Repetition in Lucretius

Author(s): Wayne B. Ingalls

Source: Phoenix, Vol. 25, No. 3 (Autumn, 1971), pp. 227-236
Published by: Classical Association of Canada
Stable URL: http://www.jstor.org/stable/1087359 .
Accessed: 04/01/2015 06:35
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
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 support@jstor.org.

Classical Association of Canada is collaborating with JSTOR to digitize, preserve and extend access to
Phoenix.

http://www.jstor.org

This content downloaded from 143.107.252.19 on Sun, 4 Jan 2015 06:35:49 AM

All use subject to JSTOR Terms and Conditions

REPETITION IN LUCRETIUS
WAYNE B. INGALLS

LUCRETIUS REPEATS HIMSELF more than do otherLatin poets. The

fact is well known and has been dealt with in a numberof recentdiscussions. Rosamund Deutsch devotes much of her dissertation, The
PatternofSound in Lucretius,to the various typesof repetitionthe poet
employsfromsimple alliterationto reiteratedlines and half-lines.,First
of all she points out (85) the interestingway in which phrases are repeated in close proximity,such as
at coria et carnemtrahitet conducitin unum.
umoraquae porroferrumconduratab igni,
at coria et carnemmollitdurata calore.

[6.967-969]

She also observesthat several versesmay intervene,such as

et quibus ille modis congressusmateriai
et quibus ille modis divommetusinsinuarit

[5.67]
[5.73]

She furthernotes (89) that Lucretiusoftenvaries his repetitionsto avoid

monotony,as in
redditacorporibusprimis per inane profundum
paucula quae porro magnumper inane vagantur
multaquepraetereamagnumper inane vagantur
multa minuta modis multisper inane videbis

[2.96]
[2.105]
[2.109]
[2.116]

Finally Dr Deutsch shows (98-100) how the poet, when discussing a

particular topic, frequentlychooses a closely connected theme word
whichis repeatedseveral timesin the courseof the discussion.In passing
we might add to her observationthat these theme words are oftenthe
componentsof what might be termed "theme phrases," as the poet's
treatmentof colour in Book 2 will illustrate.Of the one hundred and
twelve verses devoted to the topic, twenty-twoend with the word color
in the plural or in one of its oblique cases, that is, fillingthe finalthree
syllablesof the line.Two reiteratedphrasesaccount forsix of the twentytwo verse-endings.They are tinctacolore(736, 747, 776) and variantque
colores (759)/variosquecolores (783)/variovecolore (825).
Anotherdiscussionof repetitionin Lucretiusis foundin CyrilBailey's
Prolegomenato his edition of the De RerumNatura.2He largelyfollows
Deutsch but adds a fewmoretypesof repetition.He observes (163) that
Lucretius oftenrepeats lines which are in effectthe axioms of his theory
'Rosamund E. Deutsch, The Patternof Sound in Lucretius (Bryn Mawr 1939).
2Titi Lucreti Cari De Rerum Natura ed. Cyril Bailey (Oxford 1947).

227
PHOENIX,

Vol. 25 (1971) 3.

This content downloaded from 143.107.252.19 on Sun, 4 Jan 2015 06:35:49 AM

All use subject to JSTOR Terms and Conditions

228

PHOENIX

when he wishes to referto them. Of this practice numerousexamples

may be found,such as
namquodcumque
suis mutatumfinibus
exit,
hocmorsestilliusquodfuitante.
continuo
[1.670-671= 1.792-793= 2.753-754= 3.519-520]

and
huncigiturterrorem
animitenebrasque
necessest
nonradiisolisnequelucidateladiei
sed naturaespeciesratioque.
discutiant,
[1.146-148= 2.59-61 = 3.91-93 = 6.39-41]

He furtherpoints out (145) that, like the regularsuccessionof particles

in the sectionsof argumentation,repeated phrases frequentlyintroduce
a new topic. For example, the words
Quodquoniamdocui,pergamconectere
remquae
ex hocaptafdemducat

