Sie sind auf Seite 1von 17

Socratic Suicide

Author(s): James Warren

Source: The Journal of Hellenic Studies, Vol. 121 (2001), pp. 91-106
Published by: The Society for the Promotion of Hellenic Studies
Stable URL:
Accessed: 23/10/2008 08:08
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at 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 at
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

The Society for the Promotion of Hellenic Studies is collaborating with JSTOR to digitize, preserve and extend
access to The Journal of Hellenic Studies.

Journal of Hellenic Studies 121 (2001) pp 91-106

Abstract: When is it rationalto commit suicide? More specifically,when is it rationalfor a Platonistto commit suicide, and more worryingly,is it ever not rationalfor a Platonistto commit suicide? If the Phaedo wants us to lear
thatthe soul is immortal,and thatphilosophyis a preparationfor a statebetterthanincarnation,then why does it begin
with a discussion defendingthe prohibitionof suicide? In the course of that discussion, Socratesoffers (but does not
necessarily endorse) two argumentsfor the prohibitionof self-killing, at least in most circumstances,which have
exerted a long and powerful influence over subsequentdiscussion of the topic, particularlyin theist contexts. In the
context of the Phaedo itself, however, this introductoryconversationplays a crucialrole in setting the agendafor the
remainderof the dialogue andoffering an initial
ofit discussion of the majorconces
thon oe
argumentas a whole. In particular,the discussion of thtenatureof suicide is intimatelybound up with Socrates'conception of truephilosophy as
a 'preparationfor death', the relationshipbetween soul and body, and the immortalityof the soul. My intentionis to
provide a readingof thatpassage (61e-69e) which asks whetherthe Phaedo can offer a philosophicallysatisfying distinction between suicide and philosophy and how it relates to other ancient philosophical attitudesto self-killing. I
argue that Socrates does not think that being dead is always preferableto being alive, and that the religious views
expressed in the passage are consistentwith his general stance on the benevolence of the gods.

ANCIENTethical philosophy is essentially concerned with the identification and provision of a

good life (eudaimonia). In the course of this investigationit often turnsto considerthe contrast
between living and being dead, and in particularit often asks the question: When does a life
become no longer worth living? By offering a recipe for a good life, it can also identify unsatisfactory lives. Of course, it is possible to claim that life is never not worth living; life per se is
always good. That is not to say that there can be no contrast between merely living and living a
good life, but it is always better to be alive than not. If this is accepted, then suicide is irrational
and should be prohibited. However, ancient philosophers do not tend to claim that life per se is

good. They are often quite preparedto consider the possibility that on occasion a life is not
worth living. On those occasions suicide may become a rationalcourse of action.'
Let me outline two possible positions from which it can immediatelybe inferredthat one
should commit suicide.
A Being dead is the best state possible.
B Being alive is the worst state possible.
These are extreme positions. Each is compatiblewith post mortem survival (of the soul, spirit
or something similar) but neitherrequiresit. Note that it is not sufficient to argue merely that
death is a good thing, or that life is miserable. It remainspossible in these cases that life is better than death (however good death is), and that death is worse than life (however bad life is).
Between the extremes of A and B is a much more common intermediateposition which asserts
a comparisonbetween living and dying ratherthan a superlativeassessment (negative or positive) of one or the other.
C Being dead is betterthan being alive.
* I would like to

thankDavid Sedley and thejournal's 1216al0 has a discussion of circumstancesunderwhich a

referees for their helpful comments.
life is not worth living. This passage probablyoriginates
1 Consider the famous Socratic maxim: o 6e in Aristotle's Eudemus(see Aristotle ap. Plut. Cons. ad
ave4raaTro; PtIo;oV Ptofl0; avOpan(p, Apol. 38a5-6, Apoll. 115B-E).
and compare Crito 47e7-48a2. Aristotle EE 1215b15-



C also is compatiblewith but does not requiresome sort of post mortemexistence. It invokes a
comparativeassessment of being alive and being dead and comes down in favour of the latter.
It allows thatliving might be a good thing (butbeing dead is better)and also thatbeing dead may
be a bad thing (but being alive is worse). C may be furtherdivided into two claims:
Cl Being dead is always betterthanbeing alive.
C2 Being dead is sometimesbut not always betterthan being alive.
Cl follows from A or B but not vice versa. C2 is incompatiblewith A and B. FromCl one can
inferthatit is always rationalto commit suicide. FromC2 one can inferthatit is rationalto commit suicide only when it is indeed the case that being dead is better than being alive.
Unsurprisingly,C2 is by far the most common standpointfrom which rationalsuicide is justified. Of course, it leaves for furtheranalysisjust what the pertinentconditions may be, and on
that question differentethical theories will give differentresponses.
The notion that life is essentially a miserablestate to be in is familiarin Greek contexts, particularly in tragic drama.2 The particulardifficulty faced by Platonism, however, is that it
appearsto promotea version of Cl. After all, some of the main tenets of Platonismthat appear
in the Phaedo itself include the preferabilityof the soul over the body, of the intelligible over the
perceptible,and - most important- of the discamate soul over the incarnatesoul. In that case it
seems that the immediate course of action for any committedPlatonist should be suicide. Yet
famously in the Phaedo Socratesplaces an almost total prohibitionon killing oneself.
This prohibitionbegins to look even more peculiar when the Platonic stance is contrasted
with that of another of the philosophical heirs of Socrates, Stoicism. Perhapsmore than any
otherancient school, the Stoics - particularlyof the ImperialRomanperiod- are associatedwith
the acceptanceand practice of rationallycalculatedsuicide. This acceptancewas derived from
their agreementwith my position C2 above. Even more peculiar,the loudest voice in condemnation of suicide in Hellenistic philosophy comes from the Epicureans,who are themselves
famous for arguingthat 'death is nothing to us' (KyriaDoxa 2). This position is dependenton
their decidedly un-Platonicanti-eschatology. Death for the Epicureansis annihilation. No good
or bad can be perceived by a subject after death;indeed there is no subjectto do any perceiving
after death.3
Readersfamiliarwith the Phaedo will know that it will be arguedthat the soul is immortaland
thatwhen releasedfromthe body it can enjoy a betterstate of being. Socratescan thereforedrink
the hemlock without thinkingthat he is being harmedby death. But here there arises the problem which I outlined above, a problemwhich also seems to be promptedby Christiannotions of
the afterlife, the soul, and the physical world: Platonism, like Christianity,may be thought to
invite and encouragesuicide.4 Both Platonismand Christianityhold out the promise of a better
existence after death (in heaven, amongst the Forms), and this makes it seem a good idea not to
spend any longer than necessary living a life here on Earth. Perhapswe should all, on finishing
the dialogue, reach for the hemlock and race to join Socrates.
3 The Epicureans' argument has not always won
See e.g. Hom. II. 22.481; Eur. Troades636-7, and
generally the material collected in Stobaeus 4.52b: approval. An influentialrecent criticism can be found in
EIIAINO; OANATOY.Glover (1977) 173-4 emphasizes Nagel (1979), which has provokeda greatdeal of discusthe difficulties in making any assessment that one's sion. For an introductionto the modem literature,see the
future life will not be worth living. CompareFeldman essays collected in Fischer (1993).
4 Pabst Battin (1982) 29, 64-5.
(1992) 217-23.



