the passage on the naming of the stars [Ph. 367-85] continues that deliberate program of editing to remove any suggestion that humans had a role to play in forming and naming the constellations' (pp. 207f.). But here Germanicus is specifically rejecting the idea of the natural connection between signifier and signified which underlies Stoic theory of language and its relation to theology.[[8]] Rejection of Stoicism is one reason for Germanicus' anti-theological stance.

There is another unexplored possibility. One of Germanicus' poetic predecessors is Lucretius. In discussing Germanicus' rejection of Aratus' weather-signs and substitution of a new meteorology which includes the planets (pp. 110f.), Possanza re-iterates that there is no room in Germanicus' poem for Zeus' semata. His argument rests on Ph. 12, sideraque et mundi varios cognoscere motus ('to learn about the heavenly bodies and the various motions of the heavens'), which he takes to refer to the 'various' movements of the planets, as opposed to the regular movements of the fixed stars. But this line brings to my mind at least Lucretius 5.774f, solis uti varios cursus lunaeque meatus / noscere possemus ('so that we would be able to know about the various motions of the sun and the movements of the moon'), in a passage where the poet specifically argues against a theological interpretation of heavenly signs. It was Lucretius who gave the Romans a rationalistic way of thinking about natural phenomena; surely his influence should be considered.

Cicero did not have Lucretius to draw on in his Aratea. This could help partially to explain the differences between his and Germanicus' translations of the Phaenomena, as well as the factors mentioned by Possanza, who states that Germanicus' narrative voice is 'engaged in an intertextual dialogue with the Phaenomena or with Cicero's translation or with both' (p. 201). According to him, Cicero is a 'negative influence of what was to be avoided because his translation in its language and meter represents the epico-tragic tradition of the old republican poetry' (pp. 115f). He explains the differences between Cicero and Germanicus thus: 'When Germanicus came to translate the Phaenomena sometime between AD 4-7, that fullness and weightiness of expression [found in Cicero's Aratea] had been disciplined and reduced by a strict regimen of Hellenistic poetics which had been adopted and mediated into Latin poetry by the Neoterics and Augustans' (p. 28).

Cicero is characterised as primitive, whereas Germanicus writes in accordance with Hellenistic poetics. However, we should not let ideas of poetic evolution blind us to Cicero's own role in constructing the opposition between his poetry and that of the neoteroi, an opposition which in any case may not obtain for the Aratea, an early poem on an Alexandrian theme, which introduced refinements the neoteroi and their successors were glad to adopt.[[9]] Consider Aratea 35f. (on the Pleiades): Alcyone Meropeque, Celaeno Taygeteque, / Electra Asteropeque, simul sanctissima Maia ('Alcyone and Merope, Celaeno and Taygete, Electra and Asterope, and also most holy Maia'). This predates Virgil's use of the device of filling a hexameter with Greek names, as at Georgics 1.437, Glauco et Panopeae et Inoo Melicertae ('to Glaucus, to Panopea and to Melicerta, son of Ino'). Cicero was on the cutting edge of this poetic practice (Quint. 12.10.33).

Alexandrian poetics were already available for Cicero, as for his predecessors Livius Andronicus and Ennius. He chose in the Aratea to combine Alexandrian aspects with the diction of earlier Latin epic. This choice is a highly appropriate one. How better to render Aratus' Homeric dialect than to draw on the diction of early Latin epic, which itself strives to imitate Homer? Not evolution, but differing principles of choice can be said to operate in Cicero's and Germanicus' translations of Aratus. Whereas Cicero reads Aratus as Callimachean epic, Germanicus reads Aratus as Callimachean epic.

Cicero is a better model for Germanicus than Possanza admits. Both play at enacting poetic secondarity. Germanicus excuses his variant version of the Orion myth with the words haec ego non primus, veteres cecinere poetae (Ph. 647, 'I am not the first to sing of these things: the ancient poets did too'.) According to Possanza (p. 198), veteres poetae can be taken as a reference to actual predecessors, including Cicero. But he misses the force of the intertextual play; the phrase is a quotation from Cicero, albeit a different passage (the Pleiades again): sed frustra, temere a vulgo, ratione sine ulla / septem dicier, ut veteres statuere poetae (Aratea 33f., 'but it is an empty and rash belief of the common people, based on no reasoning, that [the Pleiades] are seven, as the ancient poets established.') Here Cicero is sceptical of tradition. Acknowledgement of this would make Possanza's argument about Germanicus' 'disclaimer' (for the myth, in his retelling) stronger. Cicero, a self-conscious witness to his own intellectual thoroughness in the Aratea, belongs in Germanicus with the other purveyors of spurious tradition he sought to discredit. At the same time, Germanicus playfully acknowledges his poetic debt to his predecessor, critical dialogue marking respect of one author for another.


[[1]] D. Kidd, Aratus: Phaenomena (Cambridge 1997), has recently made Aratus more accessible.

[[2]] Inst. Gramm. ed. Keil 2.351.4 (correcting Possanza's 3.351.4) and 3.417.1. Further argument is required to establish that Priscian's Caesar in Arato is indubitably 'shorthand for Germanicus Caesar in Arateo carmine'(p. 223).

[[3]] See D. B. Gain, The Aratus Ascribed to Germanicus Caesar (London 1976) 16-20. Generally Possanza marshals the existing scholarship well, but add to bibliography Stephen Hinds, Allusion and Intertext (Cambridge 1998), M. Fantuzzi and R. L. Hunter, Muse e modelli: la poesia ellenistica da Alessandro Magno ad Augusto (Laterza 2002) 533-66, and K. Volk, The Poetics of Latin Didactic (Oxford 2002), especially for Manilius, something of an absent presence in Possanza's book, although he touches upon this author in n. 34 p. 103.

[[4]] On the dates of composition and revision of the Fasti, see G. Herbert-Brown, Ovid's Fasti: An Historical Study (Oxford 1994) 32f., with the bibliography in n. 1. On Ovid and Germanicus, see R. E. Fantham, 'Ovid, Germanicus and the Composition of the Fasti', PLLS 5 (1986) 243-81.

[[5]] The Fasti and Metamorphoses 'provided the models for Germanicus' handling of the erotic elements in many of the catasterism myths' (pp. 234f). See also p. 169.

[[6]] Alternatively, if one does subscribe to the theory of early composition and later revision, that still does not rule out Tiberius' authorship. Gain [3] 20 says, 'It is conceivable that Tiberius composed most of the poem many years before [the death of Augustus] and added lines 1-16 and 558-60 . . . only after Augustus' death, thus producing a sort of second edition'.

[[7]] Most clearly articulated in C. Newlands, Playing with Time: Ovid and the Fasti (Cornell 1995).

[[8]] See M. Frede, 'Principles of Stoic Grammar', in J.M. Rist (ed.), The Stoics (Berkeley and Los Angeles 1978) 27-76; E. Gee, Ovid, Aratus and Augustus (Cambridge 2000) 73f.

[[9]] See Hinds op. cit. n. 41 p. 75, and E. Gee, 'Cicero's Astronomy', CQ 51 (2001) 520-36.