[2.478-479]

introducethe proofthat the numberof atomic shapes is finite.Further

on the poet uses the same words when he begins the argumentproving
that the numberof atoms of each shape is infinite(2.522-523). He also
suggests (162) that Lucretius sometimesrepeated abnormal idiomatic
expressionsto establish their usage. He adduces the example of tanto
quique magis (meaning "by so much the more") which is firstfoundat
3.700 and is repeated at 5.343. It is difficult,
however,to see how the
of
this
two
idiom
thousand lines later
repetition
approximately
single
serves to establishits usage. Nor can I envisage the poet takingpains to
establish the usage of idioms which might troublehis futureeditors.
W. S. Maguinness has also touchedupon Lucretius' repetition,though
he, like Dr Deutsch, deals primarilywith the repetitionof singlewords.3
He thinksthat the poet's quest forclarityled him to repeat words and
phrases in a manneravoided by otherLatin writers.
view of repeAlthougheach of these scholarstakes a slightlydifferent
tition in Lucretius, theirconclusionsare fundamentallythe same. The
followingquotation fromDr Deutsch (46-47) seems a fairexpressionof
the view currenttoday:
The samepoeticdevice[repetition]
is to be foundalso in otherauthors,bothLatin and
of Lucretiusservea special function.The repetition
of
English,but the reiterations
sounds whichare unrelatedin meaningshows that the recurrence
of pure phonetic
tonesappealedto thepoet'sear fortheirownsake. But therecurrence
ofwords,which
is naturallymadepossibleby thesenseof thepassagein whichtheyrecur,is clearlya
3W. S. Maguinness, "The Language of Lucretius," in Lucretius ed. D. R. Dudley
(London 1965) 73-75.

This content downloaded from 143.107.252.19 on Sun, 4 Jan 2015 06:35:49 AM

All use subject to JSTOR Terms and Conditions

REPETITION IN LUCRETIUS

229

didacticdeviceused by Lucretiusto emphasizethe principleswhichhe is attempting

to enforceupona skepticalreader.

In short, prevailing opinion holds that the poet employed repetition

primarilyfordidactic reasons.
F. M. Smith,4the most recent writerto consider Lucretius' use of
repetition,likewise draws attentionto its practical and propagandistic
value in scientificand philosophicalexposition.He furthersuggeststhat
verbal echoes and repetitions,though not always necessarilyconscious,
oftenreveal importantthemesand conceptswhichwere identifiedin the
poet's mind. For example, the verses
cumtameninterse prostrati
in graminemolli
aquae rivumsubramisarborisaltae
propter
non magnis opibus iucunde corporacurant,

cumtempestas
etanni
arridet
praesertim
viridantis
herbas.
tempora
conspergunt
floribus

are employed in the proem of the second book (29-33), where the poet
suggests that simple pleasures ought to satisfy modern man, and are
repeated in Book 5 (1392-1396) in the descriptionof the amusementsof
primitiveman. The poet's feelingthat the ordinarydiversionswhich
satisfiedprimitiveman ought to be adequate forhis moderncounterpart
is reflectedin the verbal repetition.
But Smith's discussion,like thoseof Deutsch, Bailey, and Maguinness,
althoughit sheds lighton a new facetof the question,neverthelesstreats
repetitiononly as part of a larger topic, Lucretian thoughtprocesses.
Nor does it significantlyalter the fact that repetition,especially of
and casual manner.
phrases,has been treatedonly in a most perfunctory
Yet Lucretius repeats himselfmuch more frequentlythan most commentatorsseem to realize, or at any rate to document. This assertion
may be proven by analysing a passage of the De Rerum Natura in a
mannersimilarto that employedby Milman Parry in his endeavours to
demonstrate that the Homeric epics were oral verse compositions.' I
have chosen as a typical random sample twenty-five
verses beginning
at the firstparagraph after the middle of the poem (4.322-323). The
varietyof passages in whichthe phrasesare repeatedhas leftme satisfied
that these twenty-five
lines are fairlytypical. In the analysis, phrases
with continuousunderliningrecureitherverbatimor withsome negligible
change, and words with brokenunderlininghave analogues in the same
metricalposition.
4F. M. Smith,"Some Lucretianthoughtprocesses,"Hermathena
102 (1966) 77-82.
HSCP 41
5MilmanParry,"Studies in the Epic Techniqueof Oral Verse-making,"
(1930) 73-147. See especially118-120.