This problem can arise for any theory which tries to argue that there is nothing terriblein
death. I am not claiming, of course, that the fear of death is the primarymotivationbehind the
Christianidea of heaven or the Platonic idea of the release of the soul from the body at death,
but it is the case that if it is agreed
sense both of 'being dead' and of 'being
therethat death in
mortal') is not a bad thing for humans,and if it is in additionshown that living a life is painful
or promotesill-being in general,then it is difficult to see why it would not be betterall roundto
end one's life as quickly as
apossible. This is the majorreason why it is thoughtpeculiarfor the
Epicureans prohibit
quite as vociferously as they did. They agree with Plato thatthere
is nothingto be fearedin death(thoughfor quiteun-Platonicreasons) and also thinkthat life can
be full of pain and misery - which they identifiedas the only true evil. If so, we might ask, why
is not the rationalstep to commit suicide? In brief, the Epicureanansweris that it is sometimes
acceptableto commit suicide providedthat it is otherwiseimpossible to continueto live a good
life (see Sen. De Beata Vita19.1 and contrastSen. Ep. Mor 30; this is again an acceptanceof C2).
Otherwisesuicide is interpretedas a sign of an inabilityto regulateone's life properlyand well.5
The explanationof an acceptablesuicide is like thatofferedby the Stoics, a school knownto have
engaged closely with the Phaedo and to whom I will returnbelow for furthercomparisons.
Apart from this section of the Phaedo there is anotherPlatonic context in which such arguments are made prominent. In the Pseudo-PlatonicAxiochus Socrates is called to try to soothe
the eponymous character'sfears about death. At first Socrates tries to offer some remarkably
Epicurean-soundingargumentsbut these do not comfort the old man. What eventually proves
to be effective, however, is an approachwhich emphasizesboth the miseries to which the body
is subjectedduring life (366d-369a), and the benefits of discarnateexistence. This is a heady
combinationof my positions A and B, and thereforeC1.6
Socrates:'Andwhilethe soulis forcedto sharewiththesenseorganstheirdiseasesandinflammations
andthe otherinternalills of the body (sinceit is distributed
amongits pores),it longs for its native
heavenlyaether,nay,thirstsafterit, strivingupwardsin hopesfor feastinganddancingthere. Thus
beingreleasedfromlife is a transitionfromsomethingbadto somethinggood.'
Axiochus:'Well,Socrates,if you thinkthatlivingis bad,why do you remainalive? Especiallysince
you puzzle your brainaboutthese things and you're much clevererthanmost of us.' (Ps.-Plato,
Axiochus366a6-b4; trans.J.P.Herschbell)

Axiochus' question is a good one. It is the questionwhich promptsthe discussion of suicide in

the Phaedo since, as Cebes well appreciates,this particularstrandof thought seems to encourage suicide and is therefore in tension with the general disapprovalof taking one's own life.
Axiochus exclaims that as a result of hearing Socrates'argumentshe no longer fears death, and
even yearns for it (370e).
Latersources seem to be well aware of this tendency in Platonism- or at least the tendency
that some might understandPlatonism in this way. Cicero includes a story of a certain
Cleombrotusin his first TusculanDisputation(1.84):

Cf Englert (1994), Cooper (1999), Warren(2000)
242 and n.34. Williams (1976) 207-8 tries to answer
from the opposite perspective why we go on living. He
suggests thatpeople have 'categoricaldesires'which propel the agent to take an active interestin the future. Of
course, if it is known that all one's categorical desires
cannotbe fulfilled, then suicide may indeedbe the rational course of action. Such a life is indeed not 'worth liv-

ing'. See also Williams (1973) esp. 85-92 for furtherdiscussion of such a model.
6 As such, therefore,it seems reasonableto interpret
the Axiochusas a post-EpicureanPlatonistcriticismof and
responseto Epicureanarguments. They are shown to be
therapeuticallyineffective, and much less useful than the
Platonistperspective. For a discussion of the Axiochus
and its relationto Epicureanism,see Furley(1986).


There is a certainepigramby Callimachusabout Cleombrotusof Ambracia,who he says having read

Plato'sbookthrewhimselffroma wall to his deathalthoughnothingbadhadhappenedto him.7

Augustine also tells this story (Civ. Dei 1.22), and specifies that Cleombrotushad been reading the Phaedo (lecto Platonis libro ubi de immortalitateanimae disputavit). Augustine further
points out that Plato himself would have done the same and hurledhimself to his death on completion of the work had he not realized that suicide was prohibited.
It should thereforebe clear why the Phaedo might begin with a discussion of the permissibility of suicide. Not only will Socratesat the end of the dialogue drinkthe hemlock and so precipitate his own death,8but the general conclusions of the dialogue might be thoughtto encourage suicide. There are therefore both dramaticand philosophical reasons for the discussion
beginning as it does.
After Phaedohas introducedthe dramaticscene of the dialogue, the conversationhe narrates
swiftly turnsto the questionwhy Socrateshas begun to write poetrynow thathe has been imprisoned. Socrates answers that a dreamvision had encouragedhim to 'practiseand cultivate the
pyia'ov, 60e6), and that althoughhe has a strong suspicion that this
(govoaiciv toiEt Kcai
means that he should continue to practisephilosophy, he thought it best to try the other arts as
well, and thereforehas been setting some fables of Aesop to verse. He ends by telling Cebes to
inform one Evenus (who has been inquiringafter Socrates'new poetic endeavours)to hurryto
follow him when he 'leaves' Athens laterthat day. By 'leave Athens' Socratesmeans 'die'.9
Simmias is taken abackby this, and asks Socratesto explain what he means by this apparent
encouragementto Evenus to hasten his own death. The cryptic answer is that if Evenus is a
philosopherhe will be willing to do this, althoughit is generally thoughtthat suicide is wrong.
This comment sets up the subsequentdiscussion in which Cebes poses a question for Socrates.
How can he claim both that a philosopherwill be willing to follow someone who is dying, and
also that suicide is wrong?
It is worthpausingto see how much the opening conversationhas alreadyachieved by adumbratingor alludingto what will follow. First, Socrates'dream,much like a similardreamreported in the Crito (44a-b), a dialogue set dramaticallyonly two days before the Phaedo, offers
divine sanction of some sort for Socrates'actions. Later,it will be remarkedthat suicide is not
permitteduntil the gods send some sort of 'necessity' (62c7). This will form part of Socrates'
eventualstory of the purificationof the soul duringlife, and the eventualfinal separationof body
and soul at death (67a6).
Second, Socrateshas alreadyin this opening exchange introducedthe notion of the practice
of philosophy,and has more importantlyimplied that there are right and wrong ways to engage
in it. It is not thereforeincidentalthatwhen telling SimmiasthatEvenus will indeed follow him,
he specifies that this is the case for all those who 'partakeworthily' of philosophy (nag; Trq?
nP(p&yao; ReeaOxtv, 61c8-9). This is, of course, one of the centralthemes
atio xoxo
of the entire dialogue, namely that the true and correctpractice of philosophy is a 'preparation
7 Cf. Cic. Pro Scauro 4-5. In the Tusculansthis follows Cicero's story about a certain Cyrenaic called
Hegesias, who had such a pessimistic view of the chances
of humanhappinessthat those who listened to him often
immediately went off to commit suicide. As a result,
Ptolemy Philadelphus banned him from giving public
lectures. The story of Cleombrotusappearsin a number
of other sources, perhapsfirst in a poem of Callimachus
(AP 2.471 = Callim. 53 Gow and Page). For a discussion
of the history of this story, see Williams (1995). Cf.
Griffin(1986) 71. Riginos (1976) 180-3 gives references

to all the appearancesof this anecdote. Note also that

Cato is said to have been spurredon to suicide by reading
Plato: Sen. Ep. Mor.24.6, Riginos (1976) 183.
8 On the question whether Socrates can be said to
have committed suicide, since of course he is orderedto
drinkthe hemlock by the Athenians,see Frey (1980).
aieiugt 61b9 is not just a euphemism. Socrates
later try to demonstratethat at the point of deathhis true
self, namely his soul, will 'leave' this place. Comparethe
use of &oio&reiv and its cognates at 61el-2 and 67cl.



