VERGIL'S LATIN

Author(s): W. F. Jackson Knight

Source: Acta Classica , 1958, Vol. 1, "ROMAN LIFE AND LETTERS": STUDIES PRESENTED
TO T. J. HAARHOFF PROFESSOR OF CLASSICS AT THE UNIVERSITY OF THE
WITWATERSRAND 1922 - 1957 (1958), pp. 31-44
Published by: Classical Association of South Africa

Stable URL: https://www.jstor.org/stable/24589992

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Classical Association of South Africa is collaborating with JSTOR to digitize, preserve and
extend access to Acta Classica

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 976 12:34:56 UTC
All use subject to https://about.jstor.org/terms
 VERGIL'S LATIN

by W. F. Jackson Knight (University of Exeter)

Paradoxically, if one word had to be found to indicate the

Vergil became a great, or even the greatest, lord of language
be 'compromise'. Compromise seems a dull, or even a me
minded people, and people of spirit, can be expected to
compromise. Yet compromise was Vergil's way to the uniqu
the tones of human grandeur's most exalted voice.
Vergiiian compromise is easier to accept as a principle
principles or tendencies of Vergil's mind are remembered,
alternation and reconciliation. Vergil liked to give close att
or one character or one interest, and then equally close atten
When such attentions are repeated in turn, they alternate.
claims are balanced, and some constructive equilibrium is fou
is progressive reconciliation. This mechanism is to be expec
but it is peculiarly characteristic in Vergil. He shows its op
scales, from the arrangement of letters, words and verses
of great tragic forces. What happens is much more than me
compromise is the start, and the method for part of the way
the way is Vergil's 'universality'.1
Vergil remained himself, and characteristically himself, i
aspect of his immense artistic activity. This sounds obvious
gotten, and if it is, mistakes follow. Nor is it a fact which i
in full, and understandingly. The number of considerations w
view and harmonized is so vast that our intellect and imagin
This characteristic operation of Vergil's mind seems to be
moral nature. He was shy, doubtful and meditative, and afr
confident. With humility, however, he combined the highe
fore he wished to be thorough, and not to act unfairly or allo
did to be below standard or shabby. These propensities mig
lysed his self-expression; but fortunately he had other adv
will-power strong even by Roman standards, and a force of
be imagined.
None of this is irrelevant to the complex and elusive ques
language. The finished mastery of his poetry is deceptive,
intricacies of its origins.
1 This article could be described as a few reflections periphera
Professor T. J. Haarhoff, F.I.A.L., The Stranger at the Gate, Oxf
the Universal, Oxford, 1948; to some extent developing conceptions
J. C. Smuts.
2 I have made very free use of J. Marouzeau, Traite de Stylistique Latine, Paris,
1946, to which I refer in the notes as M, and also A. Cordier, Stüdes sur le vocabulaire
ipique dans ΐ{.neide, Paris, 1939. To them I owe many references and some conclusions.
They should be consulted for the bibliography of the subject. Since independence was
impossible, and I needed to argue from material presented by them, I have gratefully
followed their presentation. Closely relevant to the present subject is Andrew J. Bell,
The Latin Dual and Poetic Diction, London and Toronto, 1923, a provocative but
instructive work with much concerning 'the figures'. The observations of L. A. S. Jermyn,
Greece and Rome XX, 1951, pp. 26—37, 49—59, seem to me indispensable.

31

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
 The basis of Vergil's Latin is Ciceronian practice. Indeed, his Latin is perhaps
fundamentally nearer to the Latin of Cicero's prose and verse than to any other
kind of Latin known to us3. Being rational and logical, Vergil preferred to start
with a Latin of such resource and lucidity that no outline of thought need be
obscured, and no rational relationships need be misrepresented. Latin has not
quite the precision of Greek; but Cicero's best prose comes near to the ideal.
Accordingly, Vergil on the whole accepted as a substratum Cicero's judgement
on the accidence of nouns and verbs, the syntax of moods and tenses, and to
some extent the choice of vocabulary; and Cicero could sometimes even furnish
him with a metaphor capable of poetic use. Lucretius, though his Latin has even
been thought not altogether unlike the Latin spoken at Rome in his own day,
and Catullus, with his inclination to rather reckless experiments, could never
have accepted so much of Cicero's Latinity.
Vergil compromised from the start. For one thing, he perpetually compro
mised between Greek and Latin. At any time a phase derived from or affected
by a passage in Cicero or any other Latin writer might be changed by the
influence of a Greek writer or Greek writers. That was continually happening,
and it has to be assumed. For example, when Vergil writes ruvipit uocem he
recalls Greek phrases such as φωνην ρηγννναι. Servius is therefore wrong,
or incomplete, in saying that what Vergil did was merely to invert a correct
expression such as rum-pit silentium 4. Again, sensit medios delapsus in hostis 5
is not less a Greek construction because, besides its occurrence in Catullus and
Horace, it is also old Latin; Plautus wrote dixit daturus 6. The Greek back
ground can never be safely forgotten, and all ancient and modern scholars who
compare Vergilian expressions exclusively with Latin antecedents run a serious
risk, a risk which must always be recognized even if it cannot always be avoided.
Apart from Greek, Vergil had plenty of other Latin to balance with Cicero's,
even if Cicero's was his substratum; and it is a serious probability that he tried
to read it all, and did read a great part. When he was writing the Eclogues, the
Greek of Theocritus and other Hellenistic poetry gave him most of the neces
sary material, but he was already making use of early and classical Greek poetry
including lyric and tragedies, and Latin poetry also, especially poetry of Catullus
and Gallus, and perhaps several other 'neoterics'. The Eclogues already show
Vergil hard at work according to his characteristic method. He was already
'compromising' between opposites, and especially combining and compressing
together derivations from different passages of earlier literature. One result is
the comparative frequency even in the Eclogues of phrases and sentences hard