This content downloaded from 143.107.252.19 on Sun, 4 Jan 2015 06:35:49 AM

All use subject to JSTOR Terms and Conditions

230

PHOENIX
Splendida porro oculi6 fugitantvitantque tueri.7
sol etiam caecats, contra si9 tendere
et alte
propterea quia" vis magnast'2 ipsiuspergas.0
aera per purumx3gravitersimulacra feruntur14
et feriuntoculos'5 turbantia composituras.

praeterea splendor quicumque estl6 acer adurit'7

325

saepeoculosideoquodseminapossidet
ignis

330

in eorumdeniquemixta,
multaquesuntoculis23

335

multa, dolorem oculis quae gignunt"' insinuando.19

lurida praeterea fiuntquaecumque tuentur20
arquati, quia lurorisde corpore eorum2'1
semina multa22fluuntsimulacrisobvia rerum
quae contage sua palloribus omnia pingunt.24

'porro oculos-3.359.
7vitamquetueri-1.195, caeloque tuentur--1.152,4.434, 5.92, 6.50.
Ssol etiam caeli-2.210.
9contrasi-1.570.
o1dicere pergam-3.422.
sumere perge -2.347.
exsolvere pergo -1.932, 4.7.
inducere pergis -1.16.
concederepergat -1.1080, 2.237.

"proptereaquia-1.631; 2.72, 232; 3.572; 4.186, 320, 338, 349, 1220; 5.558, 920,
1446; 6.97,462, 1049.

12vismagna-6.530, 815.
xaaera per--2.146, aera per multum quia dum simulacra feruntur-4.358, aera per
multum-4.558, 5.580.
"simulacraferuntur-4.164, 176, 210, 239, 358, 735; 6.76.
16feriantoculos-4.217, 257; 6.923,feriuntoculorum-4.691.
16quicumqueest-5.177; 6.502.
"acer obhaesit-4.420.
Isquae gignatur--2.1078.
'1gignereconveniundo-2.923.
quaecumque creantur-1.169, fieri caeloque tuentur-1.152, and cf.feri in terris
orfient
caeloque tuentur--6.50.
2Idecorporeeorum-4.43, 5.154.
n2seminamulta-5.658, 6.160.
nobis-5.860
sunt
. -4.715
sis
-3.1025
ignis-6.863
mutaqu
sunt
24
constant--1.588,2.337, 694, 724
debent -1.1039
fiant -2.288
possent-1.166
tollat -1.701
omnia versat -2.882
reddat --1.1011

praestat--3.214
libant -3.11
restant-3.947

This content downloaded from 143.107.252.19 on Sun, 4 Jan 2015 06:35:49 AM

All use subject to JSTOR Terms and Conditions

REPETITION IN LUCRETIUS
E tenebris"6
autemquae suntin luce26tuemur
proptereaquia, cumpropiorcaliginisaer27
aterinitoculosprioret posseditapertos,28
lucidusaers9
insequiturcandensconfestim
qui quasi purgateos ac nigrasdiscutitumbrasso
aerisillius,nam multissx
partibushic ests2
mobiliormultisqueminutior
et magepollens.
luce replevit"
qui simulatque33vias oculorum84
atque patefecitquas anteobsederataer
(ater), continuorerumsimulacrasequuntur3
quae sita37suntin luces8,lacessuntqueut videamus."9
facerein tenebrise luce nequimus"4
quod contra40
proptereaquia posteriorcaliginisaera

231

340

345

Two factsemergeat once fromthis analysis. First, thereare no fewer

than thirtyverbatimrepetitionsin the twenty-five
lines, which is more
than I expected to find,and more, I think,than is generallyrealized.48
Secondly, most of the repetitionsserve no apparent didactic function.
C. Bailey too seemed aware of Lucretius' non-didacticrepetitionwhen
26etenebris-3.1, in tenebris-2.15, 54, 56, 58; 3.77, 88, 90; 4.231, 235, 348;
5.170; 6.38.
26suntin luce-4.347.
7"cf. 4.349; proptereaquia cum-6.462.
28posseditapertas-6.1050.
-3.122
g2editus
-6.685
percitus

abditus aer-6.1037
morbidus
-6.1097.

