American Philological Association

Contributions of the Herculanean Papyri to Our Knowledge of Epicurean Logic

Author(s): Phillip H. de Lacy
Source: Transactions and Proceedings of the American Philological Association, Vol. 68 (1937), pp.
318-325
Published by: The Johns Hopkins University Press
Stable URL: http://www.jstor.org/stable/283272
Accessed: 12-05-2015 02:26 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.

The Johns Hopkins University Press and American Philological Association are collaborating with JSTOR to digitize,
preserve and extend access to Transactions and Proceedings of the American Philological Association.

http://www.jstor.org

This content downloaded from 143.107.252.166 on Tue, 12 May 2015 02:26:09 UTC
All use subject to JSTOR Terms and Conditions

318

Phillip H. De Lacy

[1937

oftheHerculaneanPapyrito our
XXIV.-Contributions
KnowledgeofEpicureanLogic
PHILLIP

H. DE

PRINCETON

LACY

UNIVERSITY

The Herculanean papyri indicate that the Epicureans did not eliminate
the study of logic from their philosophy, but rather that they constructed
for the firsttime an empirical logic built upon a careful study of the logical
speculation available to them. They used their new logic as a basis for
their entire system of philosophy, and they referred to it constantly in
works dealing with physics and ethics. This empirical logic is a significant
contribution of the Epicureans to the development of ancient thought.

Though the Herculanean papyri have little literaryor historical value, and thereforehave persistentlydefied integration with the traditional scheme of ancient culture, they
occasionallyreveal to us certainmovementsin ancientthought
which necessitate a revision of our long established views
about the Classical world. The most significantcontributions of these papyri lie in the sphere of philosophy,and
especially revolutionaryis the material they presentbearing
upon Epicurean logic. Previous to the acquisition of this
material historiansof philosophy,influencedby such ancient
sourcesas Cicero,' had generallysupposed that the Epicureans
despised the pursuitof dialectic,rejectingeven such seemingly
essentiallogical principlesas definitionand division. In place
of logic, according to this view, the Epicureans developed a
disciplinecalled "canonic" which asserted that perceptionis
the only source of knowledgeand set forthcertain practical
rules of philosophicalmethod.2
The Herculanean papyri,on the other hand, have revealed
a positiveside of the Epicurean attitude toward logic. They
reveal that the Epicureans rejecteddialecticin orderto estab1 E.g. De Finibus 1.22; 2.18; Lucullus 97.

A recent defense of this traditional point of view occurs in C. Bailey The

Greek Atomists and Epicurus (Oxford, Clarendon Press, 1928), 232-235.
2

This content downloaded from 143.107.252.166 on Tue, 12 May 2015 02:26:09 UTC
All use subject to JSTOR Terms and Conditions

Vol. lxviii]

HerculaneanPapyri

319

lish a new logic of theirown.3 That they preferredthe name

"canonic" is only a matterof terminology,for the problems
that they discussed under that heading were essentially of
the same type as those whichthe Stoics includedunderlogic.4
The Herculanean papyrifurtherinformus that the Epicurean
logiciansstudied with care the worksof philosophersof other
schools and profitedfromtheirspeculations. Obligations to
Aristotelianlogic are especially numerous,as will be pointed
out later. A considerablenumberof the Herculanean papyri
bear upon Epicurean logic, of which the most importantis
Philodemus' HE-p'2;?tWzoc,v, On Methodsof Inference,a document devoted entirely to the defense of Epicurean logic.5
Critics who adhere to the traditionalview of the negative
characterof Epicurean logic tend to discreditthis papyrus as
the work of an heretical Epicurean.6 It must be admitted
that Philodemus'workcarrieslogical speculationfurtherthan
anythingwe findin the extant worksof Epicurus; yet at least
two of Philodemus' authoritiesfor Epicurean logic, Zeno of
Sidon I and Demetriusof Sparta,8were apparentlyEpicureans
of good standing. Also, as we shall see, the rudiments,if not
the fullelaboration,of Philodemus' logic may be foundin the
3 Cf. Sextus EmpiricusAdversus Mathematicos 7.14-15.
4 E.g. the analysis of sign relations,the meaningof words,and the methods
and validity of inference. Epicureans and Stoics, of course, solved these
problemsquite differently.
5Edited by T. GomperzHerkulanische Studien I (Leipzig, Teubner, 1865).
The exact form of the title is a matter of dispute. Other papyri dealing
withEpicurean logic are discussedby W. Crdnert,"Die AozywcaZ-rT,uarades
Chrysippusund die tibrigenPapyri logischenInhaltsaus der Herculanensischen
Bibliothek,"Hermesxxxvi (1901) 568-579. (I use Roman numeralsto refer
to columnsof papyri.)
6 E.g. Bailey Gk. Atomists 259, note 1. More favorableestimatesare given
by Gomperz op. cit., preface; R. PhilippsonDe Philodemi Libro qui est llpt'
dcKa 2ra ,ue x(,o et Epicureorum Doctrina Logica (Berlin, 1881); and A.
277yet.
Ed. Chaignet Histoire de la Psychologie des Grecs (Paris, Hachette, 1889) II
375-376.
7 Sign. XIX 5; cf. Cic. N.D. 1.59; Fin. 1.16.
8 Sign. XXVIII
13; cf.W. CrbnertKolotes und Menedemos (Leipzig,Avenarius,
1906) 100-125.

