The Aeneid and Archaism

From Vol. 2, No. 1, Spring 2017

By Keita Kashima, Oxford University


This paper examines linguistic archaism in the Aeneid. Particular attention is paid to archaic morphology from the perspective of Augustan poetry, that is to say, forms of words that belong to older strands of the Latin language and are therefore not so frequently encountered in Classical Latin we are used to reading.

Section §1 lists methodological problems and considerations. In Section §2, I examine how archaism, especially archaic morphology, may contribute to the thematic complexity of the Aeneid. In particular, I attempt to answer the following three questions:

(1) What does archaism add to the reader's understanding of temproal relationship in the Aeneid?

(2) Are there discernible patterns in the use of archaism?

(3) Does archaism contribute to characterisation?

Quite a few of the answers I give are either negative or inconclusive, mainly because there are not enough archaic forms in the Aeneid to make definitive statements about them. But some usages of archaic forms are interesting from the thematic perspective of this poem. Moreover, the general lack of archaic forms may be useful in understanding the position of the Aeneid within the Roman literary history, and perhaps even Virgil’s attitude towards the contemporary era.

§1 Preliminary remarks and methodological restrictions

Virgil’s fondness of antiquity is suggested by Quintilian’s comment (Inst.1.7.18) that the poet was amantissimus vetustatis (‘most fond of antiquity’). Quintilian describes Virgil in this way when explaining the presence of the old genitive ending –āī [i] in pictai vestis (9.26) and aquai 7.464) in the Aeneid. He also praises Virgil for his use of fossilized forms such as olli, quianam and moerus (Inst.8.3.25). Archaism brings, according to Quintilian, orationi maiestatem aliquam non sine delectatione (Inst.1.6.39). Therefore, at least Quintilian was ‘delighted’ with the presence of archaic forms in the Aeneid. In this short paper, I shall examine whether one could go further than Quintilian’s statement that they are merely ‘delightful’ for their ancient solemnity by attempting to answer the three questions which I raised in the overview.

Before surveying Virgil’s archaism, it is necessary to introduce some limiting factors. This is because one definition of ‘archaism’ would be too broad in our examination here. ‘Archaism’ in poetry may be taken to mean any linguistic feature that makes the author’s language old and solemn, and this is usually achieved by employing linguistic forms, style, and vocabulary which are obsolete and are therefore associated with older language (e.g. thou for modern English you). To examine every aspect of ‘archaism’ in the Aeneid, as defined above, would be difficult in this short paper.

Therefore, to simplify my examination, I do not take into account, in most cases, the characteristic features of pre-Augustan poetic language that are mentioned by Palmer, including asyndeton, tricola, assonance and, most notably, alliteration to the extent that it has no correspondence in Greek poetry.[ii] Three other possible methods of archaizing are the use of obsolete vocabulary preserved in the language of ancient law-codes and religious texts, the use of obsolete syntactic features, and the employment of fossilized paradigmatic forms. I shall concentrate on the latter two aspects, but especially on the last.

Despite having narrowed down the term ‘archaism’ in this paper mostly to its morphological aspect, numerous problems still remain. I give a summary of seven factors, labelled (a) – (g), which are relevant to archaic morphology:

(a) Metrical constraints. Many specialties of poetic diction must have risen under the constant metrical pressure, and the Aeneid is no exception to this. For example, the conjunction uti, which is found in 1.466, 2.507, 7.528 and 12.488, though archaic morphologically and frequent in Cato and Lucretius,[iii] is metrically convenient. In such cases, the possibility that Virgil used archaic forms for various literary effects could be reduced to an argumentum e silentio.

(b) Textual transmission. Archaic forms, which have the same metrical structure as their more recent counterparts, may have been altered through ‘normalisation’ by later scribes. Examples include divum and illi from divom and olli, or even vice-versa through ‘hyper-correction’.[iv] A thorough consideration of this problem involves a meticulous examination of various manuscripts, which cannot be done here. I follow the text provided by ‘Brepolis-Library of Latin texts’, and when an archaic form appears in the text, I shall accept its authenticity.

(c) Colloquialism. Archaic forms may have been retained as colloquialisms avoided in high prose literature. For example, the dative and ablative plural relative pronoun quis (e.g. 5.511, 8.316) may be conversational.[v] Therefore, the appearance of the dative-ablative plural quis in the poem does not necessarily mean that Virgil used it to emphasise the inclusion of such archaic forms in his poetry.