80abluitumbras-4.378.
31nammultis-2.592.
32hic/haecest-2.1066, 3.992, 4.1089, 6.238.
33quodsimul atque-3.211, 4.1041; simul atque-1.777, 4.40.
uvias oculorum-4.351.
86luce repletaest-2.806.
( vocamus-4.30
s
videmus-1.1060
rerumsimulacra vaga --4.127

recedent-4.130

feruntur-4.164,
210,239

vagari -4.724.
"7quaesita-2.802.
s3suntin luce-4.337.
agutvideamus-4.245, 255, 633.
40quodcontra--1.82, 780; 2.280.
41intenebris-see note 25; cf. in tenebris,in luci-4.235, and in luce
timemus-2.56,
3.88, 6.36.

4cf. line338 above.

*sBailey,forexample,ad loc. notesbutone.

This content downloaded from 143.107.252.19 on Sun, 4 Jan 2015 06:35:49 AM

All use subject to JSTOR Terms and Conditions

232

PHOENIX

he wrote: "Sometimeswordswhichhe had already writtenseem to stay

in his mind and come out in the same collocation in quite a different
context.. ." (145). Furtheron, Bailey calls this phenomenonthe "semiconscious runningof phrases in the poet's mind," but he never offers
an explanation forit. I thinkan explanationcan be given if we turn to
the firstpoet who employed repetitionextensivelyand consider why
he did so.
Homer's reasons foremployingrepetitionwere expoundedby Milman
Parry." Accordingto him, the repetitionswere the formulaeof an oral
traditionwhichHomer, an oral poet, inheritedfromhis predecessorsand
which provided him with a ready-madestore of diction which would
expressmuch of what he wanted to say as he improvisedhis heroicsong.
The formulae,in short,assisted Homer to compose. But formulaealso
made listeningeasier for the audience, since many of the repetitions,
used time and again, would be almost as familiarto the audience as they
were to the poet himself.
To returnto Lucretius,we can see at once an interestingparallel in
his didactic use of repetition,especially if we realize that Lucretius'
fromHomer's. As silent readingwas
audience was not so very different
in
unknown
virtually
antiquity, Lucretius' audience, like Homer's,
would have heard thepoem recitedor read aloud.45Thus, forhis audience,
Lucretius' repeated lines and phrases would have served a purpose
somewhat similar to that served by Homer's formulae.
But what is more importantis that the primaryfunctionof Homer's
formulaealso served Lucretius. They helped him to compose. It cannot
be denied that Lucretius had difficultywritinghis poem. He twice
mentionsthepatriisermonisegestas(1.832; 3.260), and revealshis constant
preoccupationwith his effortsto set forth"the nature of things"in the
"writings of our country's tongue" (naturam... rerum... patriis
exponerechartis[4.969-970]). Indeed, beforebeginningthefirstargument
of the poem, he remarksupon the difficulty
of the task in the famous
words
Nec me animifallit Graiorumobscurareperta

Latinisversibus
dificileinlustrare
esse,
cumsitagendum
multanovisverbis
praesertim

propteregestatemlinguae et rerumnovitatem;

sedtua mevirtus
tamenetsperatavoluptas
suavisamicitiaequemvis
laborem
efferre
suadetetinducitnoctesvigilare
serenas
dictisquibusetquo carminedemum
quaerentem

clara tuae possim praepanderelumina menti,

res quibus occultaspenitus conviserepossis.

[1.136-145]

44See the article cited above in note 5.

