Galenus Stromata VIII

7/27/2019 Galenus Stromata VIII

1/34

Vigiliae Christianae 65 (2011) 343-375 brill.nl/vc

VigiliaeChristianae

Galenus Christianus? Te Doctrine of Demonstrationin StromataVIII and the Question of its Source

Maty HavrdaCentre or Patristic, Medieval and Renaissance exts, Palack University

Olomouc, Czech Republic

AbstractTe article is a source-critical study o the doctrine o demonstration in the so-calledeighth book oStromataby Clement o Alexandria. Ater an overview o the doctrine,as presented in Str. VIII 3,1-15,1, it examines parallels in philosophical literature,especially in the writings o Galen. Tis examination brings to light correspondences(not all o which have been previously noted) whose number and proximity opensthe question o the relation between Galen and the source o Stromata VIII. Aterconsidering three explanations to account or these similarities, the article proposes

that Galens lost writing on demonstration could be Clements source.

KeywordsClement, Galen, demonstration, method o discovery, embryology

1. Te problem of Stromata VIII

Te collection o texts preserved in Codex Laurentianus V,3 under thetitle has always puzzled Clements readers. It seemsthat even in antiquity people were not quite sure where the book ends,the possibilities ranging rom mere sixteen paragraphs o Sthlins editionto a version which included at least the Eclogae propheticae.1 In the Lau-rentianus, the manuscript on whose testimony modern editions oStro-matadepend, the end o the eighth book is not indicated either. But inscholarly debate it is usually placed ater a series o philosophical discus-sions regarding the problems o proo, inquiry, sceptical , division,

1) C. the evidence collected by Teodor Zahn, Forschungen zur Geschichte des neutesta-mentlichen Kanons, Bd. III: Supplementum Clementinum (Erlangen 1884) 28.


2/34

344 M. Havrda / Vigiliae Christianae 65 (2011) 343-375

defnition, categories and causes, beore the start o a completely dierentsection introduced as compendia o Valentinian doctrines.2

Te uncertainty about the scope o the book is partly due to the actthat the text oStromataVIII is obviously a ragment or rather a series oragments whose interconnecting link or continuity with the precedingStromata is neither explicit nor clear. Tere have been dierent explana-tions o the origin o the text. Out o several solutions proposed since the17th century we will mention just two that set the outlines o the con-temporary debate.3 In a discussion covering all material placed ater Stro-mata VII in Codex Laurentianus, Teodor Zahn suggests that theseragments are abbreviations and excerpts made rom Clements fnishedwork, the continuation o the extant Stromata.4 In the frst study ocusedon the philosophical contents oStromataVIII, Hans von Arnim proposesa dierent view, according to which the material consists partly o anelaborated, but unfnished text o the eighth book, or excerpts made romit, and partly o excerpts made by Clement himsel rom other sources inpreparation or the planned continuation.5

2) In this sense, too, we will speak about Stromata VIII in this paper. For a detaileddescription o Codex Laurentianus V,3 c. Otto Sthlin in Clemens Alexandrinus, Bd. I

(GCS12, Leipzig 1905) xxxix-xlii.3) A precise overview o the various explanations is provided by Antonio Servino, Clem-ente Alessandrino: il problema di Stromata VIII, Quaderni del Dipartimento di flologia,linguistica e tradizione classica Augusto Rostagni 17 (2001) 97-104.4) Zahn, Forschungen, 104-130, esp. 117-119. Zahns hypothesis is taken up and developedby Pierre Nautin, La fn des Stromateset les Hypotyposesde Clment dAlexandrie, VigiliaeChristianae30 (1976) 268-302. Te main dierence o Nautins approach is the ollowing:Whereas Zahn derives all the ragmentary material ater StromataVII in Laurentianus romthe eighth book oStromata, Nautin suggests that the excerpts rom the eighth book endwith Str. VIII 24, i.e., beore the start o the section on causes, while what ollows in Lauren-

tianus comes rom the lost Hypotyposeis. Tis part o Nautins thesis is criticized by Alain LeBoulluec, Extraits duvres de Clement dAlexandrie: La transmission et le sens de leurtitres, inAlexandrie antique et chrtienne. Clment et Origne(Paris 2006) 115-117, who,apart rom presenting textual arguments, shows that there is a thematic continuity betweenthe chapter on causes (Str. VIII 25-33) and the preceding paragraphs 17-24, which, in turn,are thematically linked with the rest o the book (1-16). Details o Nautins thesis andarguments produced in its avour cannot be discussed in this paper. For the present state othe debate c. Bogdan G. Bucur,Angelomorphic Pneumatology. Clement o Alexandria andOther Early Christian Witnesses (Suppl. to Vigiliae Christianae 95. Leiden/Boston: Brill,2009) 10. C. also Andrew C. Itter, Esoteric eaching in theStromateis o Clement o Alexan-dria(Suppl. to Vigiliae Christianae 97. Leiden/Boston: Brill, 2009) 59-76.5) Hans von Arnim, De octavo Clementis Stromateorum libro (Rostock 1894) 9, 11-12. VonArnim seems to locate the end o the fnished part o the book at 16. Sthlin, while


