Classical Quarterly51.

2 537-544 (2001) Printed in Great Britain 537

MOONSHINE: INTERTEXTUAL ILLUMINATION

IN PROPERTIUS 1.3.31-3 AND PHILODEMUS,
ANTH. PAL. 5.123*

In the famous third poem of Propertius' first book the elegiac lover recounts how he
returned from a party one night, decidedly the worse for drink, to find his beautiful
Cynthia fast asleep. He was taken, he says, with the urge to ravish her there and then,
but held back-because he feared the sharp edge of her tongue (non tamen ausus
eram dominae turbarequietem, / expertae metuens iurgia saeuitiae, 17-18). Instead he
just stood watching her, toying with her hair and hands, and (ironically) worrying,
whenever she moved and sighed in her sleep, in case she should be dreaming that
'some man' was trying to rape her (neue quis inuitam cogeret esse suam, 30). In this
fashion he went on

donecdiuersaspraecurrensluna fenestras,
lunamoraturissedulaluminibus,
compositosleuibusradiispatefecitocellos (31-3)

untilthe moon racingpast the open shutters-

the meddlesomemoon withbeamswantingto linger-
openedwithherlightweightraysthe eyesat repose
-at which point she gave him a tongue-lashing anyway, accusing him of having spent
the night so far with another woman.
'Der durch das Fenster scheinende Mond als erotisches Motif auch bei Philodem.'
So Rothstein in his 1920 commentary,' quoting lines 1-3 of the following epigram:

VVKTrpLV'r, SWKePWS, ?eA'rvr'

tLAO7ravvvxe, #aCve,
,aLve, St1' eVTprpfWV
aXowvaAAovOvptSwv.
aiyare Xpvaer'v KaAA(arov eg T ra 6LAetVTwv
pya KaTrorrevetv ov kO>ovosdOavar f.
oXA eirs KaL rTVSE Kal 7jLeas, otSa, ?eXrfvr1
Kat yap ar)v bUvX)vZqAeyev 'EvSvuVwv. (Philodemus, Anth. Pal. 5.123)

Ladyof the night,two-hornedone, loverof the all-nightrevel,shine,Moon,

shine,fallingthroughthe latticedwindows.
Castyourlighton goldenKallistion;on the acts of lovers
an immortalgoddessspieswithoutoffence.
Youcountboth herandme lucky,I know,Moon;
for Endymionset alightyourownsoul.

Later commentators either also simply advert to this epigram with equal lack of
elaboration or enquiry2or do not even mention it. The same is true of most critiques

* Materialincludedin thispaperwasfirstpresentedto an audiencein the Universityof Leeds,

and I am gratefulforobservationsmadeon thatoccasion.My thanksalso to the editorsof CQ
andto ProfessorE. J. Kenneyforhelpfulcommentson an earlierdraft.
I M. Rothstein,Die ElegiendesSextusPropertius, Bd I (Berlin,19202;Dublin/Zurich,19663).
2 See LiberI (Monobiblos)(Leiden,1946);W.A. Camps,
e.g. P.J. Enk,S. PropertiiElegiarum
Propertius. Elegies Book I (Cambridge, 1961); P. Fedeli, S. Properzio. II primo libro delle elegie
(Florence,1980),ad loc.
538 J. BOOTH

of Propertius 1.3.3 Of those who press the comparison at all, Lyne and Baker go
furthest: 'the motif recalls . . . the opening of Anth. Pal. 5.123 (Philodemus), but it
has been transformed by Propertius' application and development of it';4 'even if
Philodemus' epigram is to be accepted as the model for Propertius' moon as
ETaKoTros . . . the elegist's independence of treatment produces a poem which
develops through an emotional range that is completely foreign to its Greek models'.5
But that is not very far. Admittedly, both go on to offer some comment on what they
think Propertius makes of the moon,6 but they do not consider how Philodemus'
poem may participate in this, or whether Propertius' treatment encourages any
reappraisal of Philodemus'. In a critical world attuned to the creative potential of
intertextuality, those further steps now beckon.