45Deutsch 1-3; G. L. Hendrickson, "Ancient Reading," C7 25 (1929) 182-196,
answered by W. P. Clark in C7 26 (1931) 698-700. I would agree with Dr Deutsch's
view that Clark fails to refute Hendrickson's arguments.

This content downloaded from 143.107.252.19 on Sun, 4 Jan 2015 06:35:49 AM

All use subject to JSTOR Terms and Conditions

REPETITION IN LUCRETIUS

233

There are also expressionssuch as

and

nunc age dicta meo dulci quaesita labore

percipe...

[2.730-731]

conquisitadiu dulcique repertalabore

digna tua pergamdisponerecarmina vita.

[3.419-420]

which indicate the effortthat went into his exposition of Greek Epicureanismin Latin verse. These verses also suggest that his effortwas
As the total vocabulary of
necessaryto surmounttwo major difficulties.
Latin is smallerthan that of Greek,therewas the considerableproblem
of findingequivalents forGreek termswherenone existedin Latin. But,
once the equivalents were found,thereremainedthe problemof fitting
them into Latin hexameters.It was here that the repetitionsassisted
the poet to compose. This can be easily demonstratedin his treatment
of one of the mostimportantconceptsof all the Graiorumobscurareperta,
that of the atom. The Greek philosophershad used the word &royos,
for
which Lucretius had several equivalents, such as primordiarerumand
seminarerum.Taking the last words as an example,we findthis collocation recurringin the overwhelmingmajority of cases in one metrical
position,namelyoccupyingthe finaltwo feetafterthe bucolic diaeresis."
Thereforewhen his Epicurean source used the word a&rouos,
Lucretius
had a ready equivalent around whichhe could build the rest of his verse
or which he could use to complete the line afterthe bucolic diaeresis.
Similarlythe Greek word oLbos,which Lucretius translated as rerum
natura,occurs in two or threefixedpositionsgovernedby the metrical
shape that the words take as natura changes its case. Again, once a
metricalpositionis establishedfora givenshape, the poet willrepeat the
collocation in that position most of the time. For example, of the six
occurrencesof the two wordsrerumnaturawithnaturain the nominative
singular,fouroccupy the same positionin the line,one is a slightvariant,
and the otheris quite different.
Occupyingthe line end, they are:
.... naturaquererum
rerumnatura creatrix
rerumnatura creatrix
rerumnatura repente
rerumnatura novarum

[1.498]
[1.629]
[2.1117]
[3.931]
[6.646]

and finallythe variant:

rerumprimum natura creatrix

[5.1362]

Thus we can see how the poet employedformulaeto assist him to overcome the difficultyhe found in expressingin Latin verse the subtle
concepts of Greek philosophy.
46Seminarerumoccurs finallyin the followinglines: 1.59, 176; 2.755, 678, 833, 1059,
1072; 5.916; 6.789, 1093.

This content downloaded from 143.107.252.19 on Sun, 4 Jan 2015 06:35:49 AM

All use subject to JSTOR Terms and Conditions

234

PHOENIX

It is also possible to demonstratehow a formulatakes shape in the

poet's mind and is subsequentlyemployedto assist his composition.Let
us take forexample the whole line
quae variae retinentgenteset saeclaferarum

[4.413]

Near the end of Book 2, in Lucretius' descriptionof the productsof our

alma materearth,we firstfindthe antithesisbetweenthegenushumanum
and the saecla ferarumin the line
et genus humanum,parit omnia saeclaferarum

[2.995]

The almost identicalwordsgens and hominumhad occurredearlierwhen

the poet praised Ennius, who won fame forhis poetryper gentesItalas
hominum(1.119), but this is the firstoccurrenceof the much used and
fineformulasaecla ferarum.Furtheron, the poet repeats the antithesis
in the verse
et varias hominumgentiset saeclaferarum