3/34

Galenus Christianus? 345

Te intricate cobweb o issues called the problem o Stromata VIIImay be summarized by two questions: (1) What is the role o this text in

the context o Clements extant writings? (2) What are its philosophicalsources? Te ollowing paper is an attempt to reenter the debate rom theperspective o the source-critical approach. It ocuses on the discussion othe doctrine o demonstration in Str. VIII 3,1-8,3 and a closely relatedexposition o the method o inquiry in 8,4-15,1, sections that togethercomprise more than one third o the book.6 Ater an overview o theircontents, the paper explores parallels that might help us characterize, orpossibly determine, the philosophical source o these sections.7

2. Str. VIII 3,1-15,1: Overview

a. Christian prooemium (ch. I, 1,1-2,5)

Clements elaboration o the theory o demonstration is preceded by twoparagraphs in which he outlines the project o scientifc research asinquiry o questions based on the Scriptures ().8 HereClement describes scientifc demonstration as a method by which the

accepting von Arnims assessment about the second part o the book, suggests that eventhe frst part consists o Clements own excerpts (GCS 12, xli-xlii). Von Arnims distinc-tion between the two parts is criticized by Wilhelm Ernst, De Clementis Alexandrini Stro-matum libro VIII. qui ertur (Gttingen 1910) 52. Servino, Clemente, 97, note 3,100-102, accepts von Arnims distinction, but locates the breaking point between the twoparts at the end o 5,5 (GCS 17: 82,26).6) Te frst commentary on the section o demonstration is provided by Ernst, De Clem-entis, 11-13, 15-24. C. also Reginald E. Witt,Albinus and the History o Middle Platonism

(Cambridge 1937) 32-35; Salvatore R.C. Lilla, Clement o Alexandria. A Study in ChristianPlatonism and Gnosticism (Oxord: OUP, 1971) 120-136; Friedrich Solmsen, Early Chris-tian Interest in the Teory o Demonstration, in W. den Boer (ed.), Romanitas et Christi-anitas(Amsterdam 1973) 281291; Silke-Petra Bergjan, Logic and Teology in Clemento Alexandria. Te Purpose o the 8th Book o the Stromata,Zeitschrit r Antikes Chris-tentum 12 (2008) 396-413. Str. VIII 9,1-15,1 is discussed by Jaap Manseld, Doxographyand Dialectic. Te Sitz im Leben o the Placita,ANRWII 36.4 (1990) 3184-3193.7) Tis paper is a partial outcome o a research conducted at the University o Crete,Rethymno, in the Summer Semester o 2009. I am grateul to A.S. Onassis Public BeneftFoundation or supporting the visit and to the sta o the Department o Philosophy andSocial Studies at UOC or their hospitality. I greatly benefted rom comments made onearlier drats o this paper by Alain Le Boulluec, Orna Harari and Giorgos Karamanolis.8) C. Str. VIII 2,1.


4/34


Christian researcher, relying on the Scriptures on the one hand and thecommon notions on the other, may discover the truth.9 Te discussion

is ramed by a polemic against eristic and sceptic opponents.10

b. What is demonstration? (ch. II, 3,1III, 8,3)

Having noted that research is conducted by means o demonstration,Clement advances to discuss the problem o demonstration itsel. Despitethis thematic continuity, there is a clear break between the second and thethird paragraph. Starting with the third paragraph, Clements Christianinterests recede to the background, only to come to the surace again

much later in the text in the orm o brie marginal comments.11

Also, thebeginning o the third paragraph presupposes some earlier discussion thatis missing in what precedes it in our text.12 Clement starts by proposing a

9) Str. VIII 2,4-5. By common notions () Clement presumably meansconsensual views o Christian community, such as beft even simple-minded believers; c.Str. VII 95,9, where Clement compares the distinction between simple believers and gnos-tics to that between laymen () and specialists () who create shapes whosebeauty surpasses the common notions ().In this comparison, the common notions apparently correspond to the concord or con-

ession () o the believers (c. Str. VII 90,1-2; Le Boulluec, La notion dhrsiedans la littrature grecque IIeIIIesicles, II[Paris 1985] 367-70), i.e., to a standpoint Clem-ent also describes as common aith (); or the latter concept c. Lilla, Clem-ent, 136.; Le Boulluec, in Clment dAlexandrie, Stromate V, ome II (commentaire), SC279 (Paris: Cer, 1981) [henceorth SC 279], 20.10) C. Str. VIII 1,1-2; 2,5; Le Boulluec, La notion dhrsie, 382-85.11) Te only other passages that, in my view, betray a Christian hand are ound in the fnalchapter on causes: a theological comment in Str. VIII 29,3-6 (GCS 17: 99,1-12), a bibli-cal gloss in 30,4 (99,25), an allusion to Gen 1:1 in 28,5 (98,14.) and arguably the peda-gogical examples in 25,4 (96,2-5) and 30,2 (99,19.). C. also 5,5 (82,256).12)

