1 Lucretius Death

8/14/2019 1 Lucretius Death

1/27

Lucretius, Symmetry Arguments, and Fearing Death

Author(s): James WarrenSource: Phronesis, Vol. 46, No. 4 (Nov., 2001), pp. 466-491Published by: BRILLStable URL: http://www.jstor.org/stable/4182682

Accessed: 27/10/2008 22:11

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at

http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=bap.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the

scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that

promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

BRILLis collaborating with JSTOR to digitize, preserve and extend access to Phronesis.

http://www.jstor.org
http://www.jstor.org/stable/4182682?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/action/showPublisher?publisherCode=baphttp://www.jstor.org/action/showPublisher?publisherCode=baphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/4182682?origin=JSTOR-pdf


2/27

Lucretius,Symmetry rguments,andfearing death*

JAMES WARREN

ABSTRACTThis paper identifies two possible versions of the Epicurean'Symmetry argu-ment', both of which claim that post mortemnon-existence is relevantlylike pre-natal non-existence and that therefore our attitude to the former should be thesame as that towards the latter. One version addresses the fear of the state ofbeing dead by making it equivalent to the state of not yet being born;the otheraddressesthe prospectivefear of dying by relatingit to our presentretrospectiveattitude to the time before birth. I argue that only the first of these is present inthe relevant sections of Lucretius (DRN 3.832-42, 972-5). Therefore, this argu-ment is not aimed at a prospective fear of death, or a fear of 'mortality'.Thatparticularear is insteadaddressedby theEpicureans hrough he additionalpremise(found in the Letter to Menoeceus 125) that it is irrational o fear in prospectanevent which is known to be painless when present.This still leaves unaddressedthe related fear of 'prematuredeath', which is to be removedthroughthe accep-tance of Epicureanhedonism.

TO .1 yEvETYOal T) OctVEiV i'GOV Xyo.Euripides,Troades 636The Epicureans' arguments against the fear of death have always gener-ated a great deal of discussion and analysis. Here I detail one argumentfound in the third book of Lucretius' De Rerum Natura (DRN) and attemptto show that the exact form in which it appears in Lucretius' text is dis-tinct from a related argument which is sometimes the focus of currentdebate. I hope that from this somewhat negative conclusion some positive

* This paper has benefited considerably from discussion with David Sedley andDominic Scott. I would also like to thankthe participants n a seminar on 'Epicuruson Death and Nature' held in Cambridge n the Lent term 2000 for their reactionstoa preliminarypresentationof these thoughts.A version was presentedto the SouthernAssociation for Ancient Philosophy in September2000, where the audience providedvery helpful comments.

? Koninklijke Brill NV, Leiden, 2001 Phronesis XLV114


3/27

LUCRETIUS, SYMMETRY ARGUMENTS, AND FEARINGDEATH 467resultsmight emerge.First,we shouldask why the alternative ormulationis not tobefound nLucretius,f indeed t is anargumentwhichanEpicureancould have composed. This in turn may shed some light on what theEpicureans nvisagedas the primarypointof attack n their assaulton thefearof death, and theoverallstructureormedby the variouspartsof thatassault.The two relevant passages of Lucretius, n which he invokes a sym-metryof past and futurenon-existence, rethe following.I call themTextsA and B for ease of reference.Text A: Lucr.DRN 3.832-42

et velut anteacto nihil temporesensimus aegri,ad contigendumvenientibusundique Poenis,omnia cum belli trepidoconcussa tumultuhorrida contremueresub altis aetheris auris,in dubioque fuere utrorumad regna cadendumomnibus humanisesset terraquemarique,sic, ubi non erimus, cum corporis atque animaidiscidium fuerit, quibus e sumus uniterapti,scilicet haud nobis quicquam, qui non erimus tum,accidere omnino poterit sensumque movere,non si terramari miscebitur et mare caelo.And just as in the time that went before we felt no pain - when Carthaginianscame from all sides to wage war, and the world struckby the disturbingupheavalof war shook and quiveredunder the high vaults of heaven, and it was unclearto whose kingdom should fall all men on the land and sea - so when we are (lit.'will be') no more, when the body and soul from which combination we areformed have come apart, then no doubt therewill be nothing to us (who will notbe then) which will be able to move our senses in the slightest, not even if earthand sea and sea and sky are mixed together.

Text B: Lucr. DRN 3.972-5respice item quam nil ad nos anteacta vetustastemporis aeterni fuerit, quam nascimur ante.hoc igitur speculum nobis naturafuturitemporis exponit post mortem denique nostram.Look back, then, at how the stretch of unending time before we are born wasnothing to us. Nature, therefore,offers this reflection to us of the future time afterour eventual death.

The firststep in establishing ust what Lucretius'argument onveys is todistinguish wo similarbut different laims.I shallcall themP (for 'past')i and ii.


4/27

468 JAMES WARREN

Pi. Our pre-natalnon-existencewas nothing o us before we wereborn.Pii. Looking back from withina lifetime,our pre-natalnon-existence snothingto us.The firstof these claims nothingabout our presentstateof mind. Insteadit deals with a state of affairs during the period before our birth, andasserts that at that time we felt no pain, loss, or distress.This relies onthe generallyaccepted claim thatsince before our birthwe are not, thatis to say we have not (yet) come into existence, it makes absolutelynosense to considerour relativestate of well- or ill-beingat thattime. Thereis no subjectto consider.The second propositionabove makes a differentclaim. It considersapoint of time within a life and asserts that looking back fromthe presentwe feel presentlyno distressor pain at the thoughtof the timebefore ourbirth.The generalformof what I shall call a 'Symmetryargument' s to takeas a premisean assertion aboutpre-natalnon-existenceand then claimthatthere is no relevantdifferencebetweenthis periodandits 'reflection'in the future,our post mortemnon-existence.This latter s, of course, theperiodwhich must be the objectof any fearof death. If it can be shownthatwe have no reason to take a differentview to post mortem rompre-natalnon-existence,and further hatpre-natalnon-existences universallyand justifiablynot considereda sourceof distress,then it follows that tofeardeathis irrational nd unjustifiable.This generalform holdswhether he majorpremise s propositionPi orPii. However, the force of the argumentand its conclusion are certainlyaffectedby the choice of premise,since the symmetricalrelationuponwhichtheargument elieswill generateconclusionsabout herationalatti-tudeto take towardspost mortemnon-existencewhich aresymmetricalothe starting remises nd herefore istinctnthe samewayasthosepremises.In otherwords,if the premise s Pi, thenthe claimgeneratedby the argu-ment about the futurewill be F (for 'future') :Fi. Ourpost mortemnon-existencewill be nothing o us afterour death.But if Pii is the premise,then the claim generatedaboutthe futurewillbe this:Fii. Lookingforward romwithin a lifetime,ourpost mortem non-exis-tence is nothingto us.


5/27

LUCRETIUS,SYMMETRYARGUMENTS,AND FEARING DEATH 469Notice againthe distinctionbetweenthese claims. Fi deals with theperiodafter our death and asserts that at that time we will feel no pain or dis-tress. This, of course, is the conclusion also generatedby the famousEpicureandictumdistilled into the secondKyria Doxa:

Kyria Doxa 2: o OavcrTo; ov68v irpo; i-ija&; To yap staX)OvEvLvalO-llT, TO 8'cxvaCa&iioiovOvo)V tp?STl,IgO.Death s nothing o us; for what has been dissolvedhas no sensation, nd whathas no sensation s nothing o us.