[2.1076]

and here he introducesinto the formulathe concept of the diversityof

the nations of man, whichhe firstexpressedin the famousmagna mater
passage in Book 2 with the wordshanc variaegentes(610). Coming back
to the original verse in question, we can now see how the formula
facilitatedLucretius' task. In the context he is discussing the many
apparent optical illusions.He observes that the risingsun oftenseems
to touch the mountainsfrombehindwhichit rises,thoughin fact a vast
distance separates them. To this distance, which is poeticallydescribed
as the "vast levelsofocean" (immaniaponti/ aequora), and to the "many
thousands of lands" (terrarummilia multa) he adds the formulaicantithesisbetweenmen and beasts. The line is not absolutelynecessaryhere,
but it subtlyand almost subconsciouslyaugmentsthe vastness,and was
easily adaptable to the contextby the change of quae foret and retinent
forhominum.Such examples could be multipliedto illustratehow the
poet employed formulaeto assist his composition.
To say that Lucretiusemployedformulaeto assist his compositionin
no way depreciateshis geniusor impliesthat he was forcedto use formulae, forit is possible to show with the formulagenus humanumwhat an
effectthe poet could achieve when he broke with his normal formular
usage. Several times in the descriptionof the rise of human civilization
in Book 5 the poet employs the formulagenus humanumafter a long
initial syllable (e.g., at lines 925, 1014, 1026, 1057, and 1145). Then
suddenlyin the midst of the section dealing with religionhe burstsout
with the words
O genus infelixhumanum

[1194]

By thrustingthe wordinfelixbetweenthe two membersof the frequently

This content downloaded from 143.107.252.19 on Sun, 4 Jan 2015 06:35:49 AM

All use subject to JSTOR Terms and Conditions

REPETITION IN LUCRETIUS

235

underlinesthe wretchedstate in
repeated formulaLucretius effectively
whichhe findsthe genus humanumas a resultof religion.So farwas the
poet frombeing constrainedto use formulae.
Nevertheless he found his task difficultfor the reasons he states.
Formulae helped him to overcome part of this difficulty.But this
emphaticallywas not the only reason why he chose to employ them.
There were two other factorswhich must have influencedhis decision.
First therewas the didactic value of the formulae,whichwe have already
discussed. The other considerationwas the fact that formulaewere an
integralpart of the epic style in which he chose to write the De Rerum
Natura.
Lucretius followed the example of Empedocles in setting forth a
philosophicalsystem(or at least the physical theoriesof a philosophical
system) in epic verse. The poet's high regard forhis Greek predecessor
is apparent in the lines which precede his criticismof Empedocles'
philosophy:
carmina quin etiam divini pectoriseius
et exponuntpraeclara reperta,
vociferantur
ut vix humana videaturstirpecreatus.

Such praise LucretiusotherwisereservedforEpicurus.

However, our poet did not follow Empedocles too closely, for his
and so was his philosophy.For the epic qualities
language was different,
of his work he rather imitated the greatest epic poet that Rome had
knownto that date, and the stylewhichresultedhas been well described
as "redolent of Ennius."4' Others have dealt with Ennius' influenceon
Lucretius' language and metre, and there is no need to repeat their
observationshere.48Our focus of attentionis upon the repetitions,and
I think that the most importantconsiderationwhich led Lucretius to
employformulaewas the factthat Ennius had done so beforehim.In the
remains of Ennius I was able to find the followingrepeated phrases:
lupusfemina (Ann. 68, 70), divumquehominumque(Ann. 249, 580, 581),
et simul (Ann. 91, 128, 352), stellisfulgentibusaptum (Ann. 29, 159,
and cf. stellisardentibusapta [Ann. 339]), in alto (Ann. 378, 380), summa
nitunturopum vi (Ann. 161, 412), cordesuo (Ann. 175, 548), haec effatus
(Ann. 47, 59), nox intempesta(Ann. 102, 167), in bello (Ann. 287, 327),
Romana iuventus (Ann. 469, 537, 550), ad caelum (Ann. 282, 531),
labitur uncta carina (Ann. 386, 478), olli respondit(Ann. 33, 119),
iamqueferre (Ann. 282, 593), o genitor(Ann. 113, 456), quatit ungula
terram(Ann. 224, 277, and cf. concutitungula terrain[Ann. 439]), est
47Maguinnessop. cit. (above, n. 3) 84.
48See in particular W. A. Merrill, "Parallelisms and Coincidences in Lucretius and
Ennius," Universityof California Publications in Classical Philology3 (1919) 249-254,
and Maguinness op. cit. (above, n. 3) 84 ff.