Te abrupt beginning o the eighth book ( .) suggests to many a reader thatClement himsel had preaced his text with an introduction that was later lost; c. vonArnim, De octavo, 9, and Sthlins note ad GCS 17: 80,3. Ernst, De Clementis, 53, conteststhis interpretation by pointing out that the frst page is inscribed in themanuscript: Since the inscription . . . is preserved, it would be strange i the beginning o thetreatise was corrupt. Although it is undoubtedly true that the beginning o the eighth bookis abrupt (against Zahn, Forschungen, 115.), it could also be interpreted as an indication thatthe text designated as chapter one in modern editions was already composed as a ragment,perhaps a note on Clements antisceptical source. Str. VIII 1,3-2,5 (GCS 17: 80,11-81,8)may be described as a commentary on Matt 7:7 (quoted in 1,2 [80,9.]); c. Nautin, La fndes Stromates, 268, 291. Te quotation o this verse, so important or Clement (c. Le Boul-luec, La notion dhrsie, 385-89), could have been prompted by a sentence or two in hissource, possibly by the initial clauses o 1,1 and 1,2 (GCS 17: 80,3-4.6-8).


5/34


method to arrive at the starting point o such a doctrine ( . . .), namely to explain the proposed

word by an account so clear that all who speak the same language will ol-low (3,1). It becomes apparent soon that the proposed word () is the word (demonstration),13 and so the whose starting point is looked or is presumably a doctrineconcerned with the topic described by the proposed word, namely atheory o demonstration.14

By making the starting point o this theory an explanation o the wordby which its subject matter is called (i.e., the word ), the authorollows a method o inquiry that, as he tells us, should be applied to anyobject o investigation. It consists o three successive stages: frst, we mustdefne what people o the same nation and language agree to be themeaning o the name by which the problem is called; second, havingstarted rom this point, we must inquire whether the signifed thing, withwhich the defnition is concerned, exists or not. And fnally, i it isshown to exist, we must precisely investigate its nature, what it is like, andnever transgress the given order.15

Tese guidelines have a distinctly anti-sceptic overtone. Te possibility

o arriving at the starting point o the doctrine in question is deended onthe grounds that the word means something, and somethingexistent, to those who use it, be it philosophers, rhetors or judges (3,2).Tis observation meets with a more general demand or a possibility toconfrm () statements made about the subject matter oinquiry. As long as a statement is merely what seems to be the case (), an opponent can, with equal orce, show whatever he wants tothe contrary. It is thereore necessary to confrm the statement; but ithe judgment about it were based on something doubtul as well, and the

13) C. von Arnim, De octavo, 9.14) For c. Str. VIII 4,1. Te expression also appears in Str. VII95,3; c. Ernst, De Clementis, 14; Itter, Esoteric eaching, 96. However, in the latter passage, does not reer to the doctrine o demonstration, but to the teaching o Christwho is himsel believed to be the starting point o the doctrine; c. e.g. Str. VII 5,1; 57,3.15) 3,3-4 (GCS 17: 81,17-24): , . , , , , .


6/34


judgment about the latter again on something else that is also doubtul,[the account] would go on infnitely and would be unprovable. But i

conviction about that which is not agreed is derived rom that which isagreed by all, the latter should be taken as the starting point o the doc-trine (4,1).

Clement then presents a defnition o demonstration that supposedlymeets the above mentioned standard: All human beings would agree thatdemonstration is an account that provides convincingness () tothat which is doubtul on the basis o that which is agreed ( ).16 Tis defnition is a starting point o a subsequentinquiry o the essence o demonstration, in the course o which Clementurther clarifes the concept and distinguishes it rom other scientifcmethods, such as indication () or analysis.17

Te agreed defnition is broad enough to include two kinds o dem-onstration: (1) demonstration in the most proper sense ()and (2) the one that is merely based on opinion ().18 Accordingto Clement, a similar distinction holds or conviction (), andthough he does not say it explicitly, it must also be applied to that whichis agreed (), i.e., to the premises on which demonstra-

tion broadly speaking is based. Tat there are two kinds o agreed prem-ises is confrmed in 6,2 where Clement makes a distinction between aconclusion drawn rom agreed premises on the one hand, and a conclu-sion drawn rom true premises on the other. While the ormer procedureis called deduction, the latter is called demonstration. In this passage,deduction and demonstration are obviously not used as complemen-tary terms, nor are the agreed and the true premises complementary.Rather, demonstration (in the proper sense) is presented as a kind odeduction whose premises, apart rom being agreed, are also true.19

16) 5,1 (GCS 17: 82,12-14): .17) For the essence o demonstration c. Str. VIII 7,6 (). For (6,1) c. below, note 113; or analysis (8,1), c. below, note 98.18) 5,3.19) S.-P. Bergjan, commenting on Clements distinction between deduction and demon-stration, suggests that Clement attempts to separate the two as ar as possible. Accordingto Bergjan, Clements discussion is misleading at this point, since he ails to mentionthat this distinction is in Aristotelian terms a distinction between two subclasses o syllo-gisms. Bergjan urther contends that in response to the objection that both demonstra-tion and dialectical syllogism are syllogisms Clement turns to a Stoic line o thought