for death' since it is the practice by which one purifies and liberates the soul from its bodily
incarnation.10This too will be relevantto the argumentsfor the permissibilityor otherwise of
suicide, the first of which begins with the thoughtthat the soul is 'imprisoned'within the body.
Let us now turnto the particularargumentsaboutsuicide which appearin the Phaedo."ISocrates
is respondingto Cebes'worrythathe is simultaneouslypromotingthe benefitsof death,andrecommendingit to all truephilosophers,and also subscribingto the generaldisapprovalof self-killing.
His first response is famously difficult to interpret.
Butperhapsit seemsastonishingto you if this aloneof all othersis simple,andit neverturnsout for
a human,as other things do, that at some times and for some people it is betterto be dead than alive.
Perhapsit seems astonishingto you if for those men - the ones for whom it is betterto be dead than

alive- it is notholy forthemto do themselvesa favour,butthattheymustwaitfor someotherassistant.(62a2-7)

Thereare many difficultiesin assessing this passage.12In general,however,it can be agreedthat
Socratesis tryingto offer reasonsfor Cebes' difficultieswith what he has just declared. First,he
suggests that Cebes is astonishedthat 'this' admitsno circumstantialqualifications. But what is
this 'this'? Variouscandidateshave been proposed,includingthe injunctionnot to commit suicide, or the propositionthat it is betterto be alive than dead, or simply 'death'itself. In essence,
however,the exact referentdoes not alterto any great extent the overall meaning of the passage.
Socratesfirst suggests that Cebes is amazed at the apparentlyabsolutenatureof the prohibition. In general such blanketdeclarationsare subject to qualificationsand exceptions - indeed,
Socrates himself is notorious for insisting on such qualificationsto the various ethical claims
made by his interlocutors. Here, however, no possibility of qualificationhas been canvassed.
The categorical nature of the prohibitionis also the source of Cebes' worry that Socrates is
inconsistent. There would be no tension whatsoever between the philosopher's practice for
death, and the disapprovalof suicide if it is the case that in some circumstancesit is in fact better for a person to die ratherthan live - since the philosophermay be one of those exceptions.
The second possible source of astonishmenthas tended to be under-emphasizedin discussions of this passage, which are more interestedin identifying the possible readings of the first
'source of astonishment'. However, I think that this second suggestion will offer more help in
clarifying the problem of distinguishingthese two desires for death - the one belonging to the
suicide, the otherthat of the philosopher.
The second source of astonishmentis a conflict between the generalprinciplethatone should
always act in one's own interestsand the prohibitionon suicide. These two conflict if it is sometimes the case that it is in fact in an agent's best intereststo be dead ratherthan alive (my C2
above), but nevertheless suicide is prohibited. In his exposition of this worry Socrates repeats
the idea that it is sometimes betterto be dead than alive (hence, perhaps,the repetitionof olq Se
PEI3XtovTe0vvvat at 62a5), but neverthelessclaims thatsuicide is prohibitedeven in those cases.
10Therearerepeatedreferencesto trueor correctphilosophy and true or correctphilosophersin this passage:
63alO, 64a4-5, 64b4-5, 64b9, 66b2, 67d8, 67e5.
" There is another discussion of suicide at Laws
873c-d, specifically wonderingaboutthe penaltiesappropriate for a suicide. It is made clear that the cases under
considerationexclude circumstancessuch as Socrates',

who is ordered to die RoKCeo;w4at"i; Bicrl (873c5),

and circumstanceswhere suicide is provoked by shame
or unavoidablemisfortune. The emphasis is on the correct ritualand purificatorymeasuresto be taken. On this
and other mentions of suicide in the Laws, see Cooper
(1999) 523-6.
12 Gallop (1975) 79-85 has a long discussion.



Independentgrounds are offered for the prohibition- namely that suicide in not 'holy'. In all
othercases of the assessmentof the rightnessof an action,thatassessmentmust take note of context and circumstance. What might promote the agent's well-being in one circumstancecould
be unhelpful in another. But the case of suicide is different. Under no circumstancesshould it
be attempted(because of some divine command), even if it would on occasion promote self
Socrates suggests that someone who thinks it betterto be dead ratherthan alive may have to
wait for some 'other assistant'(62a7). This implies thatwhat is unholy about suicide is precisely that the agent kills his or herself.13It is the self-reflexive natureof suicide which in this case
sparkssuspicions of contraveningdivine law. This particularcharacteristicof suicide is repeatedly emphasized in the surroundingtext. At 61d4 Cebes asks what Socrates means by saying
that it is not right (0eLgt1xv)to do violence to oneself (ob gil 0et-brv i?vatl Eavotv ptia?oat),
and at 61e5 Cebes again asks why people think it wrong to kill oneself (cKara Ti6 ovv ToteoV
here emphasizinghis point by using both the
(paoaOegri v elvait acrxoveacvxova ioKetevvwvat;)
normaland reflexive forms of the pronoun.
More tellingly, at 62a6, duringSocrates'explanationof the second source of worry,he echoes
this phrasingwhen he formulatesCebes' difficulty as follows:
Perhapsit seemsastonishingto you if forthoseforwhomit is betterto be deadthanaliveit is notholy
forthemto do themselvesa favour...
Both the reference to what is holy (here O<tov, but this is not I think relevantly differentfrom
the force of 0OegtTvin Cebes' expressions) and the self-reflexive construction(axo(t; ea-TvxoT;
e? itoitiv) parallelthe constructionswhich Cebes used to express why suicide may be thought
wrong. But here Socrates is not describing someone 'doing violence' to himself, but rather
someone 'benefiting himself'. This alterationdistils the problem at hand, since it has become
clear thaton some occasions to kill oneself might in fact be the course of action which does most
benefit the agent. But whereas the reflexive natureof promotingone's own self-interestis not
thought in the least problematic,it suddenly threatensto become so if we fill in suicide as the
particularcourse of action in question. If indeedthe groundsfor the religiousproscriptionof suicide is the fact that the subjectkills himselfthen this cannotbe compatiblewith the thoughtthat
any agent should always act to producewhat is best for himself.14
In this brief exchange alreadya serious difficulty has been outlined. It is already clear that
once it is agreed that on occasion being dead is preferableto being alive then a prohibitionon
suicide can only be sustained by reaching for an independentsource of justification. More
importantstill is the particularsource that Socratesuses: divine displeasure. Suicide in unholy.
At this early stage we alreadyhave in place the necessary elements for the subsequentproblems
about humans'relationshipwith the gods. On occasion the gods prohibitus from doing what is
best for us. Can the gods thereforealways be intenton fostering our best interests?