This content downloaded from 143.107.252.19 on Sun, 4 Jan 2015 06:35:49 AM

All use subject to JSTOR Terms and Conditions

PHOENIX

236

operae (Ann. 16, 465), permensaperumper(Ann. 71, 455), Poenos ...

oriundos (Ann. 220, 229), and finallythe most interestingphrase in
luminis oras (Ann. 114, 131).49 The finalexpressionis, of course,one of
and shows that Lucretius adopted at
Lucretius' best known formulae,s0
least one of his predecessor'sformulaeas his own. That this was the
only formulawhich Lucretius borrowedfromEnnius is unlikely,since
our poet repeats an astonishingnumber of phrases from the earlier
Indeed no fewerthan thirteenEnnian phrases are repeated in
writer.51
the De Rerum Natura, and among them are found some of Lucretius'
most memorableformulae,such as lumina solis, Acherusiatempla,and
Thus, if Lucretiusemployedhis predecessor'sformusomnoquesepultis.52
I
think
to employ formulaewas primarilystylistic,and
his
decision
lae,
followed necessarily from his decision to adopt a style "redolent of
Ennius."
To sum up. We have seen that formulaeconstitutean importantpart
of Lucretius' style. In deciding to use formulaethe poet was no doubt
influencedby two entirelypractical considerations,the didactic value
of repetitionand the assistance it would affordhim in composing.These
were not, however,the main reason why the poet employedrepetition,
forformularcompositionwas an aspect of the epic stylewhichLucretius
adopted fromEnnius to write the greatestmonumentto Epicurus that
survives,the De RerumNatura.
BISHOP'S

UNIVERSITY,

LENNOXVILLE,

QUEBEC

49This list was compiled from the Index sermonisof Ennianae Poesis Reliquiae ed.
J. Vahlen (reprintedAmsterdam 1963) and is restrictedto repetitionsin the Annales.
The followingare repeated in other works: arbores ventovacant-Varia 12, Sc. 185;
ex ore--Ann. 578, Sc. 306; monstrantviam-Sc. 321, 398 (monstrat);in somnis-Ann.
219, Sc. 36; horrescunttela-Sc. 140, Ann. 393 (-cittelis); mari magno-Ann. 445, Sc.
65, which is employed by Lucretius in 2.1.
60Used four times; 1.22, 179; 5.224, 781.
"6Merrillop. cit. (above, n. 48).
62Theyare: lumina solis (Ann. 283; Lucr. 1.5, 989; 2.108, 162, 654; 5.462); Acherusia
templa (Sc. 107; Lucr. 1.120, 3.25, 86); somnoque sepulti(s) (Ann. 292; Lucr. 1.133,
laborem(Ann. 425; Lucr. 3.999, 5.1272, 1359); virumvi(s) (Ann. 276;
5.975); sufferre
Lucr. 1.728, 2.326); caeli templa (Ann. 65, Lucr. 1.1064, 6.1228); aetherisoris (Sat. 4;
Lucr. 2.1000, 3.835, 5.143, 683); fortis equi vis (Ann. 374-fortis equus, Ann. 486equos vi; Lucr. 3.8, 764); media regione(Ann. 505, 481-mediis regionibus;Lucr. 3.140,
6.732); duri ... laboris (Ann. 345; Lucr. 3.999, 5.1272, 1359); in somnis (Ann. 219;
Lucr. 3.431, 4.34, 770, 789, 965, 972, 988, 1006, 1012, 1097; 5.62, 885, 1171, 1181);
tolliturin (Ann. 442; Lucr. 5.265, 6.507);fit copia (Ann. 407; Lucr. 5.359 [Lachmann,
mss. sit], 6.829).

This content downloaded from 143.107.252.19 on Sun, 4 Jan 2015 06:35:49 AM

All use subject to JSTOR Terms and Conditions