7/34


Tere seems to be an analogous relation between convincing and evi-dent principles in Clements account. I infnite regress is to be avoided

and the possibility o demonstration preserved, there must be, accordingto Clement, some principles o demonstration that are convincing bythemselves.20 Every demonstration is derived rom undemonstrated con-viction ( . . .). However, Clementsuggests that beyond conviction ( ) there areother principles o demonstration, namely that which appears clearly tosense perception and intellection ( ).21 An argument that starts rom these frst principles,rather than rom merely reputable premises ( ), anddraws an appropriate conclusion rom them, creates knowledge in thesouls o the auditors, rather than mere persuasion.22

In the preceding section the method o demonstration was appliedto the problem o demonstration itsel. In what ollows, Clement describesthe demonstrative method as applicable to any subject matter o inquiryand urther explains it by means o an example o a specifc researchquestion.

where he fnds the needed distinction between two valid conclusions, one a syllogism andthe other non-syllogistically concludent. (Logic and Teology, 406). I fnd Bergjanscomments unconvincing or two reasons: (1) Te act that demonstration is a kind odeduction is clearly shown in Str. VIII 6,4 where Clement says: By drawing an appropri-ate conclusion rom the premises we only make a deduction. But i each o our premises is

true, we make not onlya deduction, but also a demonstration (,). (2) I ail to see any indication in our text that Clement turnsto the Stoic line o thought mentioned by Bergjan. She reers to the distinction made inDiogenes Laertius, VitaeVII 78, between two kinds o valid arguments, those that arenon-syllogistically conclusive (also described as valid in the specifc sense) and thosethat are called syllogistic. But Bergjan does not explain how this passage is relevant toClements distinction between deduction and demonstration.20) 7,1.21) 7,2-3. Again, the distinction between convincing and evident principles does not indi-cate that the two terms are complementary. Rather, the concept o evident principles helpsClement to explain the dierence between two kinds o, introduced in Str. VIII 5,2and 7,7.22) 7,7-8.


8/34


c. Method o discovery (ch. III, 8,4V, 15,1)

Von Arnim correctly observes that the next thematic unit starts at 8,4,rather than at 9,1, as indicated by William Lowths chapter division.23 It isconcerned with a method by which the researcher, basing himsel on someprevious knowledge o the problem he investigates, proceeds to the dis-covery o something he has not known.

o start with, the researcher must get hold o dierent premises thatare appropriate () to the problem he investigates and he must alsoreormulate the problem itsel by means o an account that is agreed byall. In everything we investigate there is something we already know,

something convincing by itsel which we believe without proo, and thisshould become the starting point o research and the criterion o everysupposed discovery.24 Sometimes this previous knowledge, Clement goeson to speciy, may be the knowledge o the essence o the thing underscrutiny, accompanied by a complete ignorance concerning its activity,like, or example, in the case o stones, plants or animals whose activitiesor states or aculties, or generally speaking attributes, we ignore. Or wemay know some o these aculties or states or other attributes, but ignoreand investigate the essence, like in the case o the soul. Or we may have

knowledge o both the activities and the essences, but inquire to which othe essences the activities belong. Finally, o some things we know theiractivities and essences, but do not know their states.25

In a lengthy and eloquent section, Clement illustrates the method odiscovery26 by means o an elaboration o a specifc research question,namely whether the embryo is an animal. Again, the context o this dis-cussion is distinctly polemical. Clement introduces the question as anexample o a orm o expression which is potentially misleading, sinceits terms can be used in dierent ways. In order to answer it properly, wemust recognize the problems (), and this wecan do by a semantic analysis o its terms.27 It is done by the method o

23) von Arnim, De octavo, 10. Chapter division o the Stromata, proposed by WilliamLowth, was frst introduced in John Potters edition in 1715.24) 8,4-6. Although this is the only place where the word is mentioned in theeighth book, the concept already seems to be implied in Str. VIII 4,2, where Clement saysthat the agreed defnition o the proposed word shows the way to the discovery o what issought ().25) 9,1-5.26) C. 9,6: .27) C. 9,6-10,1.


9/34


questions and answers ( ): We ask ouropponent what he means by the words animal and embryo and

inquire whether the meaning he provides is doubtul or whether it isagreed by all.28 I he reuses to answer the question, he is shown to be aneristic person. In that case we should choose the method o exposition( ) instead and elaborate on the problem ourselves, whilegiving our opponent an opportunity to answer our exposition point bypoint when it is fnished. I he attempts to interrupt our investigation byhis questions, it becomes clear that he is not even willing to listen. 29

Now, to reach an agreement regarding the meaning o our terms, oneshould identiy what we have already known. We have already had a cer-tain notion o what the embryo is and what the animal is (i.e., we haveknown the essence) and seek to fnd out whether the activities and stateso the embryo are such as belong to the animal.30 For example, i wedefne the animal as that which is nurtured and grows, it remains to beshown that the embryo is nurtured and grows (which is obviously thecase).31 Or, i the animal is defned as that which perceives and sets itselto motion by an impulse, it is also clear what the matter o investigationshould be.32 But, as mentioned above, it is also necessary to determine

what we mean by the embryo.33

Again, in the ace o those who are shu-ing about names, the author makes it clear that we do not investigatethe word embryo nor its (incorporeal) meaning, but rather the nature othe thing under scrutiny.34 Although the semantic analysis o the terms oour question is a necessary prerequisite o research, we can only discoverthe answer by investigating the nature o the subject matter itsel.