(d) Archaism embedded in fixed phrases. The phrase paterfamilias and its inflected forms, for example, are common in Classical Latin, including Cicero.[vi] Fixed expressions are likely to persist through the medium of legal and religious phrases (within the Aeneid, see for example 5.174 sociumque salutis) in both the spoken and the written language, even when its endings become obsolete. In such cases, archaic forms are not necessarily employed specifically for literary effects.

(e) Indirect archaic syntax through Graecicism. For example, the retention of the archaic use of the prolative infinitive instead of the gerund construction (e.g. 2.10 tantus amor casus cognoscere nostros) is influenced by Greek authors, especially Homer.[vii] In this case, it cannot be accepted straightforwardly as genuine archaism. Such grammatical constructions are superficially archaic, but it remains uncertain whether Virgil consciously employed them more for archaic effects by reproducing old Latin syntax than for his imitation of the Greek poetic style.

(f) Innovation resembling archaic syntax. Penney observes that ne + the present imperative in Old Latin has an inhibitive function, whereas ne + the subjunctive a more prohibitive one that includes future reference.[viii] Ne + the present imperative is found in Virgil, but the construction now also includes future reference (e.g. 6.95 tu ne cede malis).

(g) Latin poetic tradition and selective archaism. This is perhaps the most significant factor and therefore requires closer attention. Archaism is often a trait of historic Latin poetic tradition that (as far as we know) begins with Andronicus, Naevius and then Ennius, whose literary impact is illustrated by Virgil’s numerous references to the Annales.[ix] Equally important is the activity of the neoteric school in the late Republican period, whose members rejected many linguistic features of earlier poetry. Examples include the omission of the final –s, because it was considered a characteristic rusticitas associated with boorishness, ruining the urbanitas that they aimed for.[x]

Yet Catullus does retain obsolete forms,[xi] not only because of factors (a) and (c) above, but also as a result of the mimesis of previous poets and the continuation of the traditional Latin Dichtersprache. Imitation of the established poetic language manifests itself most clearly in the case of Lucretius who, according to Skutsch, uses the genitive singular ending –ae 153 times, but the obsolete alternative –ai more (166 times).[xii] The degree to which old Latin forms are employed differs according the poets’ preference, their school of thought and their degree of imitation of earlier poetry, and some forms are fully rejected,[xiii] whilst others are not.

§2 Answering questions (1), (2) and (3)

Archaic morphology in the Aeneid must be considered against this background of Latin poetic tradition (g) and the factors which I have listed in (a)-(f), many of which are not always easy to take into account. I begin with question (1), before moving onto questions (2) and (3), which I asked in the overview.

(1) What does archaism add to the reader’s understanding of temporal relationship in the Aeneid?

At this point, it is worth quoting from Wilkinson 1990. He cites Cordier’s work, who makes the claim that words which are ‘reserved to certain spheres such as religion’ and are ‘used in archaic constructions’[xiv] occur approximately once every 40 lines.[xv] This statistic leads Wilkinson to conclude that ‘an average of 2.52 different archaisms per hundred lines seems too low a percentage to diffuse a markedly archaic flavour throughout’.[xvi] Given the ratio, it is reasonable to conclude that archaism is not frequent in the Aeneid.[xvii]

But what if one were to suppose that the poem contained archaism less frequently than the Latin hexameter poems that were widespread at the time of its ‘publication’? It could be suggested, indeed cautiously, that the Aeneid defied the hexameter tradition and introduced a new poetic language that had less intense Republican linguistic colouring than its predecessors.

This hypothesis may be supported by an observation on the high proportion of the attributive genitive (G) that precedes the preposition (P) and the governing substantive (S). Penney 1999: 263-5 demonstrates that the structure G-P-S is found 42.5% of the time (in contrast to, say, P-G-S or P-S-G) in the Aeneid, in comparison to the remarkably lower proportion in Lucan (24%), which perhaps indicates Virgil’s conscious effort to include the structure G-P-S.[xviii] Since the structure G-P-S is not a feature found in Early Latin poetry, Penney raises the possibility that it is a Virgilian innovation, though Homeric influence cannot be excluded.

Perhaps one may find in this syntactic structure some modern, linguistic attitude that is fitting for the socio-political atmosphere of the compositional era of the Aeneid, the post-civil-war decade of 20s B.C., when there was surely ambivalence of hope and anxiety for the political future, as well as idealization of and regret for the past. Virgil’s infrequent recourse to archaism perhaps indicates his discouragement that his contemporary readers idealize their past, as it proves to be harmful to the future of New Rome under Augustus. Therefore, this possibly new Dichtersprache for Roman epic hexameter poetry[xix] is perhaps Virgil’s poetic-linguistic message to stress the requirement of New Rome to distance herself from her old Republican image.