This content downloaded from 143.107.252.166 on Tue, 12 May 2015 02:26:09 UTC
All use subject to JSTOR Terms and Conditions

320

Phillip H. De Lacy

[1937

writingsof Epicurus. For these reasons Philodemus' logic

mustbe considereda legitimatepart of Epicurean philosophy.
The Epicureans, as Philodemusrepresentsthem,made the
firstthoroughgoingattempt to establish an empiricallogica logic freedfromany formala priorielement,by the reduction of every propositionto immediateexperience. The empirical logic of the Epicureans was primarilyopposed to the
Stoics, who placed logic on a completelyformallevel. The
Stoics directed in part the course of Epicurean logic, principally by raisingobjections which the Epicureans endeavored
to answer.9 In constructingtheirlogic the Epicureansutilized
the partial advances of previous thinkersin the directionof
empiricism. They took from the Empirical Physicians the
basic principlesof empirical method, observation (avc-o4l)a),
and o-ropla),and inference
record of past experience (1up'p-q
Moreover,
fromsimilarto similar(uIrT6aoLSaKac' ,uoLotrlra).10
they took from Aristotle those elements in his philosophy
whichwere adaptable to empiricism. Aristotlehad discussed
empiricalmethodprimarilyas a rhetoricaldevice, ratherthan
a genuine part of philosophy;1 neverthelesshe formulated
several logical principlesthat could be used by the Epicureans
withlittlechange:
1. All knowledgeabout reality must be derived fromperception of objects.'2 This statement is in accord with the
view that the concreteobject is the only primarysubstance
(Cat. 2 a 11-19). All knowledgeof universalsmustbe derived
fromthe observationof the qualities of particulars.'3
9 See below, p. 321.
10These three principles, implicit in the Hippocratean works (e.g., Kar'
9, llapa-yyEXta 1, ]IEpI ALalT7lS
'I-Tpe6OV1, lIEp' TEXvxqs7, HLEpI4bacos 'AvOpc,wrov
1), are expounded at length in Galen's Subfiguratio Empirica. These three
principles probably formed the basis of Nausiphanes' Tripod, and from there
Cf. Philippson De Philodemi Libro 52-55.
were taken over into Epicureanism.
"1Cf. Rhetoric 1402b 12-1403a 16; Topica lO5a 20-33, 108a 7-17; A.Pr.
70a 3-70b 38.
12 De An. 432a 3-14.
Cf. the Epicurean position stated in Sext. Emp. M.
7.216.
13A.Po. 81a 38-81b 9, 97b 7-25; Metaph. 980b 28-981a 7; Top. 108b 9-12.
Philodemus expresses this view in Sign. xvi-xvii.