3 Eduard Ftaenkel, Atti e Memorie della reale Accademia Virgilliana di Scienze,

Lettere ed Arti di Mantova, N.S. Vols. 19—20, 1926—7, pp. 217—227, stresses the
parallel and comparable importance of Vergil and Cicero — to Petrarch 'gli occhi de la
lingua nostra', and considers them to have been mutually sympathetic. He does not say
very much about their similarity of language, but his observations are constructive, not
least on Catalepton X as a highly skilful and characteristic exercise by Vergil in
developing and parodying Catullus IV. Some small agreements of conscience, if that is the
word, between Cicero and Vergil are worth noticing; for both equidem must mean
Ί indeed', not simply 'indeed', and 'igitur' must not stand first in a sentence. Cf. also
Karl Büchner, RE s.v. Vergilius.
4 Serv. A II 129.
5 All 377.
6 Plaut. Asin. 634.

32

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
to translate according to one exclusive sense or to any received Latin usage.
Vergil could perfectly easily have given us plain sense and a thin stream of
meaning. But he had a different conception of his art.
The Georgics required the use of Hesiod, perhaps Aristotle, certainly Theo
phrastus, Hellenistic didactic poets, Cato, Lucretius and countless others, in
cluding Cicero's Latin translations of Greek poetry and both poetry and prose
by Varro. All or nearly all affected Vergil's Latin. Among the compromises
which now strained his language was the very arduous adjustment of it to both
literary material of many ages and direct, familiar knowledge of the country
and the farm.

Material for the Aeneid came from literary works in Greek and Latin, in
poetry and prose, and numerous beyond all computation. Among them, in
particular, were Latin poems in the Roman epic tradition, in which the Aeneid
was to take a place. Vergil duly accepted their influence, but without neglecting
other influences. There are here good examples of his compromises, and of his
careful and sensitive judgement. He had to decide how much to accept, and how
to use the material provided by the tradition, such as rare and archaic words,
epic formulae and long compound adjectives. These questions have been
elaborately investigated, and lists and statistics published 7. Vergil learned from
Hellenistic innovations, but diverged. He accepted the Hellenistic practice of
hard thinking in the choice and vital use of words, but not to the extent of
abandoning epic tradition or even some epic formulae. He purified the epic
style which Ennius, Lucretius, Cicero in his poetry, and others offered for his
use. He even admitted expressions more appropriate to Latin drama, not only
tragedy but also comedy. But he remained traditional, and an epic poet. He did
not carry classicism to the degree of purification reached in Horace's Odes, and
perhaps later emulated by Lucan; he did not normally allow, as even Ovid was
inclined to allow, old, obscure or unexpected forms and expressions to creep
into the verse without strong reasons for their admission, and, as Quintilian
would have advised, he was carefully moderate in his archaisms. To reach his
destination in an extreme of power, Vergil travelled by countless middle ways.
Archaic and formal usages, familiar from old Latin epic, sometimes set
Vergil's Latin in obvious contrast with the Latin of Cicero's prose. Macrobius,
agreeing with Quintilian and perhaps going farther, was right to insist that old
Latin can very often explain Vergil's apparent innovations. Vergil no doubt
liked the traditional flavour, and agreed with Quintilian's belief that the sense of
antiquity adds a certain majesty8. But as always, Vergil's motives are subtle
and even elusive. His habit of compromise was a means, not an end. To achieve
it to his own satisfaction, he needed three other principles of art. His poetry
must be musical, and musically expressive. It must also be visual and tactile.
Vergil normally wrote concretely of solidly imagined people, things and actions,
avoiding the loose play of abstract ideas. Finally, but no less important, the
meanings must be compressed and condensed into a short space and few words.
Vergil's tendency to compromise, which might also be described as a willing
ness to consider everything and despise nothing, gave him great freedom to be