In Epicureanerms,death is the disruption f the atomiccomplexof bodyand soul which constitutea person,and since no furthersensationcanbe experienced,no pleasureand pain can be experiencedand thereforeno well- or ill-being in Epicureanterms. This is the argument,there-fore, that in orderfor something o count for or againstwell-beingtheremust be a subject of thatwell- or ill-being who standsin an appropriaterelationshipo the supposedgood or bad.'For an Epicureanhis relation-ship is the perceptionof pleasureor pain. In any case, afterdeaththereis no subject at all (just as before birththere was no subject)and there-fore death can be nothing o us. If this is the forceof the 'Symmetryargu-ment',thenit addsnothingsignificant o the majorEpicurean laimaboutdeath. In fact, Fi might appear to make a weaker claim than Kyria Doxa2, since it claims that death will not be bad, whereas Kyria Doxa 2famously says it is nothing to us. (The verb 'to be' is omitted from theversion preserved in Diogenes Laertius, but in that case it is natural tosupply the present tense rather than the future. Lucretius renders it as: nilmors est ad nos at 3.830.) Why Kyria Doxa 2 should do this is a ques-tion to which we will return.For now let it be observedthat the reasonsfor the conclusion given by the Kyria Doxa themselves are sufficient tojustify only the conclusion that after death I will feel no pleasure or pain,and that therefore after death it makes no sense to talk about my well- orill-being.

Fii, however, makes a different claim. It asserts something about ourpresent attitudes as we look forward from the present to a point in timeafter death. It claims that just as when looking back we feel no distressat the thought of pre-natal non-existence, so we should in the present feelsimilarly about post mortem time. This offers a new dimension to theEpicurean discussion of the fear of death. Whereas Kyria Doxa 2 and Fidealt merely with the state of affairs after death, this argument deals

I Lucr.DRN 3.863-4.


6/27

470 JAMES WARREN

directlywith ourpresentattitudesduringourlifetime.It reflectson to thefuturethe retrospective ttitudewe supposedlyhave now about the timebeforebirth,asking us to take a symmetricalprospectiveattitude.So Piiand Fii essentiallyconcernourpresentattitudes,while Pi andFi concernwhat was the case beforebirthand will be the case afterdeath.Let mecall the versionof the argumentwhich startswith Pi and generatesFi,version 1, andthe versionwhichstartswithPii andgeneratesFii, version2. I set themout here:Version 1:Pi. Ourpre-natal on-existencewas nothing o us beforewe wereborn.SYM Pre-natal on-existences relevantlyikepostmortem on-existence.Fi. Ourpost mortemnon-existencewill be nothingto us after ourdeath.Version 2:Pii. Lookingback fromwithin a lifetime, our pre-natalnon-existenceis nothingto us.SYM Pre-natal on-existencesrelevantlyikepostmortem on-existence.Fii. Looking orward romwithina lifetime,ourpost mortemnon-exis-tence is nothingto us.

IIThereare at least two reasonswhy we mightpreferLucretius o be giv-ing the secondversionof the Symmetryargument.First,as I have alreadysuggested,version 1 concludeswith little morethan a restatement f theoriginalassertion hatdeathis annihilation ndtherefore ince afterdeaththereis no subject,deathcannotbe a harm.2Second,a criticism s some-times evelledattheEpicureanshat heyought ohaveproducednargument

2 Mitsis 1988, 306 n. 6 appears to disagree with Furley's 1986 assessment of theargument see below n. 5), but in his own descriptionof the argumentat 306 expressesit clearly in the form Pi/Fi: 'We felt nothing in the time before we were born; ust so,we will feel nothing when we are dead'. His general attitude throughout he article,however, is thatthis argument s part of an Epicureanconcern to addressworries aboutthe duration of a life - and presumablythis concern must be part of a prospectivevision of one's lifetime from the point of view of the present. Sorabji 1983, 176 isuncertain whether Lucretiusintends the 'more interesting'argumentfrom Pii to Fii.Kamm 1993, 25 also offers version 2 as 'Lucretius'argument,distinct from 'Epicurus'argument n Kyria Doxa 2: 'Lucretiusrecognises that we are not disturbedmuch aboutthe fact of our non-existencepriorto our creation.If so, he asked, why are we so dis-turbedabout our non-existenceafter death?' (my emphasis).


7/27

LUCRETIUS,SYMMETRYARGUMENTS, AND FEARING DEATH 471alongthe lines of the secondform of the Symmetryargument.KyriaDoxa2, it is claimed,is fine as far as it goes, but it fails to address he primarysense in which peoplefear death, namely the prospective ear that in thefuturethey will die. Whetheror not afterdeaththere will be a subjecttoexperiencepleasureor paindoes not matter.Whatdoes matter s what canbe called the fear of mortality,or alternatively the prospective ear ofdeath'.This fear can be distinguished romthe fearof death addressedbyKyriaDoxa 2, since it is entirely conceivable that someone mightwellagree that 'being dead' is neither pleasant nor painful,but neverthelessfeel distressas he looks forwardfrom some point in his life to a timewhen that life will cease.3 A critical reading of Kyria Doxa 2 wouldemphasise hiserror,andclaimthat Epicuruss wrongto stateas his con-clusionthat 'Deathis nothingto us'. He is entitledto concludeonly thatdeath will be nothingto us - preciselywhen we are dead. He has saidnothingso far to counterthe prospective ear that my life will cease atsome future time. The second form of the 'Symmetry argument', ofcourse,does address his second type of fear, since it is concernedwith apresentprospective attitudeto futurenon-existence.4 n that case, in aspiritof charitywe shouldperhapshink hat heoccurrences f theSymmetryargumentn Lucretiusprovide he extraandotherwisemissingelementinthe Epicureans'armouryagainst the fear of death.Enough,then,of the alternatives.Whichof them is to be foundin ourtexts? Unfortunate s this might be for the assessmentof the efficacyofLucretius'therapy, he two texts from the DRN tend to favourthe firstversion of the Symmetryargument namely Pi-+Fi).At least, text A cer-tainlyoffers this argumentand text B probablydoes.That text A offers little beyond the familiar claim that death removesa possible subjectfor harm has been pointedout before, but should berestated.5 he tenses of the verbswhichLucretius mploys arequiteclear.