This content downloaded from 143.107.252.166 on Tue, 12 May 2015 02:26:09 UTC
All use subject to JSTOR Terms and Conditions

Vol. lxviii]

HerculaneanPapyri

321

2. Truth lies in the correspondenceof propositionswith

concretefacts.14
3. Propositionsabout the futureare neithertrue nor false.'5
4. Perceptionsare always, or nearly always, true. Falsity
lies in judgment.'6
5. Thought is psychologically derived from perception
throughimaginationand memory.17
6. Language is conventional in words and structure,but
the meaningsof words are the same forall men, because they
referto elementscommonto the experienceof all men.'8
7. Generictermsare extensionalin reference,forthereis no
Singleontologicalentityunderlyingthemto whichtheyrefer."9
In addition to these principles the Epicureans took from
Aristotlea numberof logical terms,e.g. fbra-yw-y',KacOOXOV, KacO

The Epicureanswerefamiliar,
as
well,withStoic logicalterms,such as ac'aO-KEV?7.Indeed,the
EKao-To', and manyothers.

Epicureans, in spite of their hostility,owed much to the

Stoics; for they were compelled by the Stoics to discuss the
crucial problem of the validity of inductive inference(E7rStoic criticismwas not silenced by the elaboration
Xo'yuT/.os).
of empirical epistemologyand methodology,for the Stoics
insistedthat unless inductiveinferencecan establishthe essential nature of things (the nature of man qua man, abOpwwrosn7
ahOpwcros),it can never prove that any inferenceis necessarily
valid.20
14 Cat. 14b 15-22; cf. the Epicurean stress on verification: Sext. Emp. M.
7.211-212.
15 Int. 19a 18-22; cf. Cic. Luc. 97; N.D. 1.70.
16 De An. 418a 11-16, 427b 11-14, 428b 18-19.
The truth of perception is
basic to Epicureanism: Sext. Emp. M. 7.210; 8.9, 63-64; Cic. N.D. 1.70.
17 A.Po.
99b 32-lOOa 9; De An. 432a 3-14; Sext. Emp. M. 7.216-226.
Cf.
the Epicurean doctrine of 7rpoXwI4s: Chaignet Psych. des Grecs 362-364; Bailey
Gk. Atomists 245-247.
18 Int. 16a 27-29; cf. Epicurus' letter to Herodotus, Diogenes Laertius 10.7576.
19 Cat. 2a 34-35, 3b 10-23; Int. 17a 39-17b 1.
The Epicurean emphasis on
extension appears in Plutarch Adversus Coloten 22; Sext. Emp. M. 8.13, 258;
D.L. 10.33, 37-38.

20 Sign. II-iv.

This content downloaded from 143.107.252.166 on Tue, 12 May 2015 02:26:09 UTC
All use subject to JSTOR Terms and Conditions

Phillip H. De Lacy

322

[1937

Philodemus,in answeringthe Stoics, maintainsthat inductive inferenceis necessarilyvalid because it is based on a

knowledgeof the essential natureof things. In the course of

notions(rpoX/'7Ets) arisein themindcorresponding

experience

to the various kinds of objects that have been experienced.

These notions, being taken directly from experience, are
reliable indices of the nature of the objects of experience.2'
They are used in reasoningas standards of inconceivability
(ro a'&av6Orov),and as such theyestablishthe essentialqualities
of things. The essential qualities of a man, forinstance,are
those qualities withoutwhich a man cannot be conceived to
be a man. The test of inconceivability,therefore,by fixing
the definitelimitsof variation in the nature of objects, provides a ground wherebyinductive inferencesmay be necessarilytrue.22 Having once explained to theirown satisfaction
the possibilityof necessary truthson an empirical level, the
Epicureans proceeded to claim that all necessary truthsare
empirical.23 An exclusively formalanalysis of propositions,
they argued, is impossible,for language is meaningfulonly
when related to the world of sense experience throughthe
medium of the empiricallyderived wpoXt4EMs.24 The whole
content and structureof thought are directly referableto
sense experience. Hence all attemptsat purelyformalanalysis,includingdefinitionand division,presupposean inductive
logic.
The reconstructionof Epicurean logic made possible by
the Herculanean papyri gives new meaning to all parts of
Epicurean philosophy. Even the writingsof Epicurus himself indicate familiaritywith the basic parts of this empirical
the empiricalapproach
logic: the use of inductiveinference,25
to meaning,2"and the appeal to inconceivability.27The in21 Cf. above, note 17.
22 Sign.

xii, xxi.