7 Cordier, passim, especially pp. 142—50, 285—301.

8 Quintilian I, 6, 39.

33

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
expressive. It led him to diverge from his predecessors in small matters and in
great. For the sake of musical sound he allowed many old genitive forms in
-um instead of -orum, such as deum, uirum and magnanimum. When, at the
death of Turnus, -orum stands at the end of both the first and second halves of
a single verse, it is unique 9. Vergil nowhere else allows the syllable to occur in
exactly these places10. Ovid was characteristically more tolerant, and used such
impressive sound-effects more freely and less thoughtfully, so making them less
impressive. Vergil wanted his poetry to present and emphasize the concrete;
accordingly, he used far fewer verbally formed nouns ending in -men -minis,
-tura and -tus than Lucretius. The nouns ending in -tus were favoured by Livy,
who seems to have learnt much from Vergil, and afterwards affected Vergil in
his turn. The nouns in -men -minis were of course a great help to versification,
as Ovid found; he used them in immense numbers and variety. Interestingly
enough, Vergil was more ready to use them if they had, or could have, a concrete
meaning, as gestamen and tegimen or tegmen. Like other Latin poets he
developed the use of nouns, especially, but not only, eloquent short nouns which
may almost personify qualities and actions, to express what prose might express
by verbs11. Another revealing practice is Vergil's care to substitute other verbs,
with richer and more vital visual content, for est and sunt, as iacet, ibat, incedo,
and most of all stat, stant12. Vergil did not invent the practice, but he extended
it. As for compression of meaning into a short space, examples are everywhere.
Most often, when Vergil writes unusual Latin, this desire for compression is
the reason, or one of the reasons. He was always trying to compress; 'ut uerba
in compendium cogeref, as Servius well described itt3. Vergil used the word
pone, 'behind'. It was not contemporary literary Latin, but old, and possibly in
Vergil's time a rustic survival14. But it was very much shorter than a tergo.
Still shorter, for its wealth of meaning, is ilicet. Vergil uses it with enormous
tragic effect to mean something like 'all was lost', or 'no hope remained'1S.
But it was not fashionable. It belongs to old Latin, especially the Latin of the
law courts and comedy16. In Terence it means something like 'the court rises';
all can go home. What is astonishing is that Vergil should have drawn from

9 A XII 648—9:
sancta ad uos anima atque istius inscia culpae
descendam magnorum haud umquam indignus auorum.
Similarly unique is A XII 903—4:
sed neque currentem nec se cognoscit euntem
tollentemue manus saxumue immane mouentem.
10 Close, but different, are A III 549, A VII 18, A XI 361.
11 Bell, pp. 155—9; Vergil should be carefully compared with other writers in his
exploitation of cor, fatum. fides, horror, lis, mens, moi,, numen, spes, and other such
nouns; a new use of nouns was one of the many lessons learnt by Tacitus from Vergil
('in some sense Vergil touched off the Silver Age' — J. R. T. Pollard).
12 Μ p. 146.
13 Serv. A I 639.
14 Μ p. 195, citing Stolz-Schmalz-Hofmann, Lateinische Grammatik, Munich, 1928,
p. 500.
15 A II 424.
16 Serv. A II 424: ilicet — confestim, mox. sane apud ueteres significabat sine dubio
'actum est'; Terentius, Adelphi: resciunt omnes rem; id nunc clamant, 'ilicet'; Eunuchus:
actum est, ilicet, peristi.

34

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
such a word such overpowering tragic and poetic force. He also used magis atque
magis, a despised expression, with high poetic success. He even went so far as
to adapt the very informal ecce tibi, writing en perfecta tibi, at a moment of
high intensity17; but, as usual, the artistic gain is great.
Vergil's humility and readiness to compromise led him to such daring and
such achievement. Indeed, it sometimes looks as if his unwillingness to call
anything common or unclean caused him to become not only capricious and
unconventional, but positively provocative, if not impertinent. However, even
so his poetry does not fail, and the price paid is nothing compared to the profit
in compression and explosive poetic power.
There are some examples of Vergil's receptive catholicity which, though they
show no great distortion of normal Latin, are sufficiently instructive to be
specially mentioned. Writing of a thoroughbred horse, Vergil says that he
'replaces soft legs', or 'feet', 'on the ground', or 'replaces his legs' or 'feet',
'softly', the proleptic adjective being beautifully expressive, mollia crura
reponit18. The meaning of mollis in this context has of course been discussed.
It is more important to look at the adjective in another context. For Ennius uses
the same words, mollia crura reponunt, but of cranes, grues: perque fabam
repunt et mollia crura reponunt19. The birds, walking about and picking up
beans, are comic; the thoroughbred is not, but exquisitely beautiful in form and
action. What exactly Vergil has done, why his daring choice of a derivation has
enabled him to express the very nearly inexpressible, and how he has altered
Latin to do so, would need a long enquiry. But Vergil's daring gave him, as
usual, success. No one can really fail to see the picture presented.
In such examples Vergil sometimes makes use of meanings and suggestions,
appropriate in an older context but less obviously appropriate in his own, new,
context, to make his own intended meaning more exactly clear. Again there is
paradox. Vergil writes of Juno, 'an incessant affliction for the Trojans', as
Teucris addita Juno 20. The verb is unexpected. Perhaps it suggests Socrates,
'applied', like a gadfly, to the Athenian democracy. But it is known where
Vergil found it. It was in Lucilius, who wrote si mihi non praetor siet additus
atque agitet me, non male sit21, 'If, with all the rest to put up with, I had not
the Praetor, too, on my hands, tormenting me, it would not be so bad.' This is
not the only rather surprising Lucilian echo in Vergil. The very inappropriate
ness of the old associations is made to add force and precision to Vergil's poetry.
Again, conventional Latin is very lightly altered.
By being reconditely allusive Vergil made his poetry defiantly direct. He was
praised in antiquity for excelling in all the styles of rhetoric, even though he
wrote poetry, not rhetorical prose. It does not follow that he always obeyed the
expressed and authoritative rules for good writing.
Quintilian has a rule against mixtura uerborum, which roughly means 'too
many words out of the natural place in a sentence', and he accuses Vergil of
disregarding it when he wrote saxa uocant Itali mediis quae in fluctibus aras
17 A VII 545.
18 G III 76.
19 Enn. Ann. 556 Vahlen 3.
20 A VI 90.
21 Lucilius XIV 469—70 Marx; Macrobius Saturn. VI, 4, 2.
22 Quintilian VIII, 2, 14; A I 109—10; Μ p. 322.