This distinction is arrivedat by the charactersA. and M. in Cicero's firstTusculanDisputation. At Tusc. 1.14, A. asks: age iam concedo non esse miseros qui mortuisintquoniam extorsisti ut faterer, qui omnino non essent, eos ne miseros quidem esseposse. quid? qui vivimus, cum moriendumsit, nonne miseri sumus?quae enim iucun-ditas, cum dies et noctes cogitandumsit iam iamque esse moriendum?4 Note, however, that this second 'Symmetry argument',although it addresses the

prospective fear of death, still does not address the furtherfear of premature death.These two are well distinguished by Striker 1988.5 Furley 1986, 76: 'The tenses of the verbs are conclusive about this; there is nostatementat all about our present attitudes'. Rosenbaum 1989, 358 takes issue withFurley, althoughhe recognises that Furley is certainly correctabout the tenses of the


8/27

472 JAMES WARRENFirst, Lucretius establishes Pi: anteacto tempore nil sensimus aegri.(3.832). The verb is clearlyaorist. Lucretiusmightstill be accused hereof beingloose in his expression n this line, though.We might be temptedto think thatnil sensimus mplies that we were there and had the poten-tial to feel pain, but did not. But this cannotbe Lucretius'claim, sincethe point he wishes to convey is of course not that we were presentatthat time and felt somethingotherthanpain, but thatwe were not thereat all and therefore could not experience anything,a point he makesexplicitlya little later(3.863-4). In any case, the periodof time in ques-tion is clearly pre-natal:we did not sufferany painthen.Nothingis saidaboutthe present,norabout our presentattitude o the past.Text A thengoes on to invoke a symmetrybetween this period and the time postmortem sic 838), and the symmetrical laim it producesabout the futureis clearly Fi, not Fii.

sic, ubi non erimus . ./scilicet haud nobis quicquam,qui non erimustum, accidereomninopoterit sensumque movere ... 3.838-41... just so when we are (lit. 'will be') no more,. .. then no doubt there will benothing to us (who will not be then) which will be able to move our senses inthe slightest,...

Lucretius s describinga state of affairsat a time after death(nonerimustum;haudpoterit), not our presentprospectiveattitude o our mortality.Text A, therefore,gives what I called version 1 of the Symmetryargu-ment and confirmsKyriaDoxa 2. It does not addressa prospective earof death.6Text B is more complicatedand its conclusions less clear, but ulti-mately it offers little if any more support or proponents f version 2 ofthe Symmetryargument Pii--Fii). I repeatthe Latin here (Lucr.DRN3.972-5).

verbs in the relevantLucretian exts. Nevertheless,he argues that Epicurus s certainlyconcerned with present attitudes since, for example, KyriaDoxa 2 holds that 'death isnothingto us'. Rosenbaumgoes on to offer a version of the SymmetryArgument 359-60) which has as its first premise: 'No one fears the time before which one existed'.and concludes 'Therefore it is not reasonable now for one to fear one's future non-existence'. These are clearly versions of Pii and Fii. If Rosenbaum is right thatEpicurus was concernedwith presentattitudes, he is nevertheless wrong in claimingthat this is the particularargument by which he set out to alleviate them.6 For my reading of the 'palingenesis' argument which immediately follows thisargument n DRN 3, see Warren 2001.


9/27

LUCRETIUS,SYMMETRYARGUMENTS, AND FEARINGDEATH 473respice item quam nil ad nos anteactavetustastemporis aeterni fuerit, quam nascimur ante.hoc igitur speculum nobis naturafuturitemporis exponit post mortemdenique nostram.

Initially we might think that this does indeed describe a present attitudeto pre-natal time. At first glance, the metaphor of viewing sustained inthese lines might be thought to invite just this interpretation. After all, inthe first line of the argument we are invited to 'look back' towards thepast, and presumably the only point from which we may currently lookback to the past is the present. Further, the metaphor of a mirror whichthese lines introduce might also point in this direction. The image is of aviewer who looks back in time at a mirror in front of him.7 The mirrordisplays an image of what is behind the viewer, namely the future. So asthe viewer looks back to his past non-existence, he is in fact seeing beforehim a mirror image of what is to come.8 And since nothing in the imagecauses any distress (976-7), the future presents nothing to fear.A proponent of version 2 of the Symmetry argument must place a lotof weight on this metaphor,and must insist that it implies a viewer locatedin the present looking at a mirror image of what he could see clearly ifhe looked over his shoulder. On this understandingthe viewer is consid-ering the relationshipbetween himself located in the present and the objectin the mirror- a reflection of post mortem time. Hence the argument isthought to concern our present prospective attitude to death. This pictureoffers the strongest indication that Lucretius is concerned with our present

7 The image of someone looking backwards n time conforms to the ancient imageof us 'backing into the future'. We can 'see' (i.e. remember) the past but the futureis not visible. Comparethe use of the Greek ontice to mean 'hereafter'(LSJ s.v. II).Also see Kenney 1971 ad 974. He insists thatspeculumexponit should be understoodas 'shows a reflection', not 'holds up a mirror', since this allows hoc (974) to referto anteacta vetustas temporisaeterni. '[T]hat is what Nature shows us as the mirror-image of futurum tempus, and it is, of course, a blank, a reflection of nothing....Naturally there is nothing horribile or triste (976) to be seen, because there is noth-ing at all to be seen.' This last remark is an overstatement.Of course, Lucretius isnot denying that nothing at all happenedbefore our birth, merely that whatever didhappen, it caused us no pain.8 Lucretiusexplains in his discussion of mirrorsat 4.269ff. why the image in themirrorappears to be twice as far away as the viewer is from the mirroritself. Weneed not think that in this passage in book 3 Lucretius wishes to imply that the sub-ject is at a temporal point half-way through his life, equidistantbetween birth anddeath, although his vision in the mirrormakes it look as though he is.


10/27

474 JAMES WARRENattitudes o past and futurenon-existence,rather hanthe stateof affairsat some time beforeor afterour lifetime.9However,despitethe intuitiveattraction f such an understanding,hetenses of the verbs used here suggest that in fact - and despite themetaphorof viewing- Lucretius s againofferingversion 1 of the Sym-metryargument.Above all, the tense of fuerit (973) suggeststhatthis isthe case: 'pre-natal ime was nothingto us'. This is anotherexpressionof premisePi of version 1 of the Symmetryargument.Admittedly, hisphrasemightalso carrya perfectsense: 'pre-natal imehas been nothingto us'. In that case it mightbe understoodas a form of premisePii ofversion2 of the Symmetryargument, f it is understoodhat 'has been'here meansin effect: 'pre-natal ime has been [sc. duringour life, up tothe present]nothing o us'. But if Lucretiuswished to makeclearthatheis talkingabout t presentlybeing the case that our pre-natalnon-existencecauses us no distress,then he could have done so easily by writingthepresenttense of the verb:sit.' This would have produced he followingtranslation,whichwould convey the pointwithoutambiguity:

Lookbackat how all theimmense mount f time beforewe wereborn s noth-ing to us.Whereas,withfuerit, the translations morenaturally:

Look backat how all the immenseamountof time beforewe werebornwasnothing o us.Both these version construe quam nascimur ante as amplifying ante-acta vetustas in the previous line - and that is how most current transla-tions construe the phrase." This construal is compatible with both versionsof the Symmetry argument. However, it is also possible, I think, to takequam nascimur ante as an adverbial phrase qualifying fuerit, in whichcase the translationwould read:

9 Feldman 1990, 23 points to this passage in order to producean argumentbegin-ning from the premise: 'The fact that he didn't exist for an infinitely long period oftime priorto his birth is not bad for anyone' (a species of Pii)."II assume that, although sit may be less easily incorporated nto a hexameter,Lucretius would not modify an argument merely metri causa. nascimur is, of course,present tense, but this offers no supportfor version2. Kenney 1971 ad loc.: 'The tenseshows that nascimurrefers to all successive generations,not merely this one'." This is certainly the case for translations nto English. Latham's 1951 Penguintranslationgives: 'Look back at the eternity that passed before we were born, and


11/27

LUCRETIUS,SYMMETRYARGUMENTS, AND FEARING DEATH 475Look backat how all the.previousmmenseamountof time was nothing o usbeforewe wereborn.