23 Sign.

viII-Ix.

above, p. 321, nos. 6 and 7.

25 D.L. 10.32, 59, 72-73, 133.
26 D.L. 10.37, 75-76.
See below, note 28.
27 D.L.
10.47, 60, 68, 71, 97.
24 See

This content downloaded from 143.107.252.166 on Tue, 12 May 2015 02:26:09 UTC
All use subject to JSTOR Terms and Conditions

Vol. lxviii]

HerculaneanPapyri

323

terestin empiricallogic,clearly,began with Epicurus himself,

and the logic of Philodemus is a natural extensionof ideas
foundin Epicurus' own works.
The Herculanean papyri also reveal that the Epicureans
applied theirempiricallogic to all fieldsof inquiry. In HepL
DqyEt6oTecolxvII-xviII

showshowtheatomictheory
Philodemus

is empiricallyderived from observation. In the new fragoccur sectionsdealing with the

mentsof Epicurus' HepLcIbovecos
and
proper use of words and the valid formsof inference,28
Epicurus uses frequentlytechnical terms of empirical logic:
ra-yco-yn
Epicurean the29vaXoyLa,30 a'&avo6rov,3' btXo7to..32
similar
of
involves
a
ology
application empirical logic. The
gods are known either by immediate experience or by analogical inferencefromimmediateexperience.33 Analogies betweenmen and god.sappear frequentlyin Philodemus'works,34
and the familiarterms of Epicurean logic recur.35 Finally,
the Epicureans treat ethics as an empiricalscience consisting
of two parts, the nature of ultimatevalue and the means by
whichthat value may be realized.36 In establishingthe nature
of ultimatevalue the Epicureans appeal not to religioussanction, moral law, or any other non-empiricallegislativeprinciple,but to the actual concreteexperienceof men. Ultimacy
28 Epicurus IHept4)bocos 28 (ed. A. Vogliano Epicuri et Epicureorum Scripta
in Herculanensibus Papyris Servata. Berlin, Weidmann, 1928) fr. 5 I, III, IV.
29 Epic. Nat. 28 (Vogl.) fr. 1 ii, 2, 5 vi.
30 Epic. Nat. (Herc. Vol.' x) IX, XV.
31 Epic. Nat. (H.V.l x) Ix.
32 Epic. Nat. 28 (Vogl.), fr. 1 iv, 5 vii, ix.
33 Cf. R. Philippson, "Zur Epikureischen Gotterlehre" Hermes LI (1916)
568-608; "Uber die Frommigkeit," Hermes LVI (1921) 355-410; "Die Gotterlehre der Epikureer," Rhein. Mus. LXXXIII (1934) 171-175. Bailey Gk. Atomists
590-594 criticizes Philippson's interpretation.
34E.g., Phld. Sign. XXII 15-28; E. Bignone, "Philodemea" Riv. di Filol.
XLVII
(1919) 414-422; Phld. Ilept 9eCov I (ed. H. Diels Abh. Preuss. Ak. Wiss.,
1915, Philos.-Hist. Klasse, no. 7) II, VII, XIII; Phld. Ilepi Ebaoflelas (ed. T.
Gomperz Herk. Stud. ii 1866) 115. Cf. Cic. N.D. 1.24, 46; Cronert Kol. und
Men. 112-113.
35 E.g. 4sTafiaLvew: Phld. Piet. (Gomperz) 6; &&Lav627TOV: H. Usener Epicurea
(Leipzig, Teubner, 1887) fr. 356. Cf. Cic. N.D. 1.25, 30, 36,
36 Cic. Fin. 1.32-33, 40-43; Tusc. 5,95-96.