Nevertheless, it remains true that Virgil was amantissimus vetustatis, for otherwise it is unconceivable, for example, why the poet would wish to use the 3rd person perfect plural ending -ēre, which is archaic and has a ritualistic tone, far more frequently than the endings -ěrunt/-ērunt.[xx] Here, one can argue that Virgil does accommodate some old morphological forms into his new Dichtersprache, and when he does so, he respects them by using them frequently. This may have a thematic correspondence to Aeneas’ accommodation of the past into the present, which one can discern, for example, in the imagery of the colours gold and purple and its Oriental, ‘Eastern’ association: Aeneas uses one of the two robes of gold and purple made by Dido (cf. 11.72-75) to wrap Pallas’ body and keeps the other for his personal use,[xxi] symbolically depicting the suppression of his Oriental identity [his Trojan origin, i.e. the past] but also its continuation in the Roman background [i.e. the present]. Virgil’s use of archaism may suggest that, just as the language of Republican poetry before the development of the neoteric school must continue to be respected but in a suppressed manner, likewise the image of old Rome and her relationship with old Italy must not be erased entirely as the Romans move onto their next chapter of politics. Therefore the first question I posed – whether Virgil’s use of archaism has any thematic significance in terms of temporal relationship – seems more than possible, although to prove this point is impossible: one would need to know how Virgilian Latin was received by his early readers, but there is simply not enough evidence on this matter. Moreover, and crucially, our evidence for Roman literature, especially pre-Virgilian literature, is too fragmentary to derive any secure linguistic argument.

(2) Are there discernible patterns in the use of archaism?

(3) Does archaism contribute to characterisation?

I now move onto questions (2) and (3), but I make some general comments on the results of the forthcoming enquiry. In many instances, the occurrence of archaism in the Aeneid is probably due to one or more factors in (a)-(g) or purely due to mere chance, therefore there are not, generally speaking, strongly systematic patterns in their employment (cf. question (2)). More disappointingly, the evidence is not substantial enough to come to a satisfactory conclusion for the role of archaism in terms of characterisation (cf. question (3)). But there are possible instances of literary effects – and perhaps very strong effects – which derive from archaic morphology, evanescent though they are. As for factor (g), I make no distinction between archaism that results from the imitation of previous poets[xxii] and one which the poet has chosen of his own accord.[xxiii]

Question (2)

I begin first with the forms olli/illi and ollis/illis.[xxiv] The form olle and its inflected forms fully ceased to be used by 200 B.C., and are not found even in Plautus and Terence.[xxv] It is likely to be an archaic feature revived deliberately by Ennius and then passed over to the Aeneid (as well as to Lucretius). Olle is restricted to the dative singular, the nominative plural (both olli) and the dative/ablative plural (ollis). These are the most frequent archaic forms in the Aeneid, and in the table below I have counted their occurrence and analyzed their position in the hexameter line (initially, internally or at the end of the line), and compared them with the data for illi and illis:[xxvi]





















Two observations can be made from the table above concerning olli/ollis:

(A) Initial olli is found more frequently than initial illi.

(B) Other uses of olli and ollis are very rare.

The five instances for (B) are 5.197 and 12.300 for internal olli, 5.358 for final olli, and 6.730 and 8.652 for internal ollis, and the following feature is shared in common: all are used to describe mortal figures (in one case, spoken by a mortal) in an elevated manner. Yet there seems to be no overwhelmingly strong correlation that connects the instances of their occurrence: 5.197, 5.358 and 12.300 can be regarded Iliadic,[xxxvi] 8.659 is about the dress and the appearance of old Gauls, and 6.730 is a part of Anchises’ speech about the reincarnation of mortal soul. Apart from the possible sense of solemnity given to mortals in various contexts, there is no systematic pattern.