This is incompatible with a version 2 understandingof the argument,sincein this case it is made clear that the time being considered at which pre-natal non-existence caused no harm is precisely the time before birth. Onthis construal Lucretius gives a clear version of proposition Pi. Of course,I need not restmy case on any controversialunderstandingon quamnascimurante. My preferredversion 1 reading is in any case strongly supported bythe tense of fuerit.It is also possible to understand the metaphor of looking into a mirror ina way which is consistent with version 1 of the Symmetry argument.Rather than focusing on the relationship between the viewer, located inthe present,and the period of time at which he is looking (namely a reflectionof future time), we should instead focus solely on the image in the mir-ror. Version 1 of the Symmetry argument is not concerned with presentattitudes to past and future non-existence. Rather it is concerned with thepast and future absence of ills. So we could understand Lucretius to beasking us simply to look at the picture in the mirror. Is there anything tofear in it? No. But it is merely a reflection of how things will be afterdeath. So since we are happy to accept that at that time before birth therewas nothing which could cause any pain, we must agree that there will benothing which can cause pain after death. Therefore, although Lucretius'striking image of the mirror perhaps tempts us towards version 2 of theSymmetry argument, it does not require us to think along those lines.'2Neither of the two texts which might be offered by proponents of ver-sion 2 of the Symmetry argument can sustain such a reading. Text A cer-tainly cannot, and Text B certainly does not in any unambiguous way. I

mark how utterly it counts to us as nothing.' Smith's 1975 revised Loeb has: 'Lookback also and see how the ages of everlasting time past before we were born havebeen to us nothing.' Bailey (Oxford, 1947): 'Look back again to see how the past agesof everlasting time, before we are born, have been as nought to us.' Brown (War-minster, 1997): 'Look back in turn and see how the eternity of everlasting time thatelapsed of old before our birth was absolutely nothing to us.' Melville (Oxford, 1997):'Look back upon the ages of time past/Eternal,before we were born, and see/Thatthey have been nothing to us, nothing at all.'12 Bailey 1947 ad 974 comments that this section is merely reinforcing what wassaid in 3.832ff., and by doing so roundsoff a particular et of arguments n this book,before Lucretiusgoes on to offer his thoughts on the genesis of particularmyths aboutpunishments n the Underworld.


12/27


13/27

LUCRETIUS, SYMMETRYARGUMENTS, AND FEARING DEATH 477

Seneca uses similararguments n a numberof occasions,most of whichalso conform o my version1 of the Symmetryargument.'7 here s, how-ever,a relatedpassage n EpistulaMoralis77.11 which offers a ratherdif-ferent perspectiveon these issues.nonne tibi videtur stultissimus omnium qui flevit quod ante annos mille non vix-erat? aeque stultus est qui flet quod post annos mille non vivet. haec paria sunt:non eris nec fuisti; utrumque empus alienum est.Would you not thinkhim an utter fool who wept because he was not alive a thou-sand years ago? And is he not just as much a fool who weeps because he willnot be alive a thousandyearsfromnow? It is all the same; you will not be andyouwere not. Neither of these periods of time belongs to you. (trans. R.M. Gummere).

Ratherthan a single subject or addresseeas in the Lucretian exts, twocharactersare described,one concernedwith past non-existenceand onewith future.The tenses of the verbs used also complicate he issue, sinceSeneca does not maintaina strictequivalencebetween the two charactershe is comparing.Thefirstwept (flevit) becausehe hadnotbeenalive (vix-erat) a thousand ears previously;'8he secondsweeps now (flet) becausehe will not be alive a thousandyears in the future.This suggests that inthis passage at least Seneca is indeedinterested n retro-and prospectiveattitudes,but not so muchin a strictsymmetry etween past- andfuture-directedattitudes. t is thereforedifficult o tell whetherSeneca is indeedmakinga specific claimaboutthe exactrelationship etweenretrospectiveand prospectiveattitudes.19Whatis more evident is that Seneca intendshere to portray wo absurd examples of regret. It is important hatthesetwo characters reweepingover the thoughtof not experiencing hingsinthe very distantpast and the far future.Even those who wish to acceptthatdeath can be a harm since it robsus of goods might shrink from the claim that death robs us of goods" De Cons. ad Marc. 19.5: mors dolorumomnium exsolutio est et finis, ultraquemmala nostra non exeunt, quae nos in illam tranquillitatem,n qua antequamnascere-mur iacuimus, reponit. si mortuorumaliquis misereturet non natorummisereatur.Ep.Mor. 54.4: si quid in hac re [sc. morte] tormentiest, necesse est et fuisse antequamprodiremus in lucem; atqui nullam sensimus tunc vexationem. Ep. Mor. 65.24: necdesinere timeo (idem est enim quod non coepisse).18 Madvig emended to: flebit quod ante annos mille non vixerit. vixerit appears in

one of the codices.'9 Cf. Rosenbaum 1989, 357: 'For Seneca, the thought is clearly directed againstnegative feelings about future nonexistence.' This may be so, but it remains to beexplained exactly how the example of an absurdretrospective attitudeis meant to beinstructive.


14/27

478 JAMES WARREN

located a thousandyears in the future.(This extremepositionis a possi-bility, however. In Cicero's Tusculans 1.9,A. begins with the claim thatboththe deadandthosewho aregoing to die aremiseri. It is pointedoutthat this would lead to the radicalconclusion hat everyone s always sub-ject to eternalmisery.)A more moderatepositionwill concedethat deathis only a harm insofar as it robs us of time and goods we could reason-ably have been expected to experience.This trades on the idea that wecan reasonably xpectto live a full life of perhaps ighty yearsor so, andso dying before this time is up robs us of something o whichwe shouldfeel entitled.Sucha counterfactual ccountof theharmof deathobviouslyloses its force in the far future,since we cannot reasonablyhope to livefor a thousand ears. Whileit mightbe right for a twenty-year-oldo feelregret f he is assuredhe will die before the age of thirty, t is less plau-sible to claim that a Roman senatorof the first centuryAD should havefelt aggrievedat the thought hat he would not participaten the millen-nialcelebrationsn31 December 999.20There s considerableoombetweenaccepting hat Seneca'sexamplesof regretare absurdandthe conclusionthatit is absurd o weep at the thoughtof non-existencehowevernear orfar in the past or futurethat non-existencemay be.2'Seneca does close this gap, and the crucialstep in his argument omesin the finalphraseof the sectionjust cited:utrumque empusalienumest.On this basis Seneca groundshis assertion hatsuchfeelingsaboutfuturenon-existenceare not justifiable.This Stoic premise s rather tronger hanthe Epicurean laim that before birthand after deathwe 'are not', sincenot only do the Stoics think thatbeforebirthand afterdeathwe are not,but they also are adamant hatthattime does not belongto us (since ourtime of birthanddeathare fated),andtherefore ts absencefrom our life-time cannotin any way be a loss to us.22Those times were never ourstolose. Seneca uses the Stoic notionof fate to underlinehis point(77.12).

20 Cf. Striker 1988, 327. Lucretius himself emphasises that the period of postmortem non-existence will extend indefinitely far into the future, no matter when weactually die (3.1073-5).21 Malcolm Schofield suggested to me that it might be possible to performa sorites-like argumentbeginning with the acceptance that it is absurd to weep at far-futurenon-existence and ending with the conclusion that it is equally absurd to weep at any

future non-existence. However, there is no suggestion that Seneca has such and addi-tional move in mind.22 These two periods of time are aliena. In some Latin Stoic texts this worddescribesobjects which are the object of the opposite of oikei6sis (e.g. Cic Fini. 3.16:alienari).