13There is a possible ambiguityhere. Does someone

commit suicide only if the fatal act is self-inflicted, i.e.
only if the agent himself slits his own wrists, or pulls the
trigger of the gun? It is surely possible that a death can
be other-inflictedand still count as suicide (e.g. Agent X
points a gun at Agent Y and forces Y to kill him). In fact,
it is notoriouslydifficultto define suicidein termsof necessary and sufficient conditions. See Frey (1980) and
Windt (1980). The United States Centers for Disease

Control and Prevention define suicide as a 'death from

injury,poisoning, or suffocationwhere there is evidence
(eitherexplicit or implicit) thatthe injurywas self-inflicted and that the decedentintendedto kill himself/herself
14Bostock (1986) 19 'It is certainlyconsistentto hold
both that some people would be betteroff dead and that
no one ought ever to commit suicide, but in thatcase one
cannot also hold that the sole basis of morality is selfinterest.'



So much for the opening exchanges. Socrateshas said nothing yet abouthis own attitudeto
suicide, but has merely been offeringpossible diagnoses of Cebes' astonishment. Cebes himself
certainlyseems to approveof these suggestions as possible causes of astonishment- if thatis the
force of his exclamationat 62a8 - andthe argumentproceedson the basis thatthese arethe problems which it must solve.
Socratesgives two possible reasonswhy suicide mightbe prohibited,each in the form of an analogy. The first suggests that humansare, when alive, in some sort of prison from which it is not
permittedto escape.'5 This likening of incarnationto imprisonmenthas obvious resonanceswith
the setting of the dialogue. Socratesis imprisonedand preparingto die. But if incarnationitself
is a certainsort of imprisonment,then Socrates'currentposition is relevantlylike thatof all mortals - and in this way Socrates'actions and attitudescan serve generallyas a model for all human
actions and attitudes. We are all incarceratedand must approachthat imprisonmentaccordingly. But, having raised this possible justification of a prohibitionon suicide, Socrates rapidly
moves on, remarkingthatthis pictureis 'impressive,and not easy to understandin full' (62b5).16
Indeed, the picture of bodily imprisonmentseems to raise more questions than it answers. No
justification is provided for our imprisonment,nor is it explained why it is not permittedto run
away.17It is certainlypossible to constructthe sort of reasonsrequired. Perhapswe have all been
imprisonedjustly because of some deep moral failing or terribleact committedin a previous life
which requiresexpiation. But this sort of thinking leads rapidly into furthertroubledwaters.
Who is punishing me? What exactly is the crime, and is it not questionablewhether 'I' committed it - since I surely cannotrememberever committingsuch an act, and it is agreedthat the
crime was committedbefore my birth?
This image of bodily incarcerationcan be comparedwith a similarPlatonicimage of the body
as a tomb (o3g.a afiia). Both convey some idea of incarnationas a less desirablestatethanwhat
came beforeor after,andboth imply thatthereis a furtherkind of existence in which we are liberated from the prison or tomb and can live freely. However, Socratespresumablyuses the prison
15The word used for
prison here, ppoupad,can also
mean 'guard post', and some commentatorshave taken
this to be its meaninghere. Gallopad loc. finds problems
even with this understanding,since if the idea is to brand
suicide an act of cowardice, it seems neverthelesspossible to imagine cases of brave self-sacrifice. Even so, the
clear and immediatelink between Socrates'currentposition, and the idea of incarnationas imprisonmentseem to
me to tell against this understandingof the word. Cf
Cooper(1999) 522, Rowe ad loc. Thereare ancientreferences to this section of the Phaedo which also retainthe
notion of imprisonment:Cic. Tusc. 1.74: vincla carceris;
Rep. 6.14: e corporumvinclis tamquame carcere; 6.15:
corporis custodiis, in custodia corporis (althoughcustodia too can also mean 'guardpost',here Cicero writes of
god 'liberating'someone from the body); Aug. Civ. Dei
16 It is often pointed out that there may be
Pythagoreanroots for this image. If so, then Socrates'
commentaboutit being 'lofty and impenetrable'might be
a furtherswipe at Philolaus' lack of clarity aboutmatters
of the soul and eschatology (61e6-9). Sedley (1995) 11:
'[W]hatwe see in the Phaedo is the paradoxicalspectacle

of Socrates having to persuade the Pythagoreansof the

truthof their own doctrine'. Huffman(1993) 327 argues
that no doctrinecan be securely ascribedto Philolaus on
the basis of this section of the Phaedo besides the assertion that suicide is prohibited(cf. 408-10).
17 Bostock (1986) 18 calls this idea a 'non-starter,
because it has no implicationsabout the basis of morality'. Presumablyit could have the requiredimplications,
but only if it were spelled out exactly why we had been
imprisonedand how we might serve our appointedsentence. Rowe ad 62bl suggests that the image of incarceration recurs at 67dl-2, 81e2, 92al and also that this
idea is not pursuedat this early stage since Socrates has
not yet outlinedthe relevantoppositionof body and soul.
But his paralleltexts refer only to the soul being 'bound
into' the body, not imprisoned,and while these two ideas
are clearlyrelatedI thinkthat it is the problemof explaining thejudicial implicationsof imprisonmentwhich causes
most difficultyfor the incarcerationimage. The notion of
'binding' could just as well be linked with the next proposed justification:that we are slaves and possessions of
the gods.



motif here since only that can give the requirednotion of an obligation to remainin this particular state. The 'body as a tomb' view conveys well the idea that what we presentlycall 'life' is
in fact merely an inferior state when comparedwith the true life we will enjoy, paradoxically,
after (what we call) death. This chimes well with the general and familiarPlatonic notion that
the physical world is inferiorto the intelligible world inhabitd by discarnatesouls. By inverting the unreflective notions of life and death, it also effects a radical inversion of a traditional
Greek motif which saw the souls after death inhabitinga shadowy and inferiorworld. Perhaps
most relevanthere is the complaintof Achilles' ghost when summonedto speak with Odysseus.
He claims that he would prefer to be a lowly bondsmanon earth than the king of all the dead
(Homer,Odyssey 11.488-91). The eventualPlatonicpicturepaintedby the Phaedo will be quite
the opposite - the presentperceptibleworld of the 'living' is a world of images and shadows.18
But this, of course, merely reinforcesthe problembeing posed for Socrates. If our present life
is so miserable in comparison,and philosophers are desperatelytrying to free themselves and
'die', then why is suicide not the quickest and most reasonableroute?
Socratesthen offers an alternativejustification. We are the possessions, indeed the slaves, of
the gods. Just as we would reasonablybe angered if one of our possessions were to destroy
itself, so would the gods be if we were to commit suicide (62b6-c4). Such angerwould reasonably bring punishment.19
Socrates at least thinks that this is better expressed than the preceding argument(62b7), but
once again problemsrapidlysurface. Damascius' commentaryon this section (1.20) points out
well that the notion of punishmentalready implies the immortalityof the soul, or at least that
something relevantly punishablewill survive suicide. However, if the idea of punishmentfor
suicide is the majorforce of this argument,then the analogybetween the gods as ourmastersand
mortalslave owners threatensto breakdown. It may be possible for a god to punish a dead person for committingsuicide, but how is a mortal slave owner supposed to punish a dead slave?
The answer,of course, is that it is not possible for humanmastersto punish suicidal slaves, but
Socratesis perfectly awareof this andphrasesthe questionaccordingly. Socratesasks Cebes the
following (62c 1-4):
If one of yourpossessionswereto kill itself,withoutyou indicatingthatyou wantedit to die, would
you notbecomeangrywithit, andpunishit if you could?
He asks Cebes to consider whether he would punish the slave were this possible. And the
answer to this is certainly affirmative.20Socrates asks Cebes to imagine himself in the role of
one of the gods, whose slaves we are agreed to be, and
certainly e able to exact punishin
ment from those who commit suicide by dealing
particularmannerwith their discarnateand
It is possible to object to the master-slave analogy by suggesting that it must be permissible
for a slave to commit suicide if his life is so miserablethat it becomes literally intolerable.21In
other words, there are limits to the obligations imposed by being someone's slave - and if one's
Cf. Horn.Od. 11.475-6: evOa vexpol Ia(ppaSee;
vaio-ol, Pporv e'i&oXaicag'ovcov.
19 Pabst Battin (1982) 39-41 provides a number of
parallels for argumentswhich try to show that suicide is
not permittedby claiming that our lives are the gifts of
god(s), or that we are the children,slaves, possessions of
20 We might begin to speculate over whether this
question implies that it was thought acceptable for an
Athenianslave-ownerto commandthe suicide of his own