1. TOWARDS AN INTERTEXT
First, how clear are the pointers to specific involvement of Anth. Pal. 5.123 here?
Though the basic theme of sexual desire (or actual sexual assault) inspired by the
sight of a sleeping beauty makes other appearances both before and after Propertius,7
viewing by moonlight is indeed a detail apparently exclusive to his poem and
Philodemus' epigram. Yet the idea of the moon shining into a bedroom is hardly so
extraordinary that it could never have suggested itself to two poets independently. A
case of true 'reference', then, or 'accidental confluence'?8 'Some would contend that
this is immaterial, arguing that intertextuality can operate on the basis of both9 but,
as it happens, there is in this instance enough evidence of deliberate allusion
('reference') on Propertius' part to impress even those inclined to 'philological
fundamentalism'.0l Wills"l has identified a range of potential 'markers' of allusion in
the form of various kinds of verbal repetition. When Propertius' lines are 'litmus-
tested' for these, the result is a striking positive. The fifth-foot/first-foot epanalepsis

3 These are very numerous.Treatmentsup to 1983 are listed by P. Fedeli and P. Pinotti,
BibliografiaProperziana (1946-1983)(Assisi,1985),52-4. S. J. Harrison,'Drink,suspicionand
comedyin Propertius1.3',PCPhSn.s.40 (1994),18-26,addslateritemsin his n. 1. See nowalso
S. Kaufhold,'Propertius1.3:Cynthiarescripted',ICS 22 (1997),87-98;G. Lieberg,'Desertusin
Properz 1.3', WSn.s. 12(1999),87-90.
4 R.
O.A. M. Lyne,'Propertius andCynthia.Elegy1.3',PCPhSn.s.16(1970),60-78 at 74. Cf.
M. Hubbard,Propertius (London,1974),21:'Philodemus' moon is put to a use Philodemusdid
notthinkof'.
5 R. J. Baker,
'Beautyand the beastin Propertius1. 3', in C. Deroux(ed.), Studiesin Latin
Literatureand Roman History II, CollectionLatomus 168 (Brussels, 1980), 245-58 at 246.
6 See nn. 18,37, 40 below.
7 J.C. Yardley,'PaulusSilentiarius,
OvidandPropertius',CQ30 (1980),239-43at 243,points
to Anth.Pal. 5.199 by the third-century
B.C.Hedylusand Ter.Eu. 601-6 (cf. n. 59 below) as
examplesof earliertreatments.He arguesthatother,nowlost, instancesof the motif in epigram
or New Comedycould well haveinspiredcertainlaterversions,includingPaulus'sixth-century
A.D. epigram (Anth. Pal. 5.275).
8 These are the termsused by R. F. Thomasin his attemptsto distinguishreader-perceived
fromauthor-designed allusion('Catullusandthe polemicsof self-reference
[64.1-18]',AJPh103
[1982],144-64;'Virgil'sGeorgicsandtheartof reference',HSCP90 [1986],171-98).
9 For example S. E. Hinds, Allusion and Intertext: Dynamics of Appropriationin Roman Poetry
(Cambridge,1998),17-50;cf. H.-C.Gunther,Eikasmos9 (1998),245. I acceptthe impossibility
of proof in most casesthata 'message'perceivedby the readercoincideswhollyor evenpartially
with any intendedby the author.This is not to say, however,that indicationsof conscious
authorialallusionshouldbe ignored.
"0The phraseis Hinds's(n. 9), 19.
J. Wills, Repetitionin Latin Poetry.Figures of Allusion (Oxford, 1996).
 INTERTEXTUAL ILLUMINATION IN PROPERTIUS AND PHILODEMUS 539

luna ... / luna (31-2) is taken by Wills (along with another at 25-6) to 'mark' non-
specific allusion to the Ariadne-episode of Catullus 64.12The idea is not unpersuasive
in itself (this allegedly neoteric mannerism is notably common in the Catullan
Ariadne-episode, and Propertius' poem has undeniable thematic and verbal links
with it), but if the epanalepsis in lines 31-2 'marks' any allusion, it 'marks' much
more convincingly a specific one-to Philodemus. For Propertius' luna ... / luna is an
exact metrical echo of Oa ve ... / Oaive at Anth. Pal. 123.1-2: same prosody, same
sedes within the elegiac couplet.13 And if still further indication of conscious
involvement with the epigram is needed, it is there-in the fancied etymological link
between luna and aa(ivevat which Propertius hints. Varro and Cicero already derive
luna from lucere, the Latin for aa(velv:luna quod sola lucet noctu (Var. L. 5.68); luna
a lucendo nominata (Cic. N.D. 2.68).14 From /alveLv later grammarians (and quite
possibly sources known to Propertius too) derived fenestra;15 and fenestras is
suggestively enclosed by the two occurrences of luna here.16 A complex web of
metrical and verbal indicators, then, may be uncovered to confirm past assumption.
Anth. Pal. 5.123 is a presence behind Propertius' lines, and its future as an intertext
now looks especially promising.