Clement then outlines the course o such investigation by proposingthe defnition o animal as that which is capable () to per-ceive or be moved by an impulse. Tis defnition comprises both a pres-

ent condition (already is able) and a prospective condition (will ever be

28) In this context the views o Plato, Aristotle and the Stoics on the meaning o the wordanimal are compared (10,3-11,1).29) 11,3-12,1.30) C. 9,8-9.31) C. 11,2; Clement does not explain this implication in so many words, but this is obvi-ously what he means when he says: I our opponent says that the animal is that which isnurtured and grows, he has an answer [to his question].32) 11,2.33) 12,2. C. below, note 73.34) 12,7-13,2.


10/34


able). Tat the prospective condition holds or the embryo is evident andneed not be investigated. What remains, then, is to consider whether the

present condition holds o it as well, either potentially (as in the case oan animal that is at rest or sleeping), or actually.35 Unortunately, the textdoes not speciy how the researcher should proceed in order to answerthis last question, which is probably due to the act that the main goal othe whole discussion is merely to illustrate how to recognize the prob-lems.

Te relevance o this section to the problem o demonstration is explic-itly shown in its last paragraph which, as von Arnim observes, shouldinclude 15,1, a passage mistakenly placed at the beginning o the nextchapter by Lowth.36 Ater the outline o the method o discovery summa-rized above, demonstration is described as a common [method] appliedto everything claimed to be discovered. It is an argument by which weconfrm something on the basis o something else, and that on the basiso which we confrm it must be agreed and recognized by the student (onthe grounds o being evident to sense perception and intellection). In thisconnection Clement recalls the dierent types o research questions men-tioned above, e.g. when the essence is known but activities or states

unknown, or when we all know the activities and states, but do not knowthe essence, like in the case o a question in which part o the body is theruling part o the soul.37

3. Te source: Preliminary observations

Te theory o demonstration developed in Stromata VIII was hardlyinvented by the Christian writer himsel. It is true that in his extant writ-ings Clement is interested in demonstration; the little he tells us about it,however, is always frmly imbedded in the context o Christian thought.38As we have seen, Clements religious interests are still clearly present in

35) 13,3-5. Te argument is summarized in 13,6-8.36) C. von Arnim, De octavo, 10-11.37) 14,1-4.38) C. esp. Str. II 9,6; 25,3; 48-49; VII 95,3-96,1. C. Harry A. Wolson, Te Philosophyo the Church Fathers, Vol. I (Cambridge, Mass. 1956) 122.; Lilla, Clement, esp. 137-39;Elizabeth A. Clark, Clements Use o Aristotle (New York/oronto 1977) 16-26; GeorgiaApostolopoulou, Die Dialektik bei Klemens von Alexandria (Frankurt a.M. etc.: PeterLang, 1977) 87-98, esp. 92-98; Le Boulluec, La notion dhrsie, 398.; Ugo Schneider,Teologie als christliche Philosophie. Zur Bedeutung der biblischen Botschat im Denken des


11/34


the frst chapter oStromataVIII, but this line o thought is abandoned,and what is developed instead is a scholastic introduction to the doctrine

o demonstration that, in the way it is presented, has no obvious relevanceto the project outlined at the beginning o the book. Tis observationseems to suggest that, starting with chapter 2 and at least within theextent o the material discussed above (3,1-15,1), the text o StromataVIII draws rom a scholarly work concerned precisely with this topic, thedoctrine o demonstration.39

Is it possible to characterize Clements source more closely? Te historyo the theory o demonstration starts with Aristotles Posterior Analyticsand there are indeed many allusions to this work, as well as to other writ-ings o the Organon, in the sections discussed above. Te most obviousexamples are the distinction between deduction and demonstration in6,2-4 and the postulation, as a way to avoid the infnite regress, o thefrst premises o demonstration, characterized as convincing by them-selves and indemonstrable in 6,7-7,2.40 Te division o the three stageso inquiry in 3,3-4 (quoted above, note 15) could be derived rom Poste-rior AnalyticsII 10, 93b30-33.41 erminology o the section is largely Aris-totelian, as the author seems to recognize when he notes that what he calls

(conclusion)an Aristotelian term, others (namely theStoics) call (8,3). Te distinction made in 5,2-3 between the twokinds o demonstration, one that provides the scientifc conviction andone based on opinion, is not strictly Aristotelian, as Aristotle reservesthe term demonstration or the scientifc deduction only.42 But itmay well correspond to Aristotles distinction between demonstrationand the dialectical deduction based on reputable premises.43 Otherelements reect post-Aristotelian epistemology too, most notably the

39)

It may be regarded as a consensual view that in 3,1-15,1 (as well as in some other partso the book) Clement draws rom one and the same philosophical source; c. von Arnim,De octavo, 10-11, who reers to this source as liber dialecticusor introductio dialectica.40) Deduction and demonstration: c. An. post. I 2, 71b23-24; op. I 100a27-29; Ernst,De Clementis, 17; Bergjan, Logic and Teology, 405.; indemonstrable premises: c.An post. I 2, 71b27 and Jonathan Barnes,Aristotles Posterior Analytics(Oxord: ClarendonPress, 1975) 99, with other reerences; convincing by themselves: c. op. I 100b18-21.41) C. the discussion o the latter passage by David Charles, Aristotle on Meaning andEssence(Oxord: OUP, 2000) 23-56 et passim.42) C. above, note 40.43) op. I 100a27-30. Alternatively, in Str. VIII 5,3 could includeAristotles rhetorical deduction as well. C. Str. II 49,2 where both the dialectical and therhetorical deductions are classifed as . Te passage is quoted by