Initial olli (A), however, in addition to its elevated sense, does seem to exhibit some patterns. First, olli is never elided, unlike illi which is always elided except in two occasions (1.210 and 2.86). Secondly, it is true that olli is often used when an Olympian talks to another Olympian (1.254, 4.105 and 12.839). The grandeur of the Olympian language is established firmly in the first occurrence of olli, when Jupiter addresses Venus (1.254: olli subridens hominum sator atque deorum). The poetic use of sator is followed by an etymological wordplay of fabor and fatorum (1.261-2), which intensifies the intellectuality of Jupiter. Therefore, early in the poem his ‘divine’ language distances the god from mortal characters. Indeed, the phrase olli subridens is used more by Jupiter (cf. also 12.839) than anyone else. And here it is possible to make a speculation; if the hexameter tradition confined the phrase to Jupiter only, one can say that Turnus’ use of the same phrase in 9.470 (olli subridens sedato pectore Turnus) points to his hubristic character, which has been exemplified earlier in the book, for example in his reaction to the metamorphosis of the Trojan ships into nymphs (9.128-158).[xxxvii] However, because the phrase itself is too rare in the Aeneid, there is no secure evidence that Virgil consciously wished to attribute the phrase to hubristic mortals, as in Turnus’ case.

Other uses of initial olli/illi seem to display no strong pattern: minor, non-Olympian gods can be described as illi (1.55), but they can also be involved in scenes where olli is found (cf. Allecto in 7.456-8); Aeneas and his men are both described at illi (e.g.1.210, 11.461) and olli (e.g. 8.94, 8.594); Aeneas and Turnus are both olli (12.788) and illi (12.720, in a simile); and perhaps paradoxically for an epic poem, warriors are described as illi (e.g. 2.86, 2.420) and farmers as olli (7.505).[xxxviii] Most occurrences of initial illi and olli, apart from the small proportion of olli when it is attributed to the Olympians, are perhaps influenced by Virgil’s mood as he composed his lines.[xxxix] The most sensible answer to question (2), therefore, is to accept the elevating effect of archaic forms in the context they are used, but to doubt the presence of definitive patterns, with the possible exception of initial olli/illi.

Question (3)

The poem is marked by the absence of passages containing a lexical cluster of archaic forms apart from (as far as I can find) one instance in the description of the delegates led by Venulus returning from Diomedes (11.225-251). In these lines, the neuter supine construction with a direct object (11.230: pacem…petendum), olli (11.236), the possible archaism in introgressi (11.248)[xl] and the archaic infinitive farier beside infit (242) are found, which does enhance the formality of the scene leading up to Venulus’ declaration of Diomedes’ response. But since the scene has no other correspondence in its unusual frequency of archaic morphology, any attempt to derive patterns from this single sample is hazardous. The passage from the council of the Italians is remarkable in its high usage of archaism, most likely in order to depict the Italians as rustic people who have long settled in Italy, in contrast to the Trojans, the wandering newcomers.

The paucity of evidence is directly relevant to the question of characterization I have posed in question (3): not enough archaic forms are uttered by or are ascribed to Aeneas, Turnus or any other character in the poem. For example, even if Virgil grants Dido to utter a sentence containing an inverse relative attraction in 1.573 (urbem quam statuo vestra est), an early syntactical feature found nowhere else in classical literature,[xli] in order to emphasise the solemnity of her diplomatic pronouncement, the only other occasion of her use of archaic morphology is accingier in 4.493. It is therefore difficult to see how archaism alone can contribute to characterization continuously.

One remark on Turnus and Aeneas, however, may be fascinating. Only these two characters have the privilege to utter what is probably the short-vowel subjunctive of the sigmatic aorist of the verbal root *fak- (cf. facio), that is, faxo, in contrast to Classical faciam.[xlii] Essentially, faciam displaced faxo whose use became marginalized to certain grammatical constructions, which too were eventually constructed with faciam. Examine the grammatical constructions of faxo in Aeneas and Turnus’ speeches:

Turnus (9.154-5): haud sibi cum Danais rem faxo et pube Pelasga/esse ferant…

Aeneas (12.315-6): … ego foedera faxo/firma manu…

Turnus’ lines in prose would go as follows: faxo ferant sibi haud rem cum Danais et pube Pelasga esse. Faxo here governs the present subjunctive ferant (‘they know, realise’ in this context) without the subordinating particle ut, and this construction is not infrequent in early authors such as Plautus and Terence.[xliii]

What about Aeneas’ lines, which in prose syntax are ego foedera firma manu faxo? Here, faxo takes a direct object (foedera) which, as de Melo 2007: 364 suggests, is probably due to the merger of the verb’s function with the more familiar form faciam. According to my research, this is the only example of the construction faxo + the direct object in the entire corpus of Latin literature. Though tentative, therefore, I think one could argue that Turnus’ conservative use of faxo gives him a Republican, Old Italic character,[xliv] in order to depict him as a figure ill-equipped for the imminent political change in Latium, whereas Aeneas proves himself to be a ‘modern’ figure, and a leader who can simultaneously respect a memory of old Italy through his innovatory syntax of the extra-paradigmatic faxo,[xlv] and through the initial alliteration (foedera faxo/firma).