2. READING BETWEEN THE POEMS

So to the close comparative reading on which intertextual interpretation must rest.

(a) Reading Propertius throughPhilodemus

Line 31 Propertius' moon, diuersas praecurrens .. fenestras, immediately differs
from Philodemus' in being spontaneously active (rather than invoked by the lover).
The exact nature of the action is much debated. The usual meaning of praecurrere +
accusative is 'race ahead of' (literally or figuratively) (OLD s.v. Ib, 3, 4),17but there is
no sense in this here: windows (or window shutters) do not compete, in speed or in
anything else, nor can they be 'anticipated'.18 'Race past' is what most inter-

12Wills(n. 11), 135,followingD. O. Ross,Jr,Backgrounds to AugustanPoetry:Gallus,Elegy

andRome(Cambridge,1975),55.
13
Arguablythereis in Propertius'repetitionluna.. ./ lunaalso a less close, but still notable,
echo of Philodemus'ZeA vr ... ZeAXrvr, 'Moon ... Moon',prominentlylocatedat the end of
the firstandlast hexameterof Anth.Pal.5.123.
'4 Bothrecordedby R. Maltby,A Lexiconof AncientLatinEtymologies (Leeds,1991),s.v.luna.
15 See especiallyNonius36.11fenestraea Graecouocabuloconuersum est in latinum,adro TroV
/a(ve?v(recorded,along with Isid. Orig. 15.17.6,by Maltby [n. 14] s.v.fenestra);also J. C.
McKeown,Ovid,AmoresII, A Commentary on BookI (Leeds,1989)on Am. 1.5.3-4.
16 The identicalline-ending lunafenestrasoccursat Virg.A. 3.152and Ov.Pont.3.3.5 (quoted
in nn. 28 and 27 respectively;see G. Danesi Marioni,'Lunafenestras',in G. Catanzaroand
F Santucci[edd.], TrediciSecoli di Elegia Latina [Assisi, 1989], 283-7). Conceivablythis
particularpiece of etymologizing(which constitutes a 'type' additional to those helpfully
schematizedby F Cairnsat PCPhSn.s. 42 [1996],26) becameclich6d.The playcontinueswith
luminibus at the end of line 32. Thisprovidesan etymologizinglink betweenfenestras at the end
of the previousline and lunaat the beginningof the sameline(32);furthermore, lunawas itself
claimedto derivefromlumen(by Firmicus;see Maltby[n. 14],s.v. luna).For positioningas a
'marker'of etymologisingsee R. Maltby,AevumAntiquum 6 (1993),270,Cairns(above),33, and
for repetition,Wills(n. 11),470-1.
17W.Wimmel,'Lunamoraturis sedulaluminibus.Zu Properz1.3.31-2',RhM 110(1967),70-5
at 73, n. 7 givesfurtherexamples.
18 The samegoes for 'Abbilder' ('images')of windows,i.e. moving,window-shaped patchesof
illuminationinsidethe room,whichis howWimmel(n. 17),74, followedby Baker(n. 5), 248-9,
540 J. BOOTH