12/34


characterization o the frst principles o demonstration as that whichappears clearly to sense perception and intellection.44

Te Stoics were also preoccupied with demonstration, and at least thedefnition o demonstration in 5,1 (quoted above, note 16) bears a traceo the Stoic inuence.45 Stoic terminology is occasionally used.46 It is clear,however, that the author sympathizes more with the Aristotelian traditionand even distances himsel rom the Stoics (c. 4,3; 8,3).

Tese preliminary notes create a vague picture o a source grounded inthe Aristotelian tradition and elaborating the doctrine o demonstrationin a syncretistic manner. But as we start looking or terminological andmethodological parallels in post-Hellenistic philosophy, the picturebecomes much sharper. Not surprisingly perhaps, the most signifcantparallels are ound in the writings o another syncretistic author o Clem-ents time who was occupied with the doctrine o demonstration, namelyGalen. Exploring them will help us determine the doctrinal backgroundo Clements source with more precision.

4. Clement and Galen

Galens interest in logic in general and demonstrative method in particu-lar is well known. His early treatise On Demonstration is now lost, but hisviews about this matter can partly be reconstructed on the basis o hisextant writings.47 In 1910 Wilhelm Ernst already noted some terminolog-

44) C. Sextus, M. VII 217, reerring to Teophrastus; c. Pamela Huby, Teophrastus oEresus: Sources or his Lie, Writings, Tought and Inuence, Commentary, IV: Psychology(Leiden etc.: Brill, 1999) 93-95. For other Aristotelian elements c. below, notes 64 and65; Solmsen, Early Christian Interest, 283.45)

C. Cicero,Ac. II 26; Diogenes Laertius, VitaeVII 45; Sextus,M. VIII 314; 422. (par-allels noted by Ernst, De Clementis, 15.).46) C. esp. Clements distinction between the (corporeal) voice, the (incorporeal) mean-ing (also called the thing) and the nature o the investigated matter in Str. VIII 12,7-13,2,which employs the Stoic concept o the incorporeal , also described as the signifedthing itsel ( . . .) by the Stoics; c. Sextus,M. VIII 12, notedby Sthlin in his GCS edition, adp. 87,23-25. For as meaning c. Pierre Hadot,Sur divers sens du mot pragmadans la tradition philosophique grecque, in P. Aubenque(ed.), Concepts et catgories dans la pense antique(Paris: Vrin, 1980) 309-319.47) C. Iwan von Mller, ber Galens Werk vom wissenschatlichen Beweis, inAbhand-lungen der kniglich-bayerischen Akademie der Wissenschaten, phil.-philol. Kl. 20 (1897)405-478; Barnes, Galen on Logic and Terapy, in F. KudlienR.J. Durling (edd),Galens Method o Healing (Leiden etc.: Brill, 1991) 50-102; Riccardo Chiaradonna, Le


13/34


ical correspondences between Clement and Galens works Introduction toLogic, On the Terapeutic Methodand On the Doctrines o Hippocrates and

Plato.48 In an important article published in 1973 Friedrich Solmsenremarks that i we examine the ragments o Galens treatise On Demon-stration with AristotlesAnalyticsas well as StromataVIII in mind, we arestruck by the number o tenets they have in common.49 And in a studypublished in 1996 the list o parallels between StromataVIII and the writ-ings o Galen is urther extended by eun ieleman, who also considerstheir number and coherence striking.50 And striking indeed it is. Inwhat ollows I list the most important correspondences, including manythat, to my knowledge, have not been previously observed.

a. On the Terapeutic Method

In the frst book o On the Terapeutic Method Galen introduces themethod by which he will proceed in the treatise. In this connection herecalls the methodological discussion in his work On Demonstration inwhich it was shown that the frst principles o every demonstration arethings clearly apparent to sense perception and intellection and withevery subject matter o inquiry we must replace the name with an

trait de Galien Sur la dmonstration et sa postrit tardo-antique, in R. Chiaradonna,F. rabattoni (eds.), Physics and Philosophy o Nature in Greek Neoplatonism (Leiden/Bos-ton: Brill, 2009) 43-77.48) Ernst, De Clementis, 11-24, indicates the ollowing parallels: (1) Str. VIII 6,1 andGalen, Inst. log. 11,1: the defnition o; (2) Str. VIII 6,2 and Galen, Inst. log. 11,2:the description o demonstration as a conclusion drawn rom true premises; (3) Str. VIII7,2 and Galen, Inst. log. 1,5; 8,3; 16,6-7; 17,7: the description o the frst premises as that