Are there other linguistic traits like this for Aeneas and Turnus? Apart from Turnus’ other use of the sigmatic future, iusso (11.467), which may be an intensification of his Republican image, there is no discernible difference between the two characters as far as the morphology, syntax and the vocabulary of their speech are concerned. In this case, then, one could argue that Virgil did not want to polarize too greatly the two characters even from the linguistic perspective, in order to maintain the disturbing imagery of fratricide, which may be implied already in the name ‘Turnus’.[xlvi]


Archaic morphology is rare in the characters’ speeches, and even when archaic forms are employed their literary impact is mostly to evoke an evanescent sense of antiquity. The conclusion I therefore come to is that archaic forms contribute only temporarily and marginally to characterization (cf. question (3)) and that they are used without fixed patterns, with potentially a few exceptions in the case of olli/illi and ollis/illis (cf. question (2)). Modern forms pervade the poem, but certain archaic forms are still persistent, and this may indicate some thematic significance as argued in my answer to question (1). Overall, I feel that Quintilian is probably right in concentrating on the maiestas and the delectatio of archaism, without seeing any strongly systematic thematic concerns developed by the use of archaic morphology. But what one can appreciate from a survey of archaic forms in the Aeneid is the development of Latin poetry and its diction, and how remnants of the archaic language are combined with Virgil’s initiative to compose a new type of Roman hexameter poetry that is set in the heroic age but is also contemporary and national.[xlvii]



[i]. This is the ending before the familiar –ae. The long ī in –āī is analogical to the 2nd declension nouns, which have the genitive ending –ī. For the form –ās, which is older than –āī and which derives from reconstructed Indo-European *-eh2-(e)s, see endnote 13.

[ii]. Cf. Leonard R. Palmer, The Latin Language, (London: Faber and Faber Limited, 1954), p.109.

[iii]. Cf. Roland G. Austin, P. Vergili Maronis Aeneidos liber primus (Oxford: Clarendon Press,1971), ad loc.

[iv]. In the case of olli/illi, one has Quintilian’s statement above and therefore can be confident that Virgil used the form olli, but the case of divom/divum is more obscure, and similarly edit and edat in 12.801. Cf. Richard J. Tarrant, Aeneid. Book XII, Cambridge Greek and Latin Classics (Cambridge: Cambridge University Press, 2012).

[v]. Manu Leumann, “Die lateinische Dichtersprache”, schweizerische Zeitschrift für klassische Alterumswissenschaft (=Schweizerische Geisteswissenschaftliche Gesellschaft, Museum Helveticum. Basel:BSchwabe), Band 4 (1947): p.127.

[vi]. E.g. Cic. Ver. 2.3.120.

[vii]. For example, John H. W. Penney, “Archaism and innovation in Latin poetic syntax”, Aspects of the Language of Latin Poetry, Proceedings of the British Academy 93 (1999): p.256, quotes Homeric δῶκε δ' ἄγειν (e.g. Il.1.347).

[viii]. John H. W. Penney, “Archaism and innovation in Latin poetic syntax”, Aspects of the Language of Latin Poetry, Proceedings of the British Academy 93 (1999): p.253.

[ix]. Otto Skutsch, The Annals of Q. Ennius (Oxford : New York: Clarendon Press ; Oxford University Press, 1985), p.13.

[x]. Cf. Robert Coleman, “Poetic diction, poetic discourse and the poetic register’, Aspects of the Language of Latin Poetry, Proceedings of the British Academy 93 (1999): p.33 and Leonard R. Palmer, The Latin Language, (London: Faber and Faber Limited, 1954), p.109. Catullus 116 has therefore encouraged a philological debate, as it contains numerous archaic features (the old instrumental qui for ut in line 3, which is wholly spondaic, and the loss of s in dabis in line 8). Some argue for the early date of its composition, but Catullus is possibly mocking Gellius, who cannot appreciate the urbanitas of Callimachus’ poetry, through the language of the rusticitas.

[xi]. For a list of non-classical forms, see the introduction (xxviii-ix) of Kenneth Quinn, Catullus - The poems, Rev ed. (London: Bristol Classical Press, 1996).

[xii]. Otto Skutsch, The Annals of Q. Ennius (Oxford : New York: Clarendon Press ; Oxford University Press, 1985), p.61.The Aeneid restricts –ai to only four occasions (see endnote 39).