preters have chosen to understand, though with some anxiety, since praecurrere for
praetercurrere19 appears to be unattested elsewhere.20Augustan Latin does, however,
offer the first instances of other compound verbs of movement in prae-, rather than
praeter-, used transitively to denote something passing, or being passed by, something
else.21Propertius himself seems to supply one example in praeuehi at 1.8.19: utere
felicipraeuecta Cerauniaremo ('travelling past Ceraunia, [Cynthia,] deploy propitious
oar', sc. 'row safely past .. .);22 a similar experiment here with praecurrerecan there-
fore hardly be deemed impossible.23
But then what of diuersas .. . fenestras? Diuersus, when used attributively of in-
animate things, most frequently denotes two or more of the same 'turned/facing in
different directions' or 'positioned apart from each other' (OLD s.v. 1, 2, 3). Either of
these meanings is conceivable if fenestras is taken to refer to actual window-apertures,
and if a room with more than one of them is envisaged.24Yet to insist on this in the
case of a bedroom seems curiously pointless.25On the other hand, fenestras taken as
'window shutters' makes particularly good sense if diuersas can imply that they are
wide open26-'positioned apart' in that each has been swung back from the other in an
arc diverging from the central point where the two meet when closed.27 If this is
Propertius'meaning, what he would appear to want to emphasize is the absence of any
impediment to the shaft of moonlight entering Cynthia's bedroom.28The contrast
wishesto understand fenestrasand also luminibusin the pentameter.In anycase,suchmeaning
for these wordsis just as unparalleledas some of the suggestionsfor diuersosand praecurrens
rejectedforthatreasonby Wimmel-and a lot moreobscure.
19 praetercurrereitself is not attestedbeforefourth-century
20The one other prose(Wimmel[n. 17],73).
Propertianinstanceof praecurrere (1.13.25) has the 'normal'figurative
meaning of 'surpass',but thisprovesnothingaboutthe presentpassage.
21
praelabiandpraenatare in Virgil,praefluerein Horace.Therangeis extendedin SilverLatin:
praefestinareandpraegrediin Tacitus(who showsa likingfor compoundsin prae- = praeter-),
praenauigare in Seneca,ValeriusMaximus,andPliny.
2 The text is uncertainat this
point (I readhere uterewith PDV'Voinsteadof ut te with
NAFV2),butpraeuecta,if correct,can only be f. sing.voc. SeefurtherCamps(n. 2), ad loc.
23 Admittedly, thereare no otherinstancesof a prae-compoundreferringto aerialprogress
past a structureon the ground;most relateto progressover land or waterpast a (sometimes
extensive)topographicalfeature(see Wimmel[n. 17], 73, n. 7). Since the personifiedmoon,
however,was conventionallysupposedto driveacrossthe sky in a chariot(cf. Theoc. Id. 2.163;
Tib.1.8.21,2.4.18;Ov.Am.2.5.38),herheavenlyprogressmaynot havebeenfeltto be as different
as it wouldat first appearto be from the terrestrialkind by ship or horsewhich otherprae-
compoundscan imply.
24Anotherpopularinterpretation, 'window(s)oppositethe bed', is untenable:withoutdirect
mention of a bed (or anythingelse) the referencemust be to one fenestrain relationto
(an)other(s).Furtherdiscussionin Camps(n. 2) and Fedeli(n. 2), ad loc.
25 Contrast the appropriatelywell-lit 'little room' used as a miniaturelibrary in Pliny's
Laurentinevilla:cubiculum inapsidacuruatum, quodambitum solisfenestrisomnibussequitur(Ep.
2.17.8).
26 So J. P.Postgate,PCPhS 11 (1892),12,contrastingHor. Carm.1.25.1iunctas.. .fenestras.
Theattractionof openshutterson a moonlitnightis easilyunderstood.Cf.nn. 27, 28 below.For
the mechanismof wooden windowshutters,see H. Bliimner,Die rimischenPrivataltertiimer,
Handbuch derAltertumswissenschaft 4.2.2 (Munich,1911),102-3.
27 Exact
parallelsfor diuersuswith this sense are to seek, but the same basic idea of
corresponding bilateraldivergencefroma centrallineis also presentin Festusp. 514 L ualgos...
dici quidiuersassurashabeant(quotedby T. Birt,Die Cynthiades Properz[Leipzig,1921],106)
and Hyg.Astr.2.6palmasdiuersastenderead caelum(quotedby Camps[n. 2], ad loc.). A similar
notion lies behind the figurativeuse at Prop. 1.10.15possumego diuersositerumconiungere
amantes.For bivalveshutterscf. Ov. Pont. 3.3.5-6 nox erat et biforesintrabatlunafenestras,I
menseferemedioquantaniteresole.
28 A comparablesituationat Virg.A. 3.150-52 [Penates]uisi ante oculosastareiacentis/ in
 INTERTEXTUAL ILLUMINATION IN PROPERTIUS AND PHILODEMUS 541

with Philodemus would then be nicely pointed: his EvTrpr1wv Ovp('ov, literally 'well-
pierced shutters', suggests some sort of all-over latticing29which would have the effect
of filtering the beam.

Line 32 The single epithet which Propertius gives the moon-against Philodemus'
opening string of three, and striking for that in itself30-is sedula, 'zealously
attentive'. This does not correlate at all with any of Philodemus' epithets. Is, then, the
very absence of any equivalents significant? Perhaps. That Propertius does not
attempt to echo vvKreptvq, literally 'belonging to the night', is unremarkable, for this
is common, purely ornamental, and, of the moon, even superfluous. But omission of
anything corresponding to the other two, distinctly choice, epithets31 is more
interesting. If the shining moon is not called 'two-horned' (StKepws), are we to think
of a full moon, which will inevitably cast a more powerful light?32And if not called
'lover of the all-night revel' (ctAo7ra,vvvxE),33 could the moon even be supposed to
dampen nocturnal frolics?Then, what of sedula itself? This is a more equivocal word
than any of Philodemus' epithets, capable of denoting not only a welcome degree of
attentiveness ('dutiful', 'zealous') but also, from the recipient's point of view, a
misplaced one ('obtrusive', 'meddlesome').34The crucial factor here is the meaning
and syntactical relation to sedula of the enclosing phrase moraturis ... luminibus.The
meaning of lumina most obviously suggested by the context is '(moon)beams', but
the word is a common poeticism for 'eyes', a sense also clearly appropriate in view of
the moon's semi-personification35and one encouraged by the reference to her 'spying'
(Ka-ro7Trrdtv) in Philodemus (Anth. Pal. 5.123.4). The future participle moraturis is
trickier. The claim has been made that it 'represents, as often, the apodosis of an
implied condition, [the implied protasis being] si sedula non fuisset',36 i.e. with
beams/eyes which would have lingered, if she (sc. the moon) had not been zealously
attentive. The 'zealous attentiveness' is taken to consist of 'sticking to her time-
table'.37The construction itself, especially with protasis explicit, is certainly well

somnis multo manifesti lumine,qua se I plena per insertasfundebat lunafenestras. In Virgil too an
unusualexpression,i.e. insertasapparently= nonsertas,'not joined',indicatesthe unshuttered
natureof thewindow(so suggestsServiusinteralia;see furtherR. D. Williams,Virgil.AeneidIII
[Oxford,1962],ad loc).
29 See A. S. F Gow and D. L. Page, The GreekAnthology. The Garlandof Philip II (Cambridge,
1968); D. Sider, The Epigramsof Philodemus(Oxford, 1997), ad loc.
3
Arguablysedulais additionallyemphasizedby the relativelyunusualword-patternof the
pentameter:in Book 1 this is one of only five examplesof a four-wordline wherean adjective
qualifying beforea polysyllabiclast word.
a nounin the firstfoot is placedimmediately
31On the see Gow-Page(n. 29), Sider(n. 29), ad loc.
rarityof 8LKeppwSand LAiorravvvvXo
32Cf. Virg.A. 3.152,Ov.Pont.3.3.6(quotedin nn. 28 and27 respectively).
3 'The
rravvvXswasan all-nightfestival,overwhichSelenemightbe thoughtto preside.Such
festivalsoften providedan opportunityfor seduction'(N. Hopkinson,A HellenisticAnthology
[Cambridge,1988],ad loc).
34Cf. Hor.Ep. 1.13.5(to the man deliveringHorace'swritingsto Augustus)ne ... odiumque
libellis I sedulus importes opera uehementeminister, and see C. O. Brink, Horace on Poetry III:
EpistlesBookII (Cambridge,1982)on Ep.2.1.260.Ovidat Ars 3.699-700avoidsambiguityby
qualifying seduluswith male: coniugisad timidasaliquis male sedulusaures I auditos memori detulit
ore sonos.
35 Improbable is a referenceto Cynthia'seyes (so e.g. Hubbard [n. 4], 21, following Paley: 'as its
lightcomesthroughthe openedwindowit plays"thebusybodyto eyesthatwouldelse havebeen
laggard"').morariwithoutsomnoor similarwouldbe a veryobscurewayof alludingto the eyes'
inclinationto remainclosed.
36 D. R. Shackleton Bailey, Propertiana(Cambridge, 1956), 13.
37So Lyne(n. 4), 75, whoclaimsthatthisis whatsedula'mostobviously'refersto.
542 J. BOOTH

enough attested in post-Augustan Latin,38but (i) it is not at all clear that sedula could
mean 'attentive' to a duty unspecified or unhinted at in the text (or even in the
intertext),39 and (ii) it does not (to my mind) make much sense to set up a dis-
agreement between the moon herself and her beams (or eyes).40The more satisfying
meaning is 'willing to linger'. This volitive/intentional nuance by no means 'lacks
grammatical warrant',41but is a usage of the future participle that also emerges
clearly in the Silver Age42and must (like the development of the verbal compounds in
prae- for praeter- discussed above) have had its beginnings in an earlier period
(Propertius himself uses a very similar expression of the Sun's rays at 3.20.12: Phoebe,
moraturae contrahe lucis iter).43 Here the volitive nuance allows moraturis...
luminibus (as ablative of respect) to explain the epithet it encloses:44 the moon is
sedula in the lover's estimation because of what he perceives as willingness on her part
to let her lumina linger (the meaning is reinforced metrically by the actual lingering
over the last three long syllables of moraturis).The uncomplimentary sense of sedula,
then, would seem to be a distinct possibility: the presence perceived as benign by the
lover in Philodemus' epigram may well prove here to be an officious busybody.45

Line 33 With this line (in which word-order cleverly reflects sense-compositos
ocellos is actually prised apart by leuibus radiis) Propertius carries events in his own
poem beyond the point at which Philodemus' epigram stops. The full consequences of

38 See R. Westman,Das Futurpartizip als Ausdrucksmittel

beiSeneca(Helsinki,1961),199-206;
for more exampleswith protasisexplicit,see E. C. Woodcock,A New LatinSyntax(London,
1959),?199.
39Contrastthe muchmorepreciseindicationof whatthe negativehypothesiswouldbe at e.g.
Ov. Met. 8.409-10 cui [cornu]benelibratouotiquepotentefuturo,I obstititobstipafrondosusab
arboreramus:theweapon'wouldhavebeen'(futuro)ableto hit the target-if onlythebranchhad
not got in the way.At firstsight Hor. S. 2.8.43-4 'haecgrauida'inquit/ 'captaest, deteriorpost
partumcarnefutura',cited by Fedeli (n. 2), ad loc., seems to offer a parallelfor the alleged
conditional constructionin Propertius,with a single word, grauida,hinting at the implied
negativeprotasis;but themeaningof grauidais unambiguousandrequiresno supplementation.
40 Pace Lyne (n. 4), 75: 'Propertiushandlesthis somewhatartificialdichotomywith great
charm.'
Haupt's'dasverweilenwollte'.
41 So ShackletonBailey,loc. cit. (n. 36),rejecting
42See Westman(n. 38), 104-5, 147-8 on 'Bereitschaftsbedeutung' and 'Absicht'.Westman
observesthat the future participlefrom verbs of 'bleibenund bestehen'seems particularly
susceptibleto extensionof meaning;see also Leumann,Hofmann,and Szantyr,Lateinische
Syntax undStylistik(Munich, 1965), 390. The generaltrendtowardsfreeruse of the future
participlein Latin poetryis noted by R. G. M. Nisbet and M. Hubbard,A Commentary on
Horace,OdesII (Oxford, 1978), on Carm.2.3.4 moriture.A timely caution againstgeneral
scholarlyreluctanceto countenancelinguisticinnovationby Latinpoetsis issuedby E. J.Kenney,
'Vt eratnouator:anomaly,innovationand genrein Ovid, Heroides16-21', in J. N. Adamsand
R. G. Mayer(edd.),Aspectsof the Languageof LatinPoetry(Oxford,1999),399-414;cf. id. CR
48 (1998),311-12.
43Cf. also Prop.4.5.61uidiegoodoratiuicturarosariaPaesti,whereuictura= 'willingto/ready
to bloom'. ShackletonBailey (n. 36), 242 classifiesthis an instanceof the futureparticiple
denoting'apparentprobability'.
44V. Eckert(Untersuchungen zur EinheitvonProperzI [Heidelberg,1985]),whilealso taking
moraturis... luminibus to dependon sedula,notes the lack of parallelsfor such a construction
with this adjective(293, n. 92). Propertius'adventuroususe of the ablative,however,is an
acknowledgedfeatureof his style;see H. Bausch,StudiaPropertiana de liberioreusu ablatiui
(Diss.Marburg,1920).
45 The possibilitymay be sensed more quicklyby the readerworkingbackwardsfrom a
'modem'intertext:'Busyold fool, unrulySun, / Whydoes thou thus / Throughwindowsand
throughcurtainscall on us?'(JohnDonne,TheSunRising).
 INTERTEXTUAL ILLUMINATION IN PROPERTIUS AND PHILODEMUS 543

the moon's action in waking Cynthia are seen only when Cynthia speaks,46 and at
first sight the action itself seems to be described in entirely neutral terms. But not so,
perhaps, when read against the epigram with all its implications. The lover there
half-explains in the last line why he counts the moon (Selene) as a friend and
considers her gaze benign: it is because she has herself known love-for the mortal
Endymion.47 The particular relevance of this myth to the love-situation sketched
in the epigram, however, is left inexplicit. In fact it confirms the assumption that
Kallistion is asleep,48 for eternally asleep was how Selene found Endymion.49
Propertius himself does not mention him. He has no need: the involvement of
Philodemus' epigram in his lines so far is enough to prompt recollection of the
well-known story of the moon's love as he embarks on the final chapter of his own.
Contenting himself now with little more than 'just looking' at Cynthia, he might, in
view of the moon's similar position vis-a-vis her own loved one,50 reasonably have
expected her apparently lingering gaze to indicate her solidarity with him. Her
subsequent waking of the sharp-tongued Cynthia with her leuibus radiis thus almost
amounts to treachery, and Propertius' choice of words suggests as much. For
although leuis in its primary sense of 'lightweight' is appropriate enough of the
borrowed light of the moon, it can also mean 'fickle' or 'unreliable'.51 From the
Propertian lover's point of view, that is precisely how the moon will look-apparently
siding not with him, but with Cynthia, to whom, by her active intervention, she gives
the chance to complain of the lover's desertion face to face.52 The moon has
successfully vied with the lover for Cynthia's attention, too, in that her beams elicit a
response, whereas his attempted gifts have (as he sees it) been spurned (lines 25-6).

(b) Reading Philodemus throughPropertius

The effect of Propertius' poem on the reading of Philodemus' own can be dealt with

4 A subtle hint at what is to come, however,may be detectablein compositos... ocellos.

composites= 'calm','composedin sleep'(cf. Ov.Am. 1.4.53benecompositussomnouinoqueand
McKeown[n. 15],ad loc.), but the meaning'composedin poetry','written(about)'perhapsalso
suggestsitself.It wasby wayof hereyesthatCynthiawasfirstintroducedat 1.1.2as Propertius'
domineeringmistress,anda reminderof thathere,whenshe is aboutto showhermettle,wouldbe
appropriate. Forcomponere withthissense,cf. Ov. Tr.2.362compositepoenas... amoredediand
furtherOLDs.v.,8b.
47Such, reputedly,was her passion for this beautifulCarianshepherdthat she regularly
descendedto earthto visithim-a prettyexplanationforhercyclicinvisibility.
48 Sider(n. 29), pp. 113, 115 also assumesthat Kallistionwas envisagedas remaining asleep
throughoutthe encounter,but CQ'sanonymousrefereepertinentlypoints out that the plural
LAE6vwrwvrather'suggestsmutuallovemaking'.
49 According to some accounts, perpetualsleep was granted to Endymion(along with
immortalityand perpetualyouth)by Zeus,eitherat his own request(e.g.Apollod.Bibl. 1.7.5)or
bySelene'scontrivance(e.g.Cic. Tusc.1.92);accordingto others,everlastingsleepwasimposedas
a punishmentfor his attemptedrapeof Hera(scholiaston Ap. Rhod.4.57-8).
5s QuitehowSelene'spassionforEndymionwassupposedto be satisfiedwe maywellwonder.
An epigramof Meleager(Anth.Pal. 5.165.5-6)impliesthat it neverwas; Cicero(Tusc. 1.92)
intimatesthatthe goddessdid at leastmanagea kiss;Lucian(Dial. Deor 19 [11]2) sidestepsthe
issuewitha coy aposiopesis.
51See OLDs.v. 15, and cf. Prop.1.15.1saepeego multatuaeleuitatisduratimebam.It is not
difficultto imagineleuibusradiisbeing articulatedhere with a sarcasticedge which belies its
superficialsuggestionof 'die leisen,sanftenStrahlen'(Rothstein[n. I], ad loc.).
52 Thatthe moon shouldproveto be Cynthia's activeally is particularly giventhe
appropriate,
associationof Cynthia'sname with 'Cynthian'Apollo, and hence also with Apollo's sister,
Artemis-Diana,the virgingoddesswithwhomthe moon was herselfwas sometimesidentified;
see E. N. O'Neill,'Cynthiaandthemoon',CPh53 (1958),1-8.
544 J. BOOTH

more briefly. It is a destabilizing one-a highlighting of how much is commonly

assumed, and yet cannot in truth be known, about the situation in the epigram. We
cannot be sure that the narrator is in bed with Kallistion,53and is not stumbling into
the room drunk after visiting elsewhere. Cynthia's lover's admission of his inebri-
ation, together with her accusations against him (never explicitly denied!), in effect
query the apparent innocence of the epigrammatic lover's passing reference to 'all-
night revel'.54We cannot be sure that the narrator of the epigram is expected and
welcome-that he is not the sort of boorish intruder that Propertius envisages55(and
fails to recognize as potentially none other than himself). -ra LAE('VTIcV Epya
represents, after all, only the epigrammatic narrator's perspective on the situation:
there is no telling whether Kallistion's own would be the same.

3. ASSEMBLING THE MESSAGE(S)

To sum up, an intertextual reading here both (i) projects added significance on to
Propertius' own lines and (ii) retrojects uncertainty on to Philodemus' epigram.
(i) Propertius in effect takes issue with the epigrammaticlover's o~ia ('I know'). The
moon, he contests, in spite of her love for the eternally sleeping Endymion, is not
necessarily sympathetic, or even harmless, towards one who comes upon his own
beloved asleep (irrespective of what designs he may actually have on her). If the
bedroom shutters happen to be open, the moon will sabotage his pleasure. For, far
from casting a diffused, undisturbing light, she will come busybodying in, throwing a
full beam across the bed quite long enough to wake the girl-who, after all, is not
Endymion and will react with predictableannoyance.56To put it another way, romantic
trysting by moonlight is a myth.57
(ii) Philodemus' epigram is noted, rightly enough, for its 'sensuous tone and
exquisite phrasing'.58 But Propertius' subsequent handling of its central theme
retrospectivelyopens up a shocking possibility: it could, all the same, be presenting the
prelude to an opportunist rape.59
'Erotisches Motiv auch bei Philodem'? Indeed, but there is much more to it than
that.

Universityof WalesSwansea JOAN BOOTH

[email protected]

53'Kallistionis lyingasleepalongsidethenarrator'(Sider[n. 29], 115).

4 Cf. n. 33 above. 55 Seep. 537.
56
By initiallylikeningCynthiaasleepto certainfiguresof myth(lines 1-8) whosereposewas
'frozen'foreverin worksof visualart(seeLyne[n.4], 66, 75),Propertiuscreatedthe illusionthat
her beauteousslumberwas similarlyundisturbable. The artisticconnectionmay well havealso
broughtto mind iconographicdepictionsof Endymion'strulyeverlastingsleep (see e.g. LIMC
3.2, 551-61,plates5,22, 61,73, 78, 81, 83,85, 87a,93,94,98), and,ironically,it is the intervention
of Endymion'sown admirer,the moon, whichultimatelyshattersthe illusionof permanencein
Cynthia'scase.
57 Ovid'sLeanderat Her. 18.59-64.appearsto contradictPropertius' contentionof themoon's
unhelpfulness to a (male)lover.He describesheras innostrasofficiosauias(60),officiosaperhaps
pickingup Propertius'sedula,and, likethe narratorin Philodemus'epigram,cites herown love
for Endymionas reason for her to show him favour(61-3); see furtherE. J. Kenney,Ovid,
HeroidesXVI-XXI (Cambridge,1996), ad loc. The Propertianelegiac lover has apparently
recoveredfaithin themoon'spotentialgoodwillwhenhe requestsherco-operationin lengthening
his first night with a new girl-friend at 3.20.14: longius inprimo, Luna, morare toro.
58 Gow-Page (n. 29), 379.
59 Cf. Ter.Eu. 601-6, wherethe youngman Chaereaswaggeringly
recountshis assaulton a
sleepinggirlownedby a courtesan.