35

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
This is not very terrible, especially if it is the only bad case which Quintilian
could find. There is here, however, a notable case of Vergil's care in choosing
the middle way. Quintilian's mixtura is a fault which often beset many good
Roman poets, though usually in tolerable measure; but it is not unlike that
'interlacement of word order' which is praised as a practice valuable to Vergil
for the enrichment of meaning, as of course it is 23. He could, however, dispense
brilliantly with the practice of interlacement as he could, apparently, with every
other device; and so achieve the sweet, heavenly power of his verses beginning
deuenere locos laetos ..where Aeneas at last reaches the homes of bliss 24.

There was an amusing difference of opinion concerning the repetition of

words25. Quintilian noted that the best writers did not greatly worry to avoid
repetitions, though some people were quite childish in their efforts to find
synonyms26. The Auctor ad Herennium had already observed that synonyms
were regularly used by writers to avoid repetitions, but he himself liked repe
titions and classified their uses27. Cicero to some extent concurred 28. As usual,
Vergil himself partially agreed with both sides. His repetitions are sometimes
exquisitely artistic according to many classifiable schemes, as uidemus Italiam.
Italiam primus conclamat Achates, Italiam...29 in three verses, ab loue...
loue... Iouis 30 in two verses and node... nocte ... nocte... noctesZi in four.
There can be great poetic might in a repetition, as inde domum, si forte pedem,
si forte tulisset32. But Vergil often admits casual repetitions in great frequency
23 J. W. Mackail, The Ae?ieid of Vergil, Oxford, 1930, Intro, p. lxxxiv; 'It is in the
manipulation of language that Vergil stands apart from other poets. His sensitiveness to
language is unique, more especially the way in which he perpetually — it might almost be
said, in every line of the Aeneid — gives words and phrases a new colour by variation,
sometimes obvious, sometimes so delicate as to escape notice, of the normal or classical
diction. Language always remains with him a fluid medium, and he handles words so as
to make them different. The interlacement of word order, to which a highly inflected
language like Latin lends itself, is carried by him to the utmost limit, and the phrase,
within which no division by punctuation is possible, may extend over several lines. Words
which are logically or syntactically inseparable may be at long distances from one another,
and his cross-patterns of language, while they seldom fail to convey the effective meaning
desired, almost defy analysis. It was this that led his detractors to say that he did not write
Latin; and there is this much of truth in the criticism, that his Latin is a language of his
own, and one in which he was, up to the last, perpetually experimenting.'
Cf. however Fraenkel p. 226 and especially notes 1 and 2 for Vergil's freedom from
such awkward arrangements as Catullus LXVI 18 non, ita me diui, uera gemunt, luuerint;
Vergil shows a slight tendency to such writing in the Eclogues, but scarcely any after
them. A thorough inspection of hyperbaton throughout Vergil might reveal secrets
concerning his mind and thought.
24 A VI 637—8:
deuenere locos laetos et amoena uirecta
fortunatorum nemorum sedesque beatas.
cf. Μ p. 335. 'C'est lä un des exemples les mieux faits pour montrer le role que peuvent
jouer les bons ecrivains dans la defense de la langue, en reagissant contre ceux qui par
l'abus des precedes de style aboutissent a la negation meme du style.'
25 Μ pp. 267—276.
26 Quintilian VIII, 3, 51; X, 1, 7.
27 Auct. ad Heren». IV, 42, 54.
28 Cicero, De Or. 3, 206; Orator 135.
29 A III 522—4.
30 A VII 219—20.
31 G I 287—90.
32 AII 756.

36

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
and, to us, rather offensively; as solemus... solebam... solent33 in five verses.
Meanwhile, his care to find synonyms is often very thorough and successful, as
when in eleven verses ten words mean water in some form, and eight of them
are different34.
Vergil freely used parataxis when constructing clauses, though a simple
sequence of statements was accounted sermo inliberalis. Vergil could write
elaborate paragraphs. But on the whole he inclined to a narrative structure
nearer to Caesar's habit than to Cicero's or Livy's, with many connectives
meaning 'and'. Poets, of course, are not expected to write in prose periods,
though Lucretius and Catullus both exhibit long and elaborate paragraphs. But
there is a special interest here. Vergil tended to return to old Latin structure,
even to the use of present participles in the nominative to end sentences, and in
word-order which approached modern English. It is indeed probable that, by
organizing sentences in this way, Vergil altered the course of the literary future,
and even helped to decide the shape of mediaeval Latin and the languages of
modern Europe. There is here another example of Vergil's attention to both of
two extremes. This time he was inclined to favour both, in different passages.
There was apparently a rule discouraging the excessive use of pronominal
words, and words lacking a solid content of meaning35. Latin is rich in such
words, and perhaps especially forms and derivatives of the relative pronoun.
They are useful for emphasizing, with brevity, the rational organization of a
sentence with an almost mathematical logic, and that is something which Latin
likes to do. The free use of such words belongs apparently to old Latin, and
especially old legal Latin. It was not considered good in literary Latin or artistic
oratory in the classical period. Vergil has not much of this fault, but he did not
always try to avoid it. There is the famous quae cuique est forluna hodie, quam
quisque secat spem ...36 As usual, that which, occurring elsewhere or used by
another, might have been a fault, is turned to high poetic effect. There may be
no great merit in nos tarnen haec quicumque modo tibi nostra uicissim...37
But Vergil's practice represents a restriction of the practice of Ennius, as quic
quam quisquam {quemquam), quem que, quisque conueniat, neget?,38 a verse
admittedly from a drama, not epic; and even when he meant to echo Ennius he
never went as far as Ennius would freely go.
When Vergil wrote strangely, he apparently obeyed his own judgement, for
all his diffidence, without much caring for pedantic critics. He used the
adjective cuius -a -um\ die age, Damoeta, cuium pecus? an Meliboei? 39 A reply
came in the form of a parody: cuium pecus? anne Latinum? non, uerum
Aegonis: nostri sic rure loquuntur 40. It is a foolish complaint. The adjective
cuius was good old Latin, had probably survived outside Rome, and was indeed
used by Cicero. Vergil used hordea in the plural, otherwise not known 41. The