slaves, but what evidence there survives seems to run

counter to this suggestion. In Athens, unlike Sparta,it
seems to have been accepted that a master was obliged
not to kill his slave. See Todd (1993) 184-92.
21 Pabst Battin (1982) 47: '[A]lthougha well-treated
slave might have some obligationto remain,a mistreated
slave does not. Analogously, the person who escapes
from an unusuallycruel servitudein life cannotbe said to
have done wrong.'



master is particularlycruel then the status of being a slave does not preclude escaping as intolerable life. Whetheror not this is an appropriateobjectionto the institutionof chattelslavery at
Athens, it is clear that Socrates is assuming throughouta benevolent master - since only this
would produce the relevant analogy with the gods. Cebes' and Simmias' next objection also
makes it clear that this is Socrates'assumption.
As forwhatyou weresaying,thatphilosophers
shouldbe willingandreadyto die,thatseemsstrange,
namely,thata god is ourprotectorandthatwe arehis
shouldnotresentleavingthisservicein whichthey
aregovernedby thebestof masters,thegods,fora wise mancannotbelievethathe will lookafterhimself betterwhenhe is free.(62c9-d7)22
In describinghow suicide is wrong, Socrates'argumenthas made our lives so good (since we
are looked afterby the best masters),and has made our lives so clearly not our own to dispense
with at will (since we belong to the gods), thatany decision to end one's own life now looks like
utter foolishness. Cebes cleverly point out here that the particularpeople we are interestedin,
namely the philosophers,are surely the very people best capable of seeing the benefits and the
duties of continuingto live in the gods' care. So they shouldwish to die least of all. But Socrates
claims that the very markof a philosopheris his desire to die, since philosophy on his conception is a practicefor death.23
Socrates'answerto this worry is to allow that Simmias'and Cebes' objectionwould be powerful indeed if there were not 'better'masterswaiting for us after death and were Socrates not
convinced thathe is going to meet bettermen once he is dead thanthose he meets generallyduring life (63b4-c7). This then generates the remainderof the dialogue, which is promptedby
Simmias' requestthat Socratesshould try to convince the others of the truthof this claim.
Of course, Socrates'insistence on the preferabilityof the gods and companyhe will meet after
deathdoes not of itself help him to maintainhis prohibitionon suicide, since once again it looks
as if it promotes the benefits of post mortem existence without any referenceto the mannerin
which one has lived or died. But surely this is what Socratesought to be insisting upon. If anything has emerged from the brief exchange on the various possible justifications for a prohibition on suicide, it is precisely that if they make no referenceto the mannerof death or the manner in which one has lived but merely insist upon an obligationto remainliving, or a particular
comparativeevaluationof deathand life, then no relevantdistinctioncan be maintainedbetween
the philosopher'spracticefor death, and the suicide's hasty departurefrom life.
In essence, any form of argumentwhich makes living a good or a necessary state of affairs
will seem both to prohibitsuicide and also to contradictSocrates'claim thata philosophershould
desire and practise death. Conversely,any argumentwhich makes living undesirable,or being
dead preferableto being alive will be consonantwith Socrates'conceptionof the aims of the true
philosopher,but will seem to recommendsuicide more generally to non-philosopherstoo.

22 Simmias offers the same sort of considerationat

Olympiodorusin Plat. Phaed. A 2.1 casts Cebes'
objectionin the form of two syllogisms: (a.) The philosopher wishes to die; he who wishes to die flees good mas-

ters; he who flees good mastersis a fool; so the philosopher is a fool. (b.) The philosopherflees the good (sc. by
wishing to die); but no philosopherflees the good, since
he aims always at the good; so the philosopheris not a



Thus far, the prohibitionon suicide rests on the assertionthat it is 'unholy'. Two possible ways
of understandingjust what this means in termsof the relationshipbetween humansand gods (that
we are prisonersand that we are slaves) have proved to be unhelpful. Nevertheless, something
positive does emerge from that discussion.
Just as it is permittedfor one of Cebes' slaves to commit suicide, provided Cebes indicates
that he should, so it is permittedfor us to commit suicide, providedthat our masters- the gods
- make some sort of indicationthat that is what they wish us to do. Suicide committedon the
basis of divine ordersis certainlyacceptable,indeed is the exception to the general prohibition
on self-killing. Socrates immediately draws this conclusion. It is nevertheless permissible to
commit suicide 'if god sends some necessity' (62c7-8). If the prohibitionon suicide was based
on it being unholy, this exception is perfectly reasonable. What the gods command is surely
never 'unholy'. What form such 'necessity' may take or how it is to be recognized is not stated,
but Socrates' question to Cebes asks if a slave could commit suicide if the master gives some
sign that this is his wish (armrjvavto;, 62c2). Tis suggests that the god's directionmight not
iohint or gesturemay suffice. If Socrates
be in the form of a direct command. Some less obvious
does considerhimself to be aboutto coommitsuicide, we must presumethathe thinksthathe has
had some such message or allowance - whetherthat came in the form of the Athenians'judgement or the dreamhe recountsat the beginning of the Crito (44a5-b5). The Phaedo itself offers
a possible example of such a sign. Socrates'dreamat 60e, by telling him to practisethe arts,will
also be recommendingthat he die, if it is understoodthat the specific art in question is philosophy and philosophy is a preparationfor death.
The Stoics, perhaps the most renowned advocates of rationally considered suicide in the
ancientworld, agree with Socratesthat on occasion god will provide a sign that it is time to give
up life. It is quite plausible that the Stoics took this section of the Phaedo as the startingpoint
for their own reflections on the timing and permissibilityof suicide.24There is also a story that
Zeno, the founderof Stoicism, committedsuicide after trippingover a pebble. He respondsby
quoting from Euripides'Niobe: 'I'm coming! Why are you calling me?' (DL 7.28, 31).
This connection with the Phaedo is straightforward.However, the Stoics differ markedly
from Platonic theory in one major regard. They do not hold that there is an immaterialsoul
which after death communes with gods or Forms or similar incorporealthings. The souls of
Stoic sages do - it seems - survive death but are destroyed at the next conflagration. (Some
sources reportthan Cleanthes
thoughtall souls were similarlyrobust;the restrictionwas probaBut
by Chrysippus).25 the Stoics do not make this post-mortemsurvival of the soul a
factor the decision whether to commit suicide. Their justification for a rational suicide
(eiSXoyo;iayoyf) is not the promise of a better state after death. Rather,in Stoic ethics the
only good is virtue;life is merely an 'indifferent',like wealth, health and so on. Suicide can be
the 'appropriate'(icaiicov) thing to do just as other actions can be so viewed. It is appropriate
if it is in accordancewith the usual rules of Stoic good action - if it can be given a rationaljustification(eSOXoyov
a&ioXoytagov, see DL 7.107) which accordswith the properevaluationof the
variouschoices available(see also Cic. Fin. 3.60-1). Since kathekontacan be performedby both
the virtuous sage and non-sages, both will commit suicide if it is recognized as the appropriate
thing to do, but since sages always do the right thing they will always commit suicide if it is
appropriate.Non-sages can be mistaken.26
Cf Epict. 1.29.29. For a suggestion thatthe Stoics
read the Crito in this light, see Sedley (1993).
25 See DL 7.156-7 and Hoven (1971) 44-65.