15/27

LUCRETIUS, SYMMETRYARGUMENTS, AND FEARING DEATH 479in hoc punctum coniectus es, quod ut extendas, quousqueextendes? quid fles?quid optas? perdis operam. desine fata deum flecti sperareprecando = Virg. Aen.6.376). rata et fixa sunt et magna et aeterna necessitate ducuntur:eo ibis quoomnia eunt.You have been cast upon this point of time; if you would make it longer howmuch longer shall you make it? Why weep? Why pray? You are taking pains tono purpose. 'Give over thinking that your prayerscan bend/Divine decrees fromtheir predestinedend'. These decrees are unalterableand fixed; they are governedby a mighty and everlasting compulsion.Your goal will be the goal of all things.(trans.R.M. Gummere)

This additional move provides the missing extra step. Not only is it absurdto weep at non-existence n the far future, t is equallyabsurd o feel theloss of any lifetimewhich one does not in fact enjoy. Given Stoic de-terminism, Seneca can head off any residual possibility that death atsome particular ime mightbe thoughta harmby depriving he personoftime he mighthave expectedto live.23The period immediately ollowingsomeone's death is no more theirs to be deprivedof than some time athousandyears into the future.Given that the time of one's deathis pre-destined,death cannot deprive one of anything not even a second.The next example comes from Plutarch'sConsolatio ad Apollonium.At107D Plutarchoutlines three possibilities for the nature of death, andattributes he identificationof these possibilities to Socrates. Death iseither like a deep anddreamlesssleep, or it is like a journey,or it is theextinctionof bodyandsoul.Eachpossibility s then examined n turn,andit is shown that on each account death is not to be feared. The first twopossibilitiesare familiarfrom Socrates'final address n Plato's Apology(40cSff.), and the third - the destruction of body and soul - may simplybe an expansion romthe idea of deathas a completelack of perception,which is of course the major orce of the analogy with a dreamlesssleep.The Apology s unconcerned o make a distinctionbetween death as anni-hilation and death as the absence of sensation.24Nevertheless, Plutarch

23 Such arguments hat death is an evil tend to trade on the idea thatwhile the timeof a subject's birthmust be fixed (since date of birth is essential to thatsubject's iden-tity), the time of deathis contingent and thereforea counterfactualaccount of the harmof death is possible. If in additionthe time of death is agreed to be necessary (as inStoicism), then this possibility is removed. See Sedley 1993, 316-8.24 PI. Apol. 40c6: i] yapobovptieV Etval j.& a'*ffiYNvgir8egictv Tii6rVO;E`XetV tOvteOve&na,... Of course, the Phaedo characterisesdeath as a 8takoa$ or &nackkauof body and soul (e.g. 64c4-8, 88alO-b2) and as a release from care and pain, but itmost certainly does not allow that death involves the (pOop&f the soul.


16/27

480 JAMES WARRENtreats the two as distinct positions and in his discussion of death as anni-hilation uses arguments familiar from Epicurean contexts.

yt YETv 6V OOaivaTo;rexia Ti; Otrt(pOopa Kaia't&aiuat; To tE awgaTo; xia TTI;xVi (TOrTptTovyap iv TOlTOTX1FO3KP(TaUKTilcaaic;), OVOS JT&) aKoVCaTIVv

avaloolletia yap rt; KaT axwrov ytyveTat Cic racm; aiakkayii Xuiirii;KQIpvr8o0;.? nEp yap oiuS-'ayaO6v iiV E`rEaTIV oir; oO6 KaKoV- REpi ya p TO OV

KiLI rO V?EGT9rCO4 KaOoatitp TOayaOov nbpuIce ytiyv0at, rOVarTOV TpO6iOVKai TOKaKOv *p &p r6 Ti OVsv&X' pg?vov ?ic rTv `vrcov oiV&Tepov To&rcv V6pXE1.Er; riv a&rv o0v ritv oi tevXeiyrifjavre; KarcxVat n' rp6 ri; YEV?aE;. @&irp^~~~~ a0,\^,^ s_ , , , , , ,OuVOAV'jV I iV rpOTnj;YEVEa%Ee O0VT'&yto0v OVt?EKCO6V, o01T oi1) -Tx tvT?x?VOV.K4iEviI Cp t0 ipo TOV)VO V v pO;1a, outO o)6E Ta ?0 aTIv

OD?V ?OTat npO5 gzt;-

If, however, death is a complete destruction and dissolution of both body andsoul (for this was the third of Socrates' conjectures) even so it is not an evil.For, according to him, there ensues a sort of insensibility and a liberation fromall pain and anxiety. For just as no good can attach to us in such a state, so alsocan no evil; for just as the good, from its nature,can exist only in the case ofthat which is and has substantiality,so it is also with the evil. But in the caseof that which is not, but has been removed from the sphere of being, neitherofthem can have any real existence. Now those who have died returnto the samestate in which they were before birth; therefore,as nothing was either good orevil for us before birth, even so will it be with us after death. And just as allevents before our lifetime were nothingto us, even so will all events subsequentto our lifetime be nothing to us. (trans.F.C. Babbit)

The section in italics contains Plutarch's version of the 'Symmetry argu-ment', and this again is cast in terms of version 1, not in terms of retro-and prospective attitudes.We can now add to the conclusion that there is no sign in Lucretius of aclear statement of version 2 of the Symmetry argument.We have no clearexample in any other ancient source of version 2 of the Symmetry argu-ment - of an argument concerning the symmetry of prospective and ret-rospective attitudes to past and future non-existence. Given the likelihoodthat even the texts written by non-Epicureans may have turned toEpicurean sources to construct their arguments against the fear of death,it therefore seems very unlikely that the Epicureans themselves everoffered any such argument stressing symmetrical prospective and retro-spective attitudes. In that case we can restate and stress the criticism thatEpicureans omitted to give an argument against the most debilitatingspecies of fearing death - the prospective fear of death.


17/27

LUCRETIUS, SYMMETRY ARGUMENTS,AND FEARING DEATH 481

IVAll is not lost so far as the Epicureansare concerned,however, sincealthough hereis an important istinctionbetweenthe conclusionsI havelabelledFi and Fii, they are neverthelessrelated.Indeed,Fi can be usedas a premisein a secondaryargumentwhich then producesFii, so longas a furtherpremiseis inserted. And that premise,while missing fromLucretius, s prominentn Epicurus'brief summaryof his ethical teach-ings, the Letterto Menoeceus.This will allow the Epicureanso offeranargumentagainstthe prospective ear of death. Here again are the twoclaims.Fi: Ourpost mortemnon-existencewill be nothing o us after ourdeath.Fii: Lookingforward rom within a lifetime, ourpost mortemnon-exis-tence is nothingto us.Fii can be derivedfromFi by using the following principle:

Whatevercauses no pain when present, causes only empty distress when antici-pated.25

This principlecan be extracted rom a section of Epicurus' etterwhichexplicitly deducesthat a fear of future non-existence s irrational ince,when it comes, death is annihilation. t is worthquoting this in full.(00TE VXatoa; o XYv 686tiVat TOv Oavaxov o0% O"irt i.nriYoEl iap6v, &XX'6it k-xnti?XX(OV. O yap napov OvaVxO?, 1XOpOK(LEVOV EVcO; VXXet.Letter to Menoeceus 12526