which is convincing by itsel; (4) Str. VIII 7,3; 14,3 and Galen, MM(Khn X 39,7-9);PHPIX 7,4 (CMG V 4,1,2: 586,19.): the description o the frst premises as clear tosense preception and intellection; (5) Str. VIII 8,2 and Galen, PHP II 3,12 (CMG V4,1,2: 112,4.): indierence concerning names by which premises are called; (6) Str. VIII8,4 and Galen, Inst. log. 1,2: on the necessity o having premises appropriate to the dem-onstrated matter. Ernst quotes these passages along with parallels rom Aristotle, Alexan-der o Aphrodisias and other sources in order to show that the background o Clementsdiscussion on demonstration is chiey Peripatetic.49) Solmsen, Early Christian Interest, 285. Apart rom the correspondences detected byErnst, Solmsen reers to verbatim agreement between MM(Khn X 39,9.) and Clem.Str. VIII 4,2.50) C. eun ieleman, Galen and Chrysipus on the Soul(Leiden etc.: Brill, 1996) 127. Teparallels are mentioned on pp. 20, 24., 30, 104.


14/34


account.51 Galen then goes on to search or the starting point o a truedoctrine ( ) concerning the therapeutic

method. As his project in this book is to discover therapies or all diseases,it is necessary or him to distinguish dierent kinds o disease. But onecannot fnd either dierentiae or species without a secure knowledge othat what is divided, i.e. the genus in question. Tereore, concludesGalen, in our case, too, we must explain by an account what disease is( . . .), so that we may perormits division correctly.52 But how shall we do that? How else, says Galen,than in the manner described in On Demonstration? First, we must agreeon the notion o the thing proposed without which it is impossible to dis-cover its essence. And, as we said [in On Demonstration], this notion mustbe taken as something agreed by all, or otherwise it would not be properto call it a starting-point. What, then, is the notion o being ill that isagreed by all human beings? And to what underlying thing do they mostoten reer with this word to be ill?53

Galens procedure described above is very similar to the method Clem-ent employs in Str. VIII 3,1-4,2 in order to fnd the starting point o thedoctrine o demonstration. According to Clement, there is no better o

clearer method to arrive at the starting point o such doctrine ( . . .) than to explain the proposed word by anaccount () so clear that all who speak thesame language will ollow. (3,1) Tis conorms with a general rule weshould apply to every subject matter o inquiry, namely that our argu-ment should not be based on any agreed starting point other than theollowing: what people o the same nation and language agree to be themeaning o the name by which the problem is called. (3,3).

As we know, according to Clement, arriving at the agreed defnition o

the proposed word is the frst stage o inquiry which should be ollowedby a question concerning the existence o the thing signifed and fnallyby research concerning its essence (3,4). In the passage discussed above,

51) Galen,MM(Khn X 39,7-10).52) X 40,2-11.53) X 40,11-41,1: ; , , . ;


15/34


Galen distinguishes between the frst and the third stage o this schemewhen he asks, frst, about the notion o being ill that is agreed by all

human beings and, second, to what underlying thing the word usuallyreers. According to Galen, it is necessary to agree on the notion() o the proposed thing in order to discover its essence ().54

Te similarity o methods is joined with the similarity o terms. EvenGalens phrase we must replace the name with an account () has a close parallel in Str. VIII 4,2 ().55 Te main di-erence between the two descriptions is that Galen is concerned with thedoctrine o therapeutic method and the proposed word is to be ill,whereas Clement is concerned with the doctrine o demonstration andthe proposed word is demonstration.

Clements question about the existence o the subject matter o inquiry(3,4) is not addressed by Galen in the passage quoted above. But later inthe treatise, in the context o a polemic against the Empiricist view thatgeneral terms always reer to particulars, Galen asks a rhetorical questionabout the meaning o the word disease and the existence o the thingsignifed that, again, reminds us o Clements (presumably anti-sceptic

and equally rhetorical) question concerning the meaning o the worddemonstration and the existence o the thing it signifes. Says Clement:Is the name demonstration o such kind as blituri, a mere sound thatmeans nothing? (. . .) At any rate, philosophers provide demonstration assomething existent, each party in a dierent way.56 Compare Galen: Dothe words animal and disease seem to you to signiy nothing, but have asense similar to that o blituri and scindapsus? Or is it the case that theysigniy, but yet there is no object underlying the words, as is the case withScylla and centaur?57

Other conceptual and terminological correspondences between the twotexts may be observed. Ater discussing the way people normally use thewords health and illness, Galen suggests that throughout the present

54)MM(Khn X 40,12-41,1).55) Galen,MM(Khn X 39,9.). C. Solmsen, Early Christian Interest, 285; ieleman,Galen and Chrysippus, 24. and note 69.56) Str. VIII 3,1-2.57)MMX 144,9-14, translated by Robert J. Hankinson, Galen On the Terapeutic Method(Oxord: Clarendon Press, 1991) 72. In the last quoted sentence Galen, like Clement,makes a distinction between the meaning o the word and the existence o the underlyingobject.