33 EI 20—25.
34 G IV 359—367.
35 Μ pp. 109—115 citing Ε. Η. Sturtevant on 'grammatical machinery'.
36 AX 107.
37 Ε V 50.
38 Enn. jr. tr. 422 Vahlen 3.
39 Ε III 1.
<0 Donat. Vit. Verg.
« Gl 210.

37

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
answer offered, 'by Bavius and Maevius' as Servius pleasantly says, was bordea
si dicit, cur non et tritica dicit? 42 It seems rather childish to us. Vergil may well
have had Latin authority for hordea. But he was probably more or less thinking
in Greek, as he often did, and using the plural as Greek used it for more than
one type of grain, as τα αλφιτα.
Poetry, in Vergil's day, had advanced farther than poetic criticism. There
certainly seems to have been much rather pedestrian appreciation going on in
antiquity, and in later antiquity. Macrobius, however, has passages which, from
our point of view, are appreciatively in the right direction, even if some secrets
of classical poetry had been lost before the later writers lived 43.
A list, partly taken from Servius, which Macrobius gives shows the kind of
Vergilian phrase which seemed to other Romans unusual, but also well, bene,
pulcbre, contrived. Examples are: recentem caede locum 44, 'noue dictum'; caeso
sparsurus sanguine jlammas 45, that is, 'qui ex caesis uidelicet projunditur ;
corpora tela modo atque oculis uigilantibus exit 48, that is, 'tela uitat'; exesaeque
arboris antro 47, 'antro' for 'cauernd; frontem obscoenam rugis arat48, 'drat
non nimie sed pulcbre dictum'; and uir gregis 69 for 'caper'. Macrobius exclaims:
'et ilia quam pulcbra sunt: 'aquae mons' 50, 'telorum seges' 51, 'ferreus imber'°'z
ut apud Homerum λάινον εααο χιτώνα. He likes dona laboratae Cereris ~'3,
oculisue aut pectore noctem accipit54, uocisque offensa resultat imago 5a,
pacemque per aras5B, and paulatim abolere Sychaeum incipit57. On the verse
discolor unde ami per ramos aura refulsit58 he writes: 'quid est enim aura auri,
aut quem ad modum aura refulget? sed tamen pulcbre usurpauit'. Of simili
frondescit uirga metallo59 he says 'quam bene usus est 'frondescit metallo'
On baud aliter iustae quibus est Mezentius irae 60, he writes: 'odio esse aliquem
usitatum, irae esse inuentum Moronis est'. He notes that luturnam suasi 61 is
unusual for luturnae suasi.
Macrobius' list and comments may seem simple and obvious. But they show
how Vergil's language appeared to intelligent readers in antiquity, and help to
indicate what seemed, or indeed were, at least some of the differences between

42 Ser. G I 210: hordea — usurpatiue. Bauius et Maeuius: hordea si dicit, cur non et
triticadicit?
43 Macrobius, Saturn. VI, 6, 1—9.
44 AIX 455—6.
45 A XI 82.
46 A V 438.
47 G IV 44.
48 A VII 417.
49 Ε VII 7.
50 AI 105.
51 Α III 46.
52 A XII 284.
53 A VIII 181.
54 A IV 530—1.
55 G IV 50.
56 A IV 56.
57 A I 720.
58 A VI 204.
59 AVI 144.
60 AX 714.
61 A XII 813—4.