26Rist (1969) 239-42 worriesunnecessarilyaboutthe

suicides of non-sages. If they can and should perform
kathekontaand suicide is sometimes kathekonthen they
can and sometimes should commit suicide.



Diogenes Laertius(7.130) tells us thatthe Stoics allow suicide 'for the sake of one's country
or friends, and in the case of intolerablepain, handicap,or incurabledisease'. This offers two
sorts of justification. First, suicide can be a positively beneficial altruisticact; sometimes it is
the appropriatething to do to benefit one's countryor friendsby killing oneself. Second, suicide
can be beneficial for the agent himself; it is appropriateif the alternativeis a continuedlife of
suffering. Suicide can eitherpromotepreferrednaturalindifferents(e.g. the well-being of one's
friends)or remove dispreferredindifferents(e.g. pain, handicap). The Stoics' immanentdivinity, the logos of which humanreason is a part and with which humanwishes should be aligned,
can give a sign when suicide is appropriate. Indeed, the circumstanceswhich must obtain for
suicide to be appropriatemay themselves be interpretedas this divine sign.27
The significant point for the present discussion is the following. In Stoicism, due to the
absence of any post mortemworld and due to the harmonizationof human and divine reason,
thereis no possibility of any conflict between what god commandsand what is beneficial for the
humanagent. A virtuoushumanwill be in complete agreementwith a divinity that is provident
and benevolent. The problemin the Phaedo is the apparentdissonancebetween the notion of a
preferredpost mortemstate inhabitedby providentgods, andthe apparentprohibitionon suicide
made by those gods even when it is agreedthat continuedlife is not preferableto death. Further,
Socrates'exception to the rule - the
tdivine sign - seems a purelyarbitrarymeans by which gods
do what will benefit them. The gods of the Phaedo risk lookmay may
owners who deal with their possessions on a whim.
In orderto avoid these pitfalls, Socratesmust still show how philosophy as a preparationfor
death is relevantlyunlike suicide, and also how this divine exception relates to that distinction.
This should ideally also allow the gods to act always with a view to humans'best interests. He
does achieve this, I think, but in orderto doso
so he first explains what he means by making philosophy 'a preparationfor death'.
Let me pause at this point to discuss two more unsuccessful attemptsto make the requireddistinctionbetween two kinds of death offered by two ancient commentatorson this section of the
The story of Cleombrotus,as told by Callimachus'epigram,is cited by Ammonius in order
to illustratehow someone might misunderstandthe characterizationof philosophy as a preparation for and practiceof dying. Ammonius (in Porph. Isag. 5,21 ff.) distinguishesbetween 'natural' (qpaIK6;) death, which is merely the separationof the soul and body, and 'chosen' or
'intentional'(7poatp?tlKo;) death.28Only the latteris sought afterby the truephilosopherwho
actively tries to separatethe soul from the body. However,of itself the introductionof the notion
of volition does not help a great deal, since the requireddistinctionis not between deathper se,
and the philosophicalpracticeof death,but between the voluntarydeathof suicide (as for example, achieved by Cleombrotus),and the voluntaryphilosophicaldesire to separatesoul and body.
A similar distinction is made by Olympiodorusin his commentaryon this section of the
Phaedo (A 1.12). He insists on distinguishing'killing oneself' (t6 cdyeiv cauov)29 and 'wish27 Griffin (1986) 72:
'[T]heir doctrine can be 524, 532-6 contrasts the Stoic position with that of
describedas an internalisationof Socrates'divine neces- Plato's Laws 854a-c.
28 This distinction is also made by Olympiodorusin
sity so that it becomes a dictate of man's own reason,
which tells him when life accordingto natureis no longer Plat. Phaed. A 3.11.
29 Literally,of course, 'leading oneself out'. Perhaps
possible. This modificationof Platonismwas made possible by the Stoic belief that the divinity of the world is the metaphor is supposed to recall the escape from a
immanent.' For more discussion of Stoicism and suicide, prison, as if one is acting as one's own jailer.
see Rist (1969) 233-55, Englert (1994). Cooper (1999)



ing to die' (o6 C0eeXvasto viralcetv).Thefirstof these,he claims,is whatis strictlyprohibitmodeof living. Indeed,theexactphrase
ed,whereasthesecondis theobjectof thephilosopher's
used by Olympiodorushere (To?0?Xe tv a&noOviGKetv)is used in the Phaedo to describe what

do (62c10). It is difficult,however,to see theexactcontrastOlympiodorus
Like Ammonius he stresses the voluntarynature of the true philosopher's death, but unphilosophical suicide is presumablyvoluntaryin a sense also. Perhapsby stressingthat the philosopher 'wishes' to die, Olympiodorusimplies that he has an aspirationtowardsthe separationof
body and soul but does not immediatelyrushto tear apartthe two throughsuicide. But still, this
philosophical aspirationshould nevertheless be manifested in some sort of action, namely the
concentrationon psychic ratherthan bodily concerns. What is requiredis a distinctionbetween
the unphilosophicalshort cut to the separationof body and soul and the gradualphilosophical
purificationof the soul and removal of bodily concerns which occurs throughouta life. In fact,
just such a distinctionis, I think,providedby the Phaedo, but in a somewhatrondabout way.
In the Phaedo the appeal to gods as our guardians or masters has not proved to be particularly

helpful in securingthe requireddistinctionbetween suicide and philosophicalstrivingfor death.