This content downloaded from 143.107.252.166 on Tue, 12 May 2015 02:26:09 UTC
All use subject to JSTOR Terms and Conditions

324

Phillip H. De Lacy

[1937

in value, like necessity in logic, is found in immediate experience. This is the point of view, it seems to me, that
underlies the Epicurean defense of psychologicalhedonism;
for,say the Epicureans, both men and animals naturallyseek
pleasure.37
In the fieldof practical ethics,or the study of the means of
attaining the greatest amount of pleasure, the Herculanean
papyri provide abundant information. Even here the empirical approach is dominant. Only through careful and
extensive observation, joined with a memoryof similarities
and an empiricalknowledgeof consequences,
and differences
is one able to avoid all thingswhichdo not lead to happiness.38
The purpose of rhetoric,according to Philodemus,should be
to advise the best course of action in any given situation,and
thereforerhetoricshould have a scientificmethod of investigatingmoralconduct.39 Philodemushimselfwroteextensively
on anger,pride,wealth,flattery,and kindredmoral questions.
His works,thoughfragmentary,
reveal unmistakablyhis empirical approach. He suggestsin one place the possibilityof
determiningempirically that which is most detrimentalto
friendship.40In his work on anger he declares his intention
the essenof establishingby empiricalanalysis (EirXo-y1caaOat)
tial nature of this evil.4' In the Hepl OiKOQO,LLasPhilodemus
upholds the empiricalreferenceof terms,in oppositionto the
method of dialectic.42 Throughout his ethical works PhiloInferences involving analogies
37 Cf. S.E. M. 11.96; Cic. Fin. 1.30; 2.32.
between men and animals are also found in Phld. De Diis I (Diels) xI-xv; Sign.
xxii 15-28; Lucretius iII 296-307; Polystratus IHpL 'AXOyov KaTaopovf'&Ecos
(ed. C. Wilke, Leipzig, Teubner, 1905) fr. 5-7.
38 Phld. IlEpl
TIOTOPLKijS (ed. S. Sudhaus, Leipzig, Teubner, 1892-1896), ii
30-31.
39 Phld. Rh. (Sudhaus) i 168, 246-247, 254; ii 30-48.
Sext. Emp. M. 2.49
gives a similar view of the function of rhetoric.
40 Usener Epic. fr. 552.
41 Phld. IIepl 'Op-yis (ed. C. Wilke, Leipzig, Teubner, 1914) VII; cf. XLV-XLVI.
42 Phld. IIIEpl OKovo.Ulas
(ed. C. Jensen,Leipzig, Teubner, 1906) xx-xxi;
cf. the criticism of Plato's style in Phld. llEp' KaKWv 10 (ed. C. Jensen, Leipzig,
Teubner, 1911) xxII-xxIIi.

This content downloaded from 143.107.252.166 on Tue, 12 May 2015 02:26:09 UTC
All use subject to JSTOR Terms and Conditions

Vol. lxviii]

HerculaneanPapyri

325

demus continuallyuses the same technicaltermsof Epicurean

logic: analogy, inconceivability,and inductiveinference.43
Thus the Herculanean papyri give us extensivematerialon
the nature of Epicurean logic and its application to all parts
of Epicurean philosophy. This new material not only increases the significanceof Epicurean philosophyin general;
it also enables us to appreciate the care with which the Epicureans studied the thoughtof other philosophersthat they
mightimprovetheirown philosophicalsystem. To thehistory
of logic the Epicureans contributedthe firstserious attempt
to formulatean entirelyempirical logic. It was indeed unfortunatethat this logic had so littleinfluenceon subsequent
philosophythat its true naturewas almost unknownuntil the
recoveryof the Epicurean worksfromHerculaneum.
43 Analogy: Phld. IlepL HXOUTOV
(ed. M. D'Amelio Di Alcuni Trattati Epicurei
sulla Ricchezza, Naples, 1926) 19; Ir. (ed. R. Philippson, " Philodems Buch uiber
den Zorn," Rh. Mus. LXXI, 1916), 432, 434.
Inconceivability: Phld. Ir. (Wilke) vi, (Philippson) 442; Usener Epic. fr.
606. Cf. Cic. Fin. 2.48-49.
Inductive inference (&t7rtoyX,uo's):Phld. Ir. (Wilke) iII, Iv; Div. (D'Amelio)
21, (Usener Epic.) fr. 45; HI4ptKoXaKEfas (ed. V. De Falco, "Appunti sul IHsp'
KoXaKEdasdi Filodemo," Riv. Ind.-Gr.-It. x, 1926) XI.

This content downloaded from 143.107.252.166 on Tue, 12 May 2015 02:26:09 UTC
All use subject to JSTOR Terms and Conditions