25 Cf. Cic. Tusc. 1.16: M.: quia, quoniam post mortem mali nihil est, ne mors qui-dem est malum, cui proximum tempus est post mortem, in quo mali nihil esse con-cedis: ita ne moriendumquidem esse malum est: id est enim, perveniendumesse adid, quod non esse malum confitemur.26 The combination KevoS - xratio; recurs in Epicurus' works. In the letter toAnaxarchus, quoted in Plutarch Adv. Col. 1117A, Epicurus uses both adjectives todescribe the virtues, which 'filling us with hope of rewards, are empty (KEv6;) andpointless (juat'cxa;)and bring us trouble'. The force of kenos seems to be 'empty' inthe sense of irrational,or without justification- as in the Epicurean term kenodoxia,used of ill-founded beliefs which tend to cause suffering. The sense of waraioq is'foolish' or 'vain' and when used of personsgenerally means that what they are doing

is futile. CompareLetter to Menoeceus 127 where an imagined objector claims that itis bad to be born, but, when one is born, it is best to die as quickly as possible.Epicurus claims that if this is said with conviction, the person should immediatelycommit suicide. If it is said in mockery, then the person is waraiogsince no one wouldbelieve him.


18/27


19/27

LUCRETIUS, SYMMETRY ARGUMENTS,AND FEARINGDEATH 483Therefore: Since death causes no pain when present, it causes onlyempty distress when anticipated and empty distress is noreal distress at all).Fii: Looking forward from within a lifetime, our post mortemnon-existence s nothingto us.A symmetricalargumentcould, if required,be constructed o derive Piifrom Pi, by using the symmetricalprinciple: Whatevercauses no painwhen present,causes only empty painin retrospect'.Nowhere s this prin-ciple evoked, since Pii is not controversial s is Fii. Pi is also uncontro-versial. No one has arguedseriouslythat pre-natalnon-existencewas anevil, although his thesishas been advancedas an absurdconclusionto areverse form of the Symmetryargument.This reverse argument akes asa premisethe claim that 'everyonefears futurenon-existence',and usesthe symmetryof pastand futuresustainedby bothversion I and2 of theoriginal argument o concludethat it is reasonableto regretone's pastnon-existence.28So far,I hopeto have shownthatEpicurusdid providean argumentwhichconcludes hatourpresent ear of futurenon-existence s irrational, uthedid so through his addition to his original claim of Kyria Doxa 2, notthrough he 'Symmetryargument',at least not in the form presentedbyLucretiusand other ancient authors.Before I turnto discuss this new Epicureanargument namely Fi orKyria Doxa 2 and Letter to Menoeceus 125), let me conclude my discus-sion of the Symmetryarguments hemselves.Those who object to Fi orFii sometimeshappily grantPi andPii. By doing so these objectorsdenythe very symmetryof past and futureupon which both versionsof theargument ely. Thereare two reasonswhich are generally given to sup-portthis denial. They are:

2h Seneca Cons. ad Marc. 19.5 does offer the argumentas a rhetoricalsupportforthe claim that fearing being dead is absurd:si mortuorumaliquis miseretur et nonnatorummisereatur. See also Rosenbaum 1989, 368-71, who thinks that a possibledefence is available in Seneca's version of the argumentat Ep. Mor. 77.11. This beginsnot with the simple claim that one does not regretone's past non-existence, but thatit is not reasonable to do so. The relevant 'backfire' argument would then thereforehave to begin by showing that it is reasonable to fear futurenon-existence. It is per-haps worth noting that in the palingenesis argument of 3.842-62 Lucretius also (at852-8) uses the observation that in fact we fell at present no concern for past selvesto bolster his claim that we should feel no concern for futureselves. For my readingof this argumentsee Warren 2001.


20/27

484 JAMES WARREN

a. The time of birth s essential to personal dentity,whereasthe timeofdeathis merelycontingent i.e. I couldnot havebeenbornearlier hanI was, butcould die laterthanI will).29b. Withina lifetimeourattitudesopastand uture xperiences re nevitablyasymmetrical.30The vastmajorityof the present iterature evotedto thistopictakesthesetwo positions as startingpoints,and while muchof interestand use hasbeen producedwhich involves discussions of personal dentity,and therationality of future-biasedreasoning, the discussion has driftedawayfrom heoriginal tructuref theEpicureanrgument.ndeed,f theEpicureanargumentdoes not invokepresentattitudes o pastand futureexperiences,then the strandof objectionencompassedby b. cannotstrikedirectlyatthe heartof the Epicureanposition,since thatobjection s clearlyconsid-eringthe attitudesof a subjectfrom a 'temporallyocatedperspective' othingswhich he did experienceand will experience.3'Further,whateverconclusionsaredrawnaboutourgeneralattitudes o pastandfutureexpe-riences, it remains to be seen how these are relevantto the questionofwhether t is rational o take a particular iew aboutthingswhich cannotbe experienced,whetherthey are in the past or the future.Much of thediscussionof this question s centredon the rationality r otherwiseof anapparentntuitivepreference or pains to be in the past and pleasures obe in the future,but the Symmetryargument,on eitherversion,is con-sideringnot two experiences,but two absences of experience.We displayequanimitywith regardto pre-natalevents not because, althoughtheywere distressingat the time, they are now in the past.They simplywerenot ever distressing o us becausethey happenedwhen we did not (yet)exist. So, for example,the conclusionthat we have a 'futurebias' whenit comes to our pleasures s not of directrelevance.32

29 The 'classic' statementof this position is offered(althoughtentatively) by Nagel1979, 8. More recently see Glannon 1994, 240-1, Williams 1995, Brueckner andFischer 1998.30 The 'classic' statementof this position is by Parfit1984, 165ff. Most recentlyseeBruecknerand Fischer 1993, Kaufman 1999.31 Fischer 1993, 26.32 The cases considered by Parfit 1984, 165-7 by which he generates an intuitive

'future bias' are different from the case of death insofar as they are cases in whichpain has been or will be experienced,althoughthe memoryof this pain is subsequentlyremoved by induced amnesia. In this case, it would appear, Epicurus would happilyconcede that such futurepainful experiences might be feared. Death, however, is notpainful when it is present.


21/27


22/27

486 JAMES WARREN

argumentwhich does deny this symmetry s presumablyoffered on themistakenunderstandinghat the Epicureans re offeringversion 2 of theargumentas their only available attack on the prospective ear of death.Perhaps t might thereforebe asked at this point just why Lucretiuspro-vided a Symmetryargument, f it produced conclusionno different romthat of KyriaDoxa 2, and if in doing so it has to invoke the principleofthe symmetryof past and futurenon-existence which would need to bedefendedagainsta numberof strong objections.In the structure f theargumentagainst the fear of death, version 1 of the Symmetryargumentproducesno new conclusions.However, it does producethe same con-clusion as Kyria Doxa 2 on differentgrounds,namelyon the basis whatwas the case duringpre-natal ime, and thereforeperhaps hould be seenas playinga role in Lucretius'attempts o persuadeus of those conclu-sions. It points to a particular iew we have of the time beforewe wereborn which is thoughtto be sufficiently obust as to be unquestionable,and asks us to view the futurestate of non-eistence n the sameway. Bydoing so, it is providing urtherevidence for someone who is tryingto'accustomhimself' (as Epicurus ells us we must)to the thoughtof KyriaDoxa 2. The Symmetryargument laysa confirmatoryndpersuasive olein the Epicureans'overallprojectrather hanseekingto establish a newand independent onclusion.