So Socrates tries a different tactic, and the conversation from 63b4 takes a new turn by trying to

explain to Simmias and Cebes just how it is that a philosopherpractisesdying. This sets aside
for a while the discussion of the prohibitionon suicide, but in outlining what he means by this
conception of philosophy, Socrates will be able indirectly to set out a relevant distinction
between the philosopher's desire for death and the suicide's abandonment of life.
After a brief interlude in the conversation during which Crito informs Socrates that the jailer

advises that he should talk as little as possible, Socrates explains that most people are unaware

that true philosophers are 'nearly' dead even when they - the philosophers - are living (64b5:
tense emphasizes
the presentthattense
emphasizes that this is an ongoing process of gradual death).30 This

produces an interestingreaction from Simmias, who (perhapsa little tastelessly given the circumstances)suggests that Socrates'conception of philosophersstrivingfor death will be shared
by lots of people who do indeed think that philosophersare 'nearly dead' and in fact deserve to
be so (64alO-b6). This prompts Socrates to make explicit that it is necessary to distinguish
between kinds of death (oMovOavaro'u, 64b9). First Socrates sets out a working definition of
death as the separationof the soul from the body:
of the soul fromthe body? Andis thendeaththe followTherefore,is it anythingbutthe separation
ing:the bodyitselfbeingseparateandapartfromthe soul andbecomingby itself,andthe soul itself
being separateand apartfromthe body andbeingby itself? So is deathanythingotherthanthis?
This working definition is perfectly compatibleand indeed is supposedto include what happens at the point of death. When someone's life ends, the soul and body (now a corpse) separate
entirely. The body and soul continue to exist but apart from one another - a model which
Socrates'Pythagoreancompanionswould not find objectionablein the least.31However,in what
follows Socrates sets in place the considerationsnecessary to allow a new an alternativeunderstandingof this descriptionof death. This understandingwill allow that the separationof body
and soul can occur - albeit to a limited degree - within one's lifetime.
31 This definition of death was also accepted by the
Cf Olympiodorus in Plat. Phaed. A 3.3:
See e.g. Plut. De Stoic. repug. 1052C.
cKaOaipo3v?avTov iiv 7naO6Rv,T?OvqKev 6e Ij 6O


KecXptrTat Yap TOv ItaO&v.



First, Socrates describes certainpleasures (of food, drink, and sex) as bodily. The philosopher is then agreed not to be concernedwith such pleasures and thereforein general not to be
concerned with bodily matters. Instead his attention is directed towards the soul (64e4-6). This

concentrationis swiftly characterizedas a desire to 'release' the soul from bodily concerns
(65al) and thereforeit can quickly be concludedthat the person who pays no attentionto bodily pleasures is striving as far as possible for death. So Socrates is relying on the definition of
death describedearlier,but has managedto enrichour understandingof that definitionby insisting thatthe 'separation'involved need not be purelyphysical. A concentrationon the soul rather
than the body and a correspondingneglect of the latteralso fit this description.
Socrates draws a contrastbetween the philosopherwho is concernedwith the soul, and the
personwho is concernedentirelywith the body. He points out thatthe majorityof people would
considerthose who pay no attentionto bodily concernsand pleasuresnot to be worthy of living,
and indeed to verge on being dead (65a4-7). These people are thereforeworking with a conception of true living quite opposed to that of the philosophers. For the majority,trulyto live is
to indulge in pleasures and passions. Philosophersavoid these, and so are not on this account
truly alive. But also in the philosophers'own terms, they are 'dying', since by this the philosophers mean that they are trying to separate the soul from the body. For the philosophers this is

a positive process, tending towardthe goal of a free and purified soul. So the philosopherand
'the many' agree that a life of philosophy is a pursuitof death, and the pursuitof bodily pleasures is properlycalled 'living'. But they disagree in their assessments of the relative value of
living and dying.32
Here, however, a problemsurely arises. How can Socratesmaintainthat in one case the concentrationon psychic concernscounts as a separationof body and soul, and thereforedeath,but
a concentrationon bodily concerns does not? If it were the case that the mere wholehearted
emphasison one of the body-soul pairwould count as a case of separatingand dividingthe body
and soul,hen the truephilosopherhas no more claim to be engaged in a 'practiceof dying' than
the person entirely consumedwith the pursuitof physical pleasures. Both would fit the definition of death outlined in 64c4-8.
The distinctionbetween these two is spelled out in the subsequenttext, in which Socrates
implies that ratherthan forcing a separationof body and soul, the pursuitof bodily pleasures
forces the two to become more tightly bound together. He describes at length the troublesthat
assail the soul as a result of its incarnationthroughperception,pleasureandpain (65a-67a). The
upshot of this ability of the body to troublethe soul is that a focus on bodily concerns does not
in fact leave the soul alone. In order to pursue and maximise physical pleasures, we need to
engage psychic capacities of perception,planning and the like. Indeed, not only does such a
process bind the soul and body together,it reversesthe properhierarchicalrelationshipbetween
the two. The soul becomes a tool or instrument,indeed a slave, to the body, aiding and abetting
the searchfor corporealdelight. So ratherthan effecting a separationof body and soul, concentrationon the body requiresthe two to become more closely fused, and thereforefurtherfrom
the separationwhich constitutesdeath. Indeed, on this score Socratescan agree with those who
thinkthat only this pleasure-filledpursuitis truly 'living' (65a6). It is clear thatbodily concerns
infect and pollute the soul, just as we learnthat philosophicalstudy purifies the soul.
So doesn't this purificationturnout to be what we said in the argumenta while ago, namely the separation as far as possible of the soul fromthe body, and the trainingof the soul to gatherand collect itself
in from every partof the body, and live as far as possible both in the presentand in the futurein itself,
freed from the body as though from bondage? (67c5-d2)


Cf. Burger (1984) 40.



Armed with this account of the interaction between body and soul, Socrates now at last

returnsto the possibility of some divine indicationto end one's life. (62c7). When giving his full
account of the philosopher's life of purification,Socrates makes it clear that a philosopherwill
not pre-empt the final separation of body and soul at 'natural' death until god himself releases

him from the prison of the body (67a6). This is now closely linked to the notion of purification
and the implicit message must be that the philosopheris not ready for this divine release until
the soul has been purged sufficiently of residualbodily concerns. Socratesrepeatedlydescribes
the practice of separatingsoul from body whil still alive as a practice the philosopherwill
engage in 'so far as it is possible' (iKa' ooov wovaral:64e5, 65c5-9, 67c5). Thereis a scale of
achievement. Some souls are more inextricablylinked with the body they currentlyinhabitthan
others, and the effects of the degree of contaminationby bodily concerns taint the soul after it
has been released from the body at the end of a life.
This is the final attemptedresolutionof the worriesI canvassedearlierover the role of divine
sanction. The clear implication is that god does not allow us finally to separatebody and soul
throughsuicide (and make that allowance clear througha signal) until the philosopherhas sufficiently purified the soul of bodily concerns.