16/34


treatise we shall derive our interpretation o words rom ordinary Greekusage (), as we have said in the treatise On

Demonstration. However, continues Galen, discoveries, investigationsand demonstrations concerning the actual essence o the thing will not bederived rom the opinions o the multitude ( ) but rom the scientifc premises ( ) the manner o whose discovery was elaborated in that work.58Here Galen again makes a distinction between what appears as the frstand the third stage o Clements scheme and indicates that the premisesemployed in the frst phasethe commonly accepted meanings owordsmay not yet be the scientifc ones. A similar line o thought canbe discerned in Clement. Te premises established in the frst stage aremeanings agreed by people o the same nation and language (3,3).However, as we have seen above, when discussing the nature o demon-stration Clement distinguishes between two kinds o agreed premises,those that correspond to an opinion and those that are true and evident(5,1-2; 6,2-4; 7,3). Tis distinction, o course, does not mean that thepremises established in the frst stage, apart rom being agreed by all, maynot also be true. Indeed, according to Clement, when replacing the pro-

posed name with an account, the researchers ought to make sure that theaccount is not merely what appears to be the case (), or anopponent could, with equal orce, show whatever he wants to the con-trary. Instead, they should fnd an account that is agreed and clear tothem all () andthat will show the way to the discovery o what is sought ().59 But it seems that (or Clement as orGalen) the truth o the matter can only be established when the researcherinvestigates its essence and discovers an account that is frmly based on

that what appears clearly to sense perception and intellection.

b. Te embryo question

In StromataV 5,3 Clement mentions the question whether the embryois or is not an animal () in connectionwith the so-called convertible statements (), that is tosay, statements that can equally be proposed by those who argue or the

58)MMX 41,17-42,9.59) Str. VIII 4,1-2.


17/34


opposite thesis ( ).60 Tis reects a view that the argumentspro and contrathe state-

ment the embryo is an animal are equally convincing.However, in StromataVIII Clement holds a dierent view.61 He intro-

duces the question as an example o a orm oexpression () that conuses and disturbs our mind,so that it is not easy to discover the dierences it involves.62 It is ques-tions like these that show why the researcher must frst o all recognizethe problems. (9,6-7). Te dierences that the question involves pre-sumably correspond to the various ways in which the words animal andembryo are used. First Clement discusses dierent meanings o animaland then o embryo (10,1-12,2), which enables him to ormulate theresearch question in the clearest possible manner and open the way to itssolution. No doubt that the aim o the whole discussion is to show thatwith the correct method o discovery even the paradigmatically conus-ing question about the embryo can be resolved in a scientifc manner.

Te section concerned with the method o discovery (Str. VIII 8,4-15,1) again contains many elements that recall the writings o Galen.According to Clement, with every subject matter o inquiry we must

establish dierent premises corresponding to each problem that are appro-priate to what is proposed ( . . . ).63 Tis remindsus o a passage in the second book o Galens treatise On the Doctrines oHippocrates and Plato where the author censures the Stoics or using prem-ises that are not appropriate to the question proposed ( ).64 What he means by that is explained at the

60) For convertible statement c. Aulus Gellius, Noct. Attic. V 10.; IX 16,7; Le Boulluec,SC 279, 34-36.61)

C. Manseld, Doxography, 3186, note 623, who points out that in book eight theexample derives rom another tradition than in book fve. C. also Marie-Hlne Con-gourdeau, Lembryon et son me dans les sources grecques (VIe sicle av. J.-C.-Ve sicle apr.J.-C.) (Paris 2007) 180. Te incompatibility o the two passages is indicated by the actthat in the fth book (5,3) Clement includes the embryo question among things unft orinquiry ( []).62) For c. Aristotle, Soph. el. 165b27 et passim; ieleman, Galen andChrysippus, 17-18, 20.63) Str. VIII 8,4: , .64) PHPII 2,2 (CMG V 4,1,2: 102,20-22). Galen adds that he dedicated a lengthy sec-tion to these premises in his treatise On Demonstration (II 2,3 [CMG V 4,1,2: 102,25-27]).Te concept o the appropriate premises is probably based on Aristotle, An. post. I 2,


18/34


beginning o book three where Galen presents a our-old classifcation opremises, describing the frst two classes as ollows: (1) premises taken

rom the attributes . . . according to the essence o the problem () and (2) premises taken rom the attributes, butnot according to the proposed matter o inquiry ( ).65 In order to establish appropriatepremises, it is thereore necessary to have some knowledge o the essence.Tis explains why, in Clements account, the researcher not only ought tohave dierent premises in respect to each problem, but he must alsochange the problem itsel into an account, no doubt basing himsel onwhat is previously known about its essence.66 Tis previous knowledgedoes not necessarily extend to the essence o the problem as a whole. Buti premises are taken that are not appropriate to the problem, saysClement, it is not well possible [or the researcher] to discover anything,since the nature o the problem, also called the question, remains unknownas a whole.67

In the next paragraph Clement explains that in some cases we actuallymay have previous knowledge o the essence o the problem as a whole( . . . ), while being completely ignorant

71b23; c. also 72a6 and other passages quoted by ieleman, Galen and Chrysippus, 13,note 23.65) PHPIII 1,4 (CMG V 4,1,2: 168,15-17). C. Paul Moraux, Der Aristotelismus bei denGriechen, Bd. II (Berlin/New York: De Gruyter, 1984) 719. and note 151, who reers toAristotle,An