[xiii]. For an example, cf. Leonard R. Palmer, The Latin Language, (London: Faber and Faber Limited, 1954), p.98. Of the instances of the genitive –as quoted by Priscan (1.198f), three come from the Odyssey of Andronichus, two from the Bellum Poenicum of Naevius, and one from the Annales of Ennius. The ending –as is never found in classical poetry, which suggests that there is some limit to archaizing.

[xiv]. L. P. Wilkinson, “The language of Horace and Virgil”, Oxford Readings in Vergil’s Aeneid (1990): p.419.

[xv]. I was unable to find Cordier’s work and see how Cordier categorized archaic words and whether the work took into account the factors which I have listed above. At any rate, the statistics are sufficient enough for my argument below.

[xvi]. L. P. Wilkinson, “The language of Horace and Virgil”, Oxford Readings in Vergil’s Aeneid (1990): p.419. Perhaps the only instance that defies this is the Council of the Latins in Book XI, for which see my answer to question (3).

[xvii]. Indeed the overuse of archaic words seems to have been disapproved, as Quintilian (Inst.1.7.18) writes: sed utendum modo, nec ex ultimis teneris repetenda.

[xviii]. Alternatively and/or at the same time, Lucan’s conscious effort to distance himself from Virgilian Latin.

[xix]. I stress that it applies only to epic hexameter poetry, because in the Eclogues and the Georgics the use of archaic forms is comparatively less frequent.

[xx]. Cf. Robert D. Williams, Vergili Maronis Aeneidos liber quintus (Oxford: Clarendon Press, 1960) on 5.580. See also Charles F. Bauer, The Latin perfect endings -ēre and -erunt (Philadelphia: Linguistic Society of America, 1933), who states that the ending –ēre is found 231 times, in contrast to 29 times for -ěrunt/-ērunt. In fact, -ěrunt never appears in the hexameter poetry of Virgil, Horace (Satires) and Juvenal, which may be indicative of the marginal status of –ěrunt in poetry. Whilst metrical convenience cannot be rejected in some examples, the fact that Horace and Juvenal in their hexameter poetry have-ērunt far more frequently clarifies Virgil’s elevated language within its ‘modernness’.

[xxi]. James D. Reed, Virgil’s Gaze: Nation and Poetry in the Aeneid (Princeton, N.J.; Woodstock: Princeton University Press, 2007), p.82.

[xxii]. As mentioned in various commentaries.

[xxiii]. Once again, the paucity of samples from Latin literature is problematic in deciding when the use of archaic forms is not an imitation.

[xxiv]. As I have mentioned above, olli/illi (and ollis/illis) have a high chance of scribal corruption (Section §1, point (b)). I regret that I have been unable to look at various manuscripts and the works of editors. But I hope that the general correlation found in observation (A) has not been distorted.

[xxv]. Otto Skutsch, The Annals of Q. Ennius (Oxford : New York: Clarendon Press ; Oxford University Press, 1985), p.64.

[xxvi]. I used “Brepolis-Library of Latin texts” to gather the data.

[xxvii]. 1.254, 4.105, 5.10, 5.284, 5.580, 6.321, 7.458, 7.505, 8.94, 8.594, 9.740, 10.745, 11.236, 12.18, 12.309, 12.537, 12.788 and 12.829.

[xxviii]. 5.197 and 12.300.

[xxix]. 5.358.

[xxx]. 1.55, 1.210, 2.86, 2.420, 8.452, 10.146, 11.445, 11.461, 12.172 and 12.720.

[xxxi]. 1.138, 1.397, 1.416, 2.139, 2.212, 2.503, 4.261, 4.698, 6.162, 6.479, 6.879, 7.538, 8.118, 8.358, 8.466, 9.377, 9.754, 10.130, 10.172, 10.202, 10.219, 10.335, 10.355, 10.494, 10.572, 10.619, 11.73, 11.120, 12.52, 12.140, 12.333, 12.617, 12.710 and 12. 951.

[xxxii]. 5.644, 6.472, 8.443, 9.221 and 10.858.

[xxxiii]. 6.730 and 8.659.

[xxxiv]. 2.342, 3.214, 7.144, 7.685, 10.238 and 11.422.

[xxxv]. 3.98, 7.730 and 10.757.