Suicide before that time is 'unholy', perhaps

because if the soul is released at that point it is still polluted by incarnation. That pollution is
potentiallyharmfulto the soul itself, since such bodily taint compromisesits chances of success
in the cycle of reincarnationor - as it is expressedin otherdialogues- the vision of the Forms.33
Let us return,then, to Cebes' second ource of astonishmentwhich first expressedthe inconsistency Socratesmust overcome.
Perhapsit seemsastonishingto you if forthosemenforwhomit is betterto be deadthanaliveit is not
holy forthemto do themselvesa favour,butthattheymustwaitfor someotherassistant.(62a5-7)
Now there is a new difficulty. If god sends a divine sign only when the philosopheris sufficiently purified to die properlyit is no longer the case that there are people for whom it would
be better to be dead but who are not permittedto die. Unless the soul is properlypurified it is
not betterto be dead, it is in fact betterto stay alive and try to purify it. In retrospect,therefore,
it looks as if Socrates must mean that there are in fact no exceptions to the rule that one should
do what is of benefit to oneself. Those who thinkthat it would be betterto die thanbe alive, and
who do not receive a divine sanction arejust mistaken in their assessment- like Cleombrotus.
The benefit of this explanationis that it no longer requiresdivine allowance to be a whim. It is
only grantedas and when a soul is sufficientlypurified,and thereforethe gods are indeedbenevolent. Theirdivine prohibitionon othercases of suicide is a preventativemeasurewhich tries to
ensure that as few people as possible die like Cleombrotuswith polluted souls.
There is one significantobstacle to this reading. At 62a Socratesdoes not express any doubts
aboutthe fact that for the particularpeople underconsiderationit is in fact betterfor them to be
dead than alive (indeed as I mentioned above this assessment seems to be emphasizedby repetition). This premise is what createdthe dissonancewith the benevolentgods' disapprovalof suicide. If the gods are indeed benevolent, then it ought to be the case that on reflection it is not
betterfor these people to die than be alive - if it were, the gods would send a sign to that effect.
The obstacle might be evaded if not removed entirely by emphasizing that Socrates is at this
point offering to Cebes a diagnosis of Cebes' own view, and thereforeneed not, I think,be himself endorsingthis assessment. It should also be noted thatthe reasongiven for the absoluteprohibition on suicide at that early stage was that suicide is 'unholy'. It may now have been
33Considerthe consequences for the fate of the soul
as related in the myth at 113dlff.



revealed that suicide is not unholy provided the gods will it and indicate that it is to be done.
Divine approvaland commandon specific occasions will cancel the groundsoriginally offered
for the blanketprohibition. Nevertheless, it must be agreedthat it would certainlyhave clarified
the argumentif it had been made explicit at some laterpoint that suicide is in fact, and contrary
to Cebes' initial astonishment,prohibitedonly for those for whom it is not beneficial - as must
be the case if the argumentis to be consistent.
The upshot of all this is that the stance taken in the Phaedo on the rationalityof suicide is based
upon yet another version of position C2 which I outlined at the very beginning. Despite
Platonism'semphasison the superiorityof the soul over the body, of the intelligible over the perceptible and of the discarnateover the incarnate,the ability of the soul and body to interactand
the fact that incarnationitself pollutes the body makes it impossible for a Platonistto claim on
the basis of the Phaedo that being dead is always betterthan being alive. Only when a soul has
been properlypurified is it rationalto separatethe soul from the body, not before. This is not
merely an arbitrarywhim of the gods - it is for the good of our souls, and thereforeour real
selves. There is also no reason to conclude that those people who are not philosophersare better off dead. First, their impuresouls may not fare very well once released. Also, suicide cuts
off the only possibility of improving this fate - namely, beginning to think philosophically.
According to the Phaedo, therefore,the Socratic maxim that 'an unexaminedlife is not worth
living' is not a call for all non-philosophersto commit suicide (so Cleombrotuswas indeed mistaken);it is a call to begin examiningour lives. 34
Magdalene College, Cambridge

34 This view should be contrasted with that of anyone would continue living once this is realized.
Brickhouse and Smith (1994), who on the basis of the Brickhouse and Smith note the religious prohibitionon
Apology, Crito, and Gorgias conclude (211) that suicide (211 n.6), and also speculatethatsuicide might be
'Socratesbelieves that in some way or anothereveryone itself an unjust- and thereforeharmful- act. Compare
will be betteroff dead' (my position Cl). However,they Cooper (1999) 535 on the Stoics: 'So it is no improveagree that some people may be better off in death than ment in the goodness, or diminishmentof the badness, of
others. Philosopherswill be especially well-off, whereas an agent's life to shortenthe time he is morally bad; the
the vicious will only be benefited because in death they only improvementin its goodness or diminishmentin its
can commit no more vicious acts and therefore can do badness there can be is for him to take steps to make a
their souls no more harm. The question remains why betterperson of himself.'



Bostock, D. (1986) Plato s Phaedo (Oxford)

Brickhouse, T.C. and N.D. Smith (1994) Platos
Socrates (Oxford)
Burger, A. (1984) The Phaedo: A Platonic
Labyrinth(New Haven)
Cooper, J. (1999) 'Greek philosopherson euthanasia and suicide', in his Reason and Emotion:
Essays in Ancient Moral Psychology and
Ethical Theory (Princeton) 515-41 (originally
printed in B.A. Brody (ed.), Suicide and
Euthanasia (Dordrecht1989) 9-38)
Englert, W. (1994) 'Stoics and Epicureanson the
natureof suicide',Proceedingsof theBostonArea
Colloquiumin AncientPhilosophy10, 67-98
Feldman, F. (1992) Confrontations with the
Reaper: A Philosophical Study of the Nature
and Valueof Death (Oxford)
Frey, R.G. (1980) 'Did Socrates commit suicide?',
in Pabst Battin and Mayo (1980) 35-8
Fischer, J.M. (1993) The Metaphysics of Death
Furley, D.J. (1986) 'Nothing to us?', in M.
Schofield and G. Striker (eds.), The Norms of
Nature (Cambridge)75-91
Gallop, D. (1975) Plato: Phaedo (Oxford)
Glover, J. (1977) Causing Death and Saving Lives
Griffin, M. (1986) 'Philosophy, Cato and Roman
suicide: 1', Greece and Rome 33, 64-77
Hoven, R. (1971) Stoicisme et Stoiciens face au
probleme de l'au-deld (Paris)
Huffman, C. (1993) Philolaus of Croton:
Pythagoreanand Presocratic (Cambridge)
Nagel, T. (1979) 'Death', in his Mortal Questions

Pabst Battin, M. (1982) Ethical Issues in Suicide

(Englewood Cliffs)
- and D.J. Mayo (eds.) (1980) Suicide: The
Philosophical Issues (London)
Riginos, A.S. (1976) Platonica: The Anecdotes
Concerning the Life and Workingsof Plato
Rist, J.M. (1969) Stoic Philosophy (Cambridge)
Rowe, C.J. (1993) Plato: Phaedo (Cambridge)
Sedley, D.N. (1993) 'Chrysippuson psychophysical causality', in J. Brunschwig and M.
Nussbaum (eds.), Passions and Perceptions:
Studies in Hellenistic Philosophy of Mind
- (1995) 'The dramatis personae of Plato's
Phaedo', in T.J. Smiley (ed.), Philosophical
Dialogues (Oxford)3-26
Todd, S. (1993) The Shape of Athenian Law
Warren, J.I. (2000) 'Epicurean immortality',
OxfordStudies in Ancient Philosophy 18, 23161
Williams, B. (1973) 'The Makropuloscase: reflections on the tedium of immortality', in his
Problems of the Self (Cambridge)82-100
-(1976) 'Persons, character and morality', in
A.O. Rorty (ed.), The Identities of Persons
(Berkeley) 197-216
Williams, G.D. (1995) 'Cleombrotusof Ambracia:
interpretationsof a suicide from Callimachusto
Augustine', Classical Quarterly45, 154-69
Windt, P.Y. (1980) 'The concept of suicide', in
Pabst Battin and Mayo (1980) 39-47