VIt remainsto be observedthat most recent discussions of the Symmetryargument end not to tackle explicitly the principleevoked in Letter toMenoeceus125, which, if my interpretations correct,must stand as animportantpart of Epicurus' argument against the prospectivefear ofdeath.36The principlemust surely be true in the sense that if someone

36 Furley(1986) 76 is again a notable exception. Nussbaum 1994, 202 and n. 9, 203and n. 11 rightly sees that this is the crucial premise for Epicurus'argument,althoughshe seems to agree with Mitsis that my text B deals with present attitudesto death,whereas text A does not. Cf. Alberti 1990, 170-1. Rosenbaum 1989, 370 entertainsthe idea that this might be the basis for Epicurean argumentsagainst fearing death,but later rejects it since then (371) 'the symmetryargumentwould not really be theingenious, novel contribution o Epicureanthanatology which many have thought...[It] would at bottom rely logically on already well-established and defensible Epi-curean ideals and would quite simply be superfluous,except perhapsas a rhetoricalflourish'. This begs the question.Cockburn 1997, 138 also makes this principlea cen-tral partof Lucretius' argument,but goes on to object to it, claiming that '[t]he sense


23/27

LUCRETIUS, SYMMETRYARGUMENTS,AND FEARING DEATH 487truly thinks that when some state of affairs comes about, it will not causeany pain, it is indeed irrational nevertheless to fear in anticipation beingin that state. If this person persists in fearing going to the dentist, say,then this surely must be because he thinks that when he is in the dentist'schair he will experience pain. If he knew that there would be no pain,then it does not seem rational to fear in anticipation painless dental work.This allows, of course, that one might mistakenly believe that some futurestate of affairs will be painful and therefore fear it in anticipation, but thissimply underlines how crucial it is for Epicurus to convince us of the truthof Kyria Doxa 2.By bridging the distance between fearing being dead and fearing futurenon-existence in the way I have suggested here, Epicurus can try to forcethe discussion back to his central claim that the state of being dead is notpainful in any way (and is therefore not an evil). Perhaps this explainswhat might otherwise seem a strange omission on Lucretius' part, namelythe lack of the further argument against the present fear of future non-existence found in Letter to Menoeceus 125. If Epicurus saw this prospec-tive fear as derivative from the fear of the state of being dead, and wasconvinced that it was on this latter fear that his attention should be con-centrated, then Lucretius may well also have been so convinced that hewent so far as to omit the brief argument retained in Letter to Menoeceus125, thinking that it was merely an optional supplementary step. All hisattentionwas directed,even in the supposed 'Symmetry arguments', towardsestablishing the conclusion of Kyria Doxa 2: nil mors est ad nos nec per-tinet hilum,lquandoquidem natura animi mortalis habetur (3.830-1). Itmight also be suggested that Kyria Doxa 2 itself contains an implicit useof the principle invoked in Letter to Menoeceus 125. I noted above thata critical reading would object to Epicurus' conclusion that death is noth-ing to us, since what follows only shows that death will be nothing to us.A more charitable reading would invoke the idea contained in Letter toMenoeceus 125 as understood within the argumentation of Kyria Doxa 2

in which physical pain is an evil is revealed as fundamentally in our fear of futurepain as in our aversion to present pain ... Similarly, the sense in which death isthoughtof as an evil is revealed in a fundamental way in a person's fear of her owndeath.'Thisobjection s related o Cockbum'sgeneralconcernabouta particular pproachto such questions, an approach which Epicurus seems to endorse (140): '[I]f we donot think of the transformation rom one tense to another as being a common opera-tor on a range of differentpropositionalcores we will find no incongruity in the ideathat something which is a matter of complete indifferencewhen it lies in the pastshould be a matter of deep concernwhen it lies in the future.'


24/27

488 JAMES WARRENto generatea conclusion about how we presentlyshouldview the periodof our futurenon-existence.Again, however, t shouldbe emphasised ust what Epicurushas shownso far - and what he has not. The combination f Kyria Doxa 2 and theargumentof Letter to Menoeceus 125 shows that the state of being deadis not to be feared even in prospectsince it will not be painful.But thisstill does not tackle directly the fear of mortality the fear which somecritics of Epicureanismound left intactby KyriaDoxa 2. The critic mightclaim thatby talkingof a fear of death,he is not claimingto justify thefear of not-being(and so could allow Epicurus he claim in Kyria Doxa2), but rather his fear of mortality. n otherwords,the fear is not basedon an inconsistentdreadof 'being dead',but rather n theprospective on-cern that one's life, plans,and projectsmightbe curtailedprematurely.37If this is a criticismspecificallyof the argument f KyriaDoxa 2, and ofthe Symmetryargumentas I have interpretedt, then it is a reasonableone. Nothinghereexplicitlytackles the problemof fearingthat a life willend incomplete.The criticof Epicureanismmightalso complain hatevenwith the additionof the argumentof Letter to Menoeceus125, Epicurusagain misses the point.He mightwell claim that fearingdeath is not atall like fearingan impendingpainfulevent - it is rathera concernwiththe shape and completenessof a life.This could even be expressed n hedonistic erms,again usingthe coun-terfactualaccount of the harm of death. Someonemightbe quite happywith the conclusion hatbeing dead is not painful,and therefore greethattheyshouldnot fear in anticipationhe state of beingdead(in otherwords,they might agree whole-heartedlywith KyriaDoxa 2 and Letterto Menoeceus125), but neverthelesseel anxiousat the thought hatdyingat somepointin time will rob them of pleasurethatthey would have experiencedhadthey died at a later time. Their life mightbe betterthanit will turnoutto be. Theirdeathmightbe premature.38There aregood reasons to think that the Epicureansdid concernthem-selves with the questionof the completenessof a life and thereforewiththe questionof whetherand when a death can be premature.But theydidthis via theirown particular randof eudaimonistic edonism,notthrough

37 See Striker 1988.3X Indeed, if expressed in hedonistic terms it appears that any death, even one inadvanced old age may rob me of possible pleasures.Of course, on such an accountdeathcould also be counted as a good - provided t releases me from inevitable futurepains.