[xxxvi]. 5.197: Mnestheus’ men after being incited by a speech of exhortation that reminds them of the final glory of Troy (5.189-191). 5.358: Aeneas smiles at Nisus and presents him with a gift, perhaps reminiscent of Achilles in Book XXIII of the Iliad (23. 555). 12.300: olli as a possessive dative, describing Ebysus’ burning beard in an Iliadic battle.

[xxxvii]. Philip R. Hardie, Aeneid. Book IX, Cambridge Greek and Latin classics (Cambridge: Cambridge University Press, 1994) ad loc. simply comments that ‘it adds to the awesomeness of Turnus’ appearance’, and this seems to have been Servius’ opinion too: magnam confidentiam virtutis ostendit.

[xxxviii]. One could, however, as my tutor has pointed out, make a counterargument here and claim that Virgil almost ennobles rustic life (cf. Philip Thibodeau, Playing the Farmer: Representations of Rural Life in Vergil’s Georgics, Joan Palevsky Imprint in Classical Literature (Berkeley ; London: University of California Press, 2011). The use of olli in 7.505 may be indicating the continuation of the spirit of the Georgics in the Aeneid.

[xxxix]. Likewise, apart from their vague ancient feeling, there is no correlation in the four instances of the genitive singular in –ai (3.354, 6.747, 7.464 and 9.26).

[xl]. See Nicholas Horsfall, Virgil, Aeneid 11: A Commentary, Mnemosyne, Bibliotheca Classica Batava. Supplementum. 244 (Leiden: Brill, 2003) ad loc. There is no way of telling if introgressi is a Virgilian coinage or an archaic verb, because there is no attestation of the word apart from in literature.

[xli]. See Roland G. Austin, P. Vergili Maronis Aeneidos liber primus (Oxford: Clarendon Press,1971) ad loc.

[xlii]. Since I do not wish to get into the details of the problem of faxo-faxim, I refer to Wolfgang D. C. de Melo, The Early Latin Verb System: Archaic Forms in Plautus, Terence, and Beyond (Oxford: Oxford University Press, 2007), and Michael L. Weiss, Outline of the Historical and Comparative Grammar of Latin (Ann Arbor: Beech Stave Press, 2009), p.419.

[xliii]. Cf. Wolfgang D. C. de Melo, The Early Latin Verb System: Archaic Forms in Plautus, Terence, and Beyond (Oxford: Oxford University Press, 2007) for more details.

[xliv]. There are numerous ways in which Turnus in the second half of the Aeneid is given an old Italic image, which cannot of course be examined here.

[xlv]. But it remains possible that faxo could take a direct object, but that examples of this construction are lacking.

  1. . Turnus’ characterization as Aeneas’ antagonist appears to be a traditional one, found for example in Livy 1.2. But whether all Roman authors were aware of the possible etymology of the name of Turnus remains uncertain. Since what I write below is not all relevant to the purpose of my work, I write an extensive footnote here. Those not interested in the Etruscan language should skip the paragraph immediately below.
  2. is probably true (given the archaeological evidence; see below) that the Etruscan equivalent of the god Cupid was ‘Turnu’, perhaps deriving from Turan, who is the Etruscan equivalent of Aphrodite/Venus, with some kind of suffix added. This may possibly be Etruscan -iu < IE *-ios. According to this assumption, the formation of ‘Turnu’ may have been something like Turnu (< *Turnnu, if the sequence *-ni- perhaps underwent assimilation or palatalization to *-nn-') < *Turniu (vowel syncopation) < *Turaniu (with loss of final *-s and interpretation of /o/ as Etruscan <u>) < *Turan-ios. But this derivation would be an ad hoc and, moreover, contradictory explanation, if Etruscan -ie, found e.g. in tarχunie, were from the same *-ios suffix. Alternatively, still advocating an etymological connection between Turnu and Turan, one may assume a purely Etruscan development, using the *-u suffix (Turnu < *Turan-u), found in participles such as mulu and agent nouns such as zicu (‘writer’). A third possible alternative would be to assume that an archaic inscription on the Aryballos Poupé from Caere preserves the archaic variant turannuve, which may be analysed as *Turan-nu-, with *-nu- suffix, which is possibly to be connected with other suffixes like -na. But this alternative would have to assume that <turannuve> is not a scribal error, as in the same text <turanuve> appears twice, without the double <n>. Overall, it remains difficult to speculate given the scarcity of relevant epigraphic evidence, and a simple derivation from Turan is in fact not guaranteed, if this name too is derived with a suffix (e.g.*Tur-an).