38

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
the Latin of Vergil and the Latin of others. The list shows Vergil never letting
anything alone without hard thought and at least some act of originating will.
Occasionally there is perhaps an appearance, deceptive or not, of change for the
sake of change and no more. More often in these and other such instances there
is compression of meaning, a clear enrichment of content, and an access of
emotion, drama, music, or colour, due to an unexpected word or usage.
The practice is something like a great extension of the normal figure of
abusio, κατάχρησις, that is, the sudden use of an unexpected word C2. Caesar
advised writers to avoid an abnormal or unusual word, inaudttum et insolens
uerbum, like a dangerous rock at sea 63. Fronto was not content with usual and
ordinary words, solitis et usitatis uerbis 64. He advised out-of-the-way words,
uerba non obuia C5, and disagreed with Cicero, who thought that in the best
orators very few unexpected and improbable words, insperata atque inopinata
uerba, could be found 66. These writers were thinking mainly of vocabulary and
words in themselves unfamiliar, the true abusio, rather than words unexpected
in their context. But the interest is not the less for that. There was clearly, and
had long been, a useful difference of opinion. One of the points at issue was
whether we should write smoothly, or surprisingly, administering shocks.
Characteristically, Vergil took everything into consideration, and might almost
be said to have devised a way to secure the advantages offered by both the
opposing doctrines. It is as if he had set before himself the ideal of compression,
and then, subject to that ideal, had worked out his compromise of usual words,
unusually manipulated.
This policy led to a great number of small divergences from Ciceronian
Latin, many of them indicated by Servius, and many, too, revivals of οία Latin
practice. When Vergil uses a mood or tense which would not be expected in
Ciceronian prose, for example an indicative in indirect question, as ne quaere
doceri... quae forma uiros jortunaque mersit H7, or uiden ut geminae stant
uertice cristae? es or a very sharply significant perfect tense, as tum res rapuisse
licebit69, there are usually antecedents in old Latin. Other variations, as quem
dat Sidonia Dido 70 (for dedisset), and 'mixed conditional clauses', are no doubt
Vergilian, but most of them are near to the general tradition of Latin expression,
and some are Vergilian in the sense that Vergil exploited constructions which,
like other parts of the Latin language, had already achieved a compression of
complex meaning. This, as a general tendency of Latin, can be seen in other
matters, including a number of single words; the adjective lentus is almost a
poem in itself 71.
A few of Vergil's genders were noticed in antiquity as irregular. He was
apparently alone in making damma, usually meaning 'doe', masculine; this was

62 Μ ρ. 197—9.
63 Gell. I, 10, 4.
64 Fronto p. 50 Naber.
65 Fronto p. 98 Naber.
66 Fronto p. 63 Naber.
67 A VI 614—5.
68 A VI 779.
69 Α X 14.
70 A IX 266.
71 Bell p. 14 and passim.

39

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
apparently because he wanted the sounds of timidi dammae, without a rhyme.
The genitives of Greek names, as Achillei or Achilli, were said to be inventions
of Vergil, created to give the right sound. But his third-declension datives in -e,
fourth-declension datives in -u, and perhaps a fifth-declension genitive in -e
are from old Latin; so, directly or indirectly, are the syncopations direxti,
repostas. None are very startling, except perhaps aspris for asperis. So probably
are some usages of the gerund which might seem strange. Servius, explaining
ardescitque tuendo 12, calls the gerund 'passive', as also in frigidus in siluis
cantando rumpitur an guts73; this is in contrast to cantando tu ilium (sc.
superabis) ? 74, which he calls 'active'. Of course, cantando rumpitur might be a
retention from some expression cantando rumpas or rump ere potes after a
Vergilian adaptation. The best parallel from another writer is from the Bellum
Hispaniense, erumpendo naues incendunt75; some such source may have in
fluenced Vergil. No principle seems to be infringed by uoluenda dies16, and inter
agendum77 is correct; both may perhaps be called old-fashioned. Vergil's
extended use of cases, especially the dative and ablative, and his omission of
prepositions, are particularly characteristic and important, but no more can be
said about them here.

Miscellaneous details are plentiful. Old-fashioned are Vergil's active verbs

which might normally have been deponent, such as populat for populatur, and
perhaps some intransitives which might have been transitive, as siliqua
quassante, uoluentia plaustra, and uoluentibus annis; other unusual forms are
recens as an adverb; multa, plural, adverbial, for multum; ambo and duo as
accusatives; adeo meaning something like 'indeed', and atque involving some
such sense as 'suddenly'; proprius meaning 'appropriate' or 'correct' rather than
'his own'; put ο in an old Latin sense, 'ponder'; and such old forms as stridunt
for strident and steterunt for steterunt.
Vergil preferred to treat Latin as timeless, and as still fluid, that is, still more
fluid than it actually was in his own time. He was not only using language, but
also creating language, as a poet should. He was perhaps capricious and self
willed; but to a poet quidlibet audendi semper fuit aequa potestas. The main
thing was to have the greatest possible volume of material to use, and the most
extensive resource. Vergil used old Latin as he used contemporary Latin. There
were two words for 'son', natus, poetical, and filius, prosaic. He used both; but
each only in contexts which were appropriate. He similarly used the more
dignified proelium and the less dignified pugna both appropriately, and not far
apart7S. But he used the prosaic fluuius very freely, and did not clearly
distinguish it from the more poetic synonyms; a word may be useful for itself,
apart from any associations. At, or near, the opposite extreme, are such phrases

72 A I 713.
73 Ε VIII 71.
74 Ε III 24.
75 ße//. Hisp. 36.2.
76 a IX 7.
77 ΕIX 25.
78 Μ ρ. 167.

40

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
as Anchisa generate, which seem to be principally valued for their Ennian
quality. They are occasionally used in considerable numbers, and contribute
importantly to some elaborate Vergilian passages79.
Ellipse is part of the reason for many Vergilian irregularities. Sometimes it
may count as the whole reason. On occumbere morti Servius says nouae
locutionis forma et penitus remota, rather strong language, and quotes Ennius,