25/27

LUCRETIUS, SYMMETRYARGUMENTS, AND FEARING DEATH 489the specificarguments urveyedhere. In short,the Epicureansry to dis-sociate the idea of a complete life from notions of temporaldurationandattempt o deny any cumulativenotionof pleasures.Theythereforewouldargue that so long as you achieve katastematic leasure,death cannotrobyou of any further oodwhichyou 'wouldhave experiencedhadyou diedlater'.They try to arguethat complete pleasurecan be experienced n afinite time. This is not the place to delve into or criticise this particularfacet of Epicureanhedonism,but this concernto provide an Epicureanstory about what constitutesa completelife shouldthereforebe seen asthe second complementarytrandof reasoning n Epicureandiscussionsof the fear of death.Once Epicurushas reducedas much as possible ofthe complex of anxieties about death to the fear of being dead, and hascountered hese with Kyria Doxa 2 and the Symmetryargument,he canset about puttingin place his own positive account of a complete lifewhich, he hopes, is immuneto fears of premature eath.39One remainingobjectionshouldbe outlined. It may be arguedthat theattitudewe normally ake towardsdeathcan be justifiedon groundsotherthan its rationality,namelyon its utility.While it mightcertainlybe thecase that the stateof beingdeadinvolvesnothingwhich should be fearedeven in prospect,neverthelesswe could claim that it is best for us to tryto avoid that state for as long as possible.We mighteven call this instru-mentalfear of death an evolutionaryproduct. t is best for each individ-ual to live for as long as possible perhaps n order o procreateas oftenand successfullyas possible- and one way in which this aim might beachieved is throughan innateaversionto death. This fear must be keptdistinct from the fear of pain, which mightof itself be said to make usavoid certainpotentiallyfatal situations.The fear in questionis not thefearof dyingin some particularway; it is theverygeneralandbasicaver-sion to ceasingto be. Epicurus,of course,wouldwholeheartedly ndorseany mechanismwhichallowsus to avoidpain.But still he insists that lifewould be better, ndeedmorepleasant, f we were to stop fearingdeath.Thequestionhere is whether pecifically earingdeathcan be seen to servesome functionsuchthat if it were not presentour lives wouldbe rendered

39 For Epicurean views on complete lives and their relation to the particularEpicureanbrand of hedonism see Mitsis 1988, 320-22, and Warren 2000, 236-44.Lucretiustouches upon these themes at DRN 3.935-45. Note the metaphorof the leakyjar, borrowed from Plato's Gorgias, which also is used by Lucretius to describe the'Danaids' at 3.1002-10.


26/27

490 JAMES WARRENunliveable.Much more could be said about this question,'" ut it is clearenough thatEpicureans o not - or at least are not supposed o - becomesuicidaldare-devils.Presumablynoughof the functionalaversions o riskand pain remain,once the debilitating ear of beingdead is abolished, oallow an Epicureano navigatesafely through he world.4' t seemsto metherefore hatthe fear of pain can carrymost of the weight supposed obe borne by the instrumentalear of deathand that therefore here is noreason to believe that it is impossibleto live a recognisablyhumanlifewithoutfearingdeath.Thatis not to say, of course, thatan Epicureanifeis identicalto those which otherslead. It is not intended o be.MagdaleneCollege, Cambridge

4' Rorty 1983 has a good discussion of the 'functional fears' of death, and con-cludes that there is an irreconcilabledilemma between the irrationalityof the meta-physicalfear of death,andthe inevitablefunctional ear of death.Thisdilemma,however,is judged not to be debilitating.Cf. Oaklander1994, 349. Haji 1991, 177 points outthat this 'sociobiological conjecture' 'may provide the beginnings of an explanationabout why, from moments within their lives, people care more about life's goods ofwhich death could deprive them, than about life's goods they could have enjoyed hadthey not been conceived so late.' Nevertheless, Haji is quick to insist that from anatemporal perspective there is no relevant distinction to be made between pre-nataland post mortemtimes. Bruecknerand Fischer 1993, 42, respond:'[Wie believe thatthe temporally situated perspective is at least as importantas the atemporal withrespect to the generationor recognitionof values.'

4 Diogenes of Oinoandafr. 35 Smith distinguishestwo types of fearing death,onewhich is clear and anotherwhich is not. As an example of the firsthe offers theexam-ple of avoiding something harmful,a fire for example, since it might cause death. Itis unclear, however, whetherDiogenes wishes to endorse this as the corrett reasonforavoiding fire (i.e. the reason why an Epicureansage would avoid fire) or on whatgrounds a committed Epicureancould do so without jeopardising the assertion thatdeath is 'nothing to us'. Of course, it is perfectlyacceptable- and surely plausible -for him to argue that we do and should avoid dying painfully if at all possible.

ReferencesAlberti,A. 1990. Pauradella morte e identiti personale nell'epicureismo.In A. Albertied. 1990. Logica, mente, persona. Florence: 151-206.Bailey, C. 1947. Titi Lucretii Cari de rerumnatura libri sex. 3 vols. Oxford.Brueckner,A.L. and Fischer,J.M. 1993. Death'sbadness.PacificPhilosophicalQuarterly74: 37-45.1998. Being born earlier.AustralasianJournal of Philosophy 76: 110-4.Cockburn, D. 1997. Other times: philosophical perspectives on past, present, andfuture. Cambridge.


27/27

LUCRETIUS,SYMMETRYARGUMENTS, AND FEARING DEATH 491Ernout,A. and Robin, L. 1963. Lucrece de rerumnatura2. Paris.Feldman, F. 1990. F.M. Kamm and the mirrorof time. Pacific Philosophical Quarterly71: 23-7.1992. Confrontationswith the Reaper: a philosophical study of the nature andvalue of death. Oxford.Fischer,J.M. 1993. Introduction:death,metaphysics,and morality.In J.M. Fischered.The Metaphysicsof Death. Stanford:3-30.Furley, D. 1986. Nothing to us? In M. Schofield and G. Strikeredd. The norms ofnature. Cambridge:75-91.Glannon, W. 1994. Temporal asymmetry, life and death. American PhilosophicalQuarterly31: 235-44.Haji, I. 1991. Pre-vitaland post-vitaltimes. Pacific PhilosophicalQuarterly72: 171-80.Kamm, F. 1993. Morality,mortality vol. 1: Death and whom to save from it. Oxford.Kaufman,F. 1999. Pre-vital and post mortemnon-existence.AmericanPhilosophicalQuarterly36: 1-19.Kenney, E. 1971. Lucretius: de rer-umnatura III. Cambridge.Mitsis, P. 1988. Epicuruson death and the durationof life. Proceedings of the BostonArea Colloquium in AncientPhilosophy 4:303-22.Munro, H.A.J. 1886. T. LucretiCari de rerumnatura libri sex4. Cambridge.Nagel, T. 1979. Death. In his Mortal questions. Cambridge: 1-10.Nussbaum, M. 1994. The therapy of desire. Princeton.Oaklander,L.N. 1994. On the experience of tenseless time. In L.N. OaklanderandQ. Smith edd. The new theoryof time. Chelsea, Mich.: 344-50 (originally publishedin Journal of Philosophical Research 18, 1993: 159-66).Parfit, D. 1984. Reasons and persons. Oxford.Rorty, A.O. 1983. Fearingdeath. Philosophy 58: 175-88.Rosenbaum, S.E. 1989. The symmetryargument:Lucretius against the fear of death.Philosophy and Phenomenological Research 1: 353-73.Sedley, D.N. 1993. Chrysippus on psychophysical causality. In J. Brunschwig andM. Nussbaum edd/ Passions and perceptions: studies in Hellenistic philosophy ofmind. Cambridge:313-31.Sorabji, R. 1983. Time, creation, and the continuum.London.Striker, G. 1988. Commentary on Mitsis [1988]. Proceedings of the Boston AreaColloquiumin Ancient Philosophy 4: 323-30.Warren, J. 2000. Epicurean immortality.Oxford Studies in Ancient Philosophy 18:231-61.2001. Lucretianpalingenesis recycled. Classical Quarterly(forthcoming).Williams, B. 1995. Resenting one's own existence. In his Makingsense of humanity.Cambridge:224-32.

Documents

1 Lucretius Death