The sole evidence for the identification of Turnus with Etruscan Cupid comes from a bronze mirror excavated near Orvieto. On the mirror, a young boy (=Turnu) is sitting on the left with Turan (=Venus), Atunis (=Adonis) and Aplu (=Apollo) depicted in the central part, and all the divinities are labeled (cf. Erika Simon, “Greek myth in Etruscan culture”, The Etruscan World: pp.505-6).

The reason why I think that Virgil knew that the Etruscan equivalent of Cupid was Turnu is based on line 76 in Book 10. A goddess named Venilia is said to be the mater of Turnus, and her name resembles that of Venus. This genealogy of Venilia-Turnus seems to be, based on the available literary evidence, a Virgilian innovation, perhaps to intensify the imagery of fratricide: Aeneas is like Cupid, Venus’ son, and Turnus is like Etruscan Cupid and is born of a mother who resembles Venus. For more details on this theme, see Sarah L. McCallum’s PhD dissertation, Taking Love Seriously: Amor and Erotic Elegy in Vergil’s ‘Italian Iliad’, (University of Toronto, 2012), pp.92-95.

[xlvii]. I wish to thank my tutor Dr. Antony Smith for his suggestions in our tutorial in Hillary Term 2016, and the editors of Persephone for making stylistic suggestions and improvements. All other possible errors, whether factual or stylistic, are of course my own responsibility.



Austin, Roland G. P. Vergili Maronis Aeneidos liber primus. Oxford: Clarendon Press, 1971.

Bauer, Charles, F. The Latin perfect endings -ēre and -erunt. Philadelphia: Linguistic Society of America, 1933.

Coleman, Robert. "Poetic diction, poetic discourse and the poetic register", Aspects of the language of Latin poetry (J.N.Adams and R.G.Mayer (eds.)), Proceedings of the British Academy 93 (1999): pp.21-91.

Hardie, Philip R. Aeneid. Book IX. Cambridge Greek and Latin classics. Cambridge: Cambridge University Press, 1994.

Horsfall, Nicholas. Virgil, Aeneid 11: A Commentary. Mnemosyne, Bibliotheca Classica Batava. Supplementum. 244. Leiden: Brill, 2003.

Leumann, Manu. “Die lateinische Dichtersprache”, schweizerische Zeitschrift für klassische Alterumswissenschaft (=Schweizerische Geisteswissenschaftliche Gesellschaft, Museum Helveticum. Basel: BSchwabe), Band 4 (1947): pp.116-140.

de Melo, Wolfgang D. C. The Early Latin Verb System: Archaic Forms in Plautus, Terence, and Beyond. Oxford University Press, 2007.

McCallum, Sarah L. Taking Love Seriously: Amor and Erotic Elegy in Vergil’s ‘Italian Iliad’, University of Toronto, PhD dissertation, 2012. (Available online at .pdf).

Palmer, Leonard R. The Latin Language. Great Languages. London: Faber and Faber Limited, 1954.

Penney, John H. W. "Archaism and innovation in Latin poetic syntax", Aspects of the language of Latin poetry (J.N.Adams and R.G.Mayer (eds.)), Proceedings of the British Academy 93 (1999): pp. 249-268.

Quinn, Kenneth. Catullus - The poems. Rev ed. London: Bristol Classical Press, 1996.

Reed, James D. Virgil’s Gaze: Nation and Poetry in the Aeneid. Princeton, N.J. ; Woodstock: Princeton University Press, 2007.

Simon, Erika. "Greek myth in Etruscan culture", The Etruscan World (Jean M.Turfa (ed.)). London: Routledge, 2013: pp.495-512.

Skutsch, Otto. The Annals of Q. Ennius. Oxford : New York: Clarendon Press ; Oxford University Press, 1985.

Tarrant, Richard, J. Aeneid. Book XII. Cambridge Greek and Latin Classics. Cambridge: Cambridge University Press, 2012.

Thibodeau, Philip. Playing the Farmer: Representations of Rural Life in Vergil’s Georgics. Joan Palevsky Imprint in Classical Literature. Berkeley; London: University of California Press, 2011.

Weiss, Michael L. Outline of the Historical and Comparative Grammar of Latin (Ann Arbor: Beech Stave Press, 2009).

Wilkinson, L. P. "The language of Horace and Virgil, in Oxford Readings in Vergil’s Aeneid", Oxford Readings in Vergil’s Aeneid (Stephen J. Harrison (ed.)). Oxford; Oxford University Press (1990): pp.415-428.

Williams, Robert D. P. Vergili Maronis Aeneidos liber quintus. Oxford: Clarendon Press, 1960.