79 Μ p. 193; cf. also L. R. Palmer, The Latin Language, London, 1954, pp. 112—3,
on A X 100—110:
tum pater omnipotens, rerum cui prima potestas,
infit (eo dicente deum domus alta silescit
et tremefacta solo tellus, silet arduus aether,
turn Zephyri posuere, premit placida aequora pontus):
'accipite ergo animis atque haec mea figite dicta.
quandoquidem Ausonios coniungi foedere Teucris
haud licitum, nec uestra capit discordia finem,
quae cuique est fortuna hodie, quam quisque secat spem,
Tros Rutulusne fuat, nullo discrimine habebo,
seu fatis Italum castra obsidione tenentur
siue errore malo Troiae monitisque sinistris ..
Professor Palmer writes: 'Vergil's archaisms are used with delicate and deliberate
artistry. As with Lucretius, they are dictated by the theme. It is noteworthy, for instance,
that the form fuat occurs in Vergil only in a speech of Jupiter (I.e.) a passage worth
examining in the present context. The words of the Pater Omnipotens are introduced by
the archaism infit. The scene is sketched with alliteration of Ennian intensity:
eo dicente deum domus alta silescit.
The speech itself opens with an impressive "dicolon abundans":
accipite ergo animis atque haec mea figite dicta.
His judgement, which begins with the majestic polysyllable quandoquidem, "in as much
as" (never used by Cicero in his speeches, nor by Caesar), has the balanced binary struc
ture rooted in the language of religion and law. In the last line we sense the dolo malo
of the leges sacrae and the sinister of the language of augury. Thus the archaism fuat
finds its setting in a majestic context where the father of gods and men sits in the
judgement seat.
'Marouzeau has pointed out a number of instances where such archaisms colour the
language spoken by the gods: quianam (Ennian) is used by Jupiter (A X 6), moerorum
by Venus (A X 24), ast by Juno (A I 46). No better illustration could be found of
Quintilian's dictum, "uerba a uetustate repetita ... adferunt orationi maiestatem aliquam"
(1,6,39).
'The Sibyl, too, speaks a language not of this world:
olli sic breuiter fata est longaeua sacerdos;
"Anchisa generate, deum certissima proles..."
The whole passage (A VI 317—336) describing Aeneas' arrival at the Styx is particularly
rich in archaic colouring: enim 'indeed', the assonance inops inhumataque, the anastrophe
haec litora circum, the archaic significance of putans, the locative an'tmi, and finally the
phrase ductorem classis, where an antique gem in a modern setting of glossae forms the
splendid line:
Leucaspim et Lyciae ductorem classis Oronten (334).
In this passage we may note, further, the Ennian reminiscence uada uerrunt and uestigia
pressit; the patronymic expressions Anchisa generate, Anchisa satus, which were a feature
of Latin epic style from Livius Andronicus on; the syntactical Graecism (this a "gloss")
iurare numen; and finally the un-Latin -que . . . -que, which is a 'caique' coined by
Ennius as a convenient hexameter ending on the lines of Homeric expressions such as
ολίγον τt φίλον τε, πόλεμοι τ ε μάχαι τε.'
I hope so long a quotation will be forgiven. Professor Palmer's comprehensive and
rounded commentary is particularly valuable here since my treatment in the text is
selective and fragmentary.

41

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
morti occumbant obmam80. It is hard for us to realize that the Romans felt that
the loss of obuiam, though there is still an ob- in the verb, made so much
difference. Even in this phrase, apparently, Vergil was quite audacious in his
continuous and insistent search for brevity and compression.
There are three adjectives which, differentiated in classical Latin, had all
according to Servius meant 'large' or something like it in old Latin. They are
dims, indignus and saeuus 81. Servius may well be inexact. But if the words did
once have wider meanings, it may be that Vergil took advantage of a certain
ambiguity when he wrote indigno cum Gallus amore ■peribat82, or when he
applied saeuus to Juno 8S. The adjective dirus is more interesting. It is thought
to have come to Rome from Umbrian and Sabine neighbours of the Romans,
so that it may not have been fully understood 84. Vergil noticeably used it :o
mean neither 'terrible' nor 'large, but something nearer to 'excessive' or
'exaggerated', tending to 'uncanny', 'grotesque', 'unnatural', as in dir a cupido 83,
an 'excessive', perhaps even 'impertinent', desire8S. It is possible that in these
three instances Vergil used antiquity and obscurity to increase his freedom, just
as he used simple verbs in preference to compounds, for instance the older form
tendere, which has necessarily more possible meanings than contendere, so that
his art of ambiguity, light allusion, and penumbral meanings could be
helped. But Vergil was equally capable of finding contendere in an old passage
and adapting it as tendere. It is hard to keep pace with the constructive agility
of his mind.
Servius does not apparently quote Sallust when he comments on Vergil's
gerundives such as ardescitque tuendo\ but elsewhere he cites more parallels
from Sallust than from any other writer. It is unnecessary to suggest that he
does so because Sallust was in particular use as a school text. More probably
Vergil admired Sallust's artistic treatment of the Latin language, and joined him
in enterprise and experiment. Cicero complained of historians whose only merit
was their brevity; Vergil thought that there was much to be said for that quality,
and certainly sympathized with Sallust in his liking even for Cato's Latin.
Servius, or Daniel's Servius, quotes Sallust for requierunt, intransitive87,
aeuoque sequenti 88, the meaning of cultus in qui cult us habendo .. .89 (two
quotations from Sallust), the fifth-declension genitive in -e (die) 90, the use of
fcrent for essent91, falsus meaning 'deceptive'92, a possible form uinus {οίνος)

80 Serv. A II 62; Enn. Sc. 135—6 Vahlen 3; but cf. also Enn. Ann. 398 Vahlen 3
occumbunt multi letum ferroque lapique; perhaps even in this small matter Vergil struck
a balance between two Ennian phrases. In general ancient criticism tends to overlook the
double parenthood accountable for many, or most, Vergilian inventions.
81 Serv. ad G I 37, Ε VIII 29, A I 4 and A II 226.
82 Ε X 10.
83 AI 4 et al.
84 Serv. Α III 235.
85 A VI 373; cf. G I 37 and A IX 185.
86 I owe this interpretation to the kindness and insight of Mr. John D. Christie.
87 Serv. Ε VIII4.
88 Ε VIII 27.
89 G I 3.
90 G I 208.
91 G I 260.
92 G I 463.

42

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
 instead of uinum93, the meaning 'untruthful' for nanus u, certos meaning
'reliable men' 95 (two quotations from Sallust), ficlens animi 9li, foedauit junere
u ο It us 97, and uos agitate fugam 98.
This is a seleaion from notes in the earlier part of Servius' commentary; a
full list would be long. There cannot be serious doubt that Vergil, from the
Eclogues onwards, worked in sympathy with Sallust's experiments, and learnt
from him procedures which were important in determining his own treatment
of language throughout his lifetime, especially his choice of words, and of
contexts for them. Sallust helped to teach Vergil the extension of the figure of
abusio which was to become characteristic of all his work.
Agrippa's famous criticism of Vergil's Latin90 is of course hard on Vergil,
but it seems to have been an honest reaction to carefully observed facts; too
carefully observed, perhaps, for passivity might have allowed receptive sym
pathy to grow. Others too have been misled, including Byron and Shelley.
However, Agrippa helps to show that Vergil's Latin seemed very peculiar to at
least some contemporaries, and to confirm our own conception of what Vergil's
peculiarities were.
By far the greater number of Vergil's first hearers and readers accepted him
with whole-hearted acclaim. Latin was still, even then, experimental. There was
scarcely yet a standard Latin, as there has been a standard English for two
hundred and fifty years. But Vergil's contemporaries, if well read, must have
noticed echoes of many periods and many writers, almost continuously. The less
well-read probably found his Latin about as remote and strange as we find the
Authorised Version of the Bible, or a little less so; and perhaps sometimes it
seemed to them about as artificial and elaborate as Gerard Manley Hopkins'
poetry seems to us.
But all alike, and at all times, could and should have been captivated by the
Vergilian music; and most have been. The music depends on the Latin, and the
Latin would not have all its music if Vergil's artistic mind had not been so
capacious.
After all, to judge of Vergil's Latin it is not so necessary to read Servius and
other commentators as to read Vergil himself, and other Latin poets for com
parison. If we do so, and look out not for oddities but for the ordinary, regular,
personal quality of the text before us, we notice something in Vergil which sets
him apart from the rest, and which can be called part of his universality. The
other poets are always themselves, and perhaps even a little self-conscious some
times. Each has a style, and sometimes it seems an exaggeratedly individual style.

93 G II 97; Servius has apparently misunderstood Vergil, but that is irrelevant here.
94 A I 392.
95 A I 576.
96 AII 61.
97 AII 286.
98 AII 640.
99 Donat. Vit. Verg'M. Vipsanius a Maecenate eum suppositum appellabat nouae
κακοζηλίας repertorem, non tumidae neque exilis sed ex communibus uerbis atque ideo
latentibus.' There is very little at all offensive in this famous notice except the words
suppositum and κακοζηλίας. There is even a pleasing suggestion of the middle way in
nec tumidae neque exilis, and the remaining words might almost be taken as an acknow
ledgment of Vergil's success.

43

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms
Vergil, of course, leaves fingerprints everywhere. But they are less important.
What is important is the catholicity, the intellectual capacity, the patience which
enabled him to amass all the resources which his vision required, and the
economy and variety with which he used them. Being so greatly gifted, he
could, with a small vocabulary and under fairly rigid rules, do something diffe
rent every moment. Variety of expression is one of his great secrets of success.
He had the full power of rare genius to break the rules and 'snatch a grace
beyond the reach of art'. Sheer resource, and abundance of material at his
immediate command, gave him immense flexibility, even though he often seems
formal. He needed all of it, because in his art not only syllables but even single
letters counted; not only vowels but also consonants; not only quantities but
also accents; and so on, beyond the limits of our knowledge. Pauses must not be
too regular. They can, however, be regular in the right place for a short passage
which is poetically, dramatically and emotionally such as to require just such
pauses. Dactyls, spondees and elisions must match the iridescent flash of the
many-sided systems of thought, which they must follow, and in their turn
engender. Comedy and tragedy must co-operate to create truth within a two
word phrase. Elemental forces must conflict and crash about us in the story, and
the whole life of worlds must seem to us tangibly at stake. Yet the artist himself
must keep something in reserve and never lose control, but coldly check
himself against excess in strengthening or protracting or repeating any single
note. It is, oddly enough, the end of the way of which compromise was
the beginning. But it was always a compromise composed of fairness and
honesty, a sympathy for finding a value in all things, a courageous will to pay
any price, however high, and unresting effort, never withheld, until everything
needed, everything that could conceivably be needed, and needed for a task
never imagined before, had been acquired. Tantae molts erat.

44

This content downloaded from

194.214.29.29 on Sun, 04 Apr 2021 20:47:51 UTC
All use subject to https://about.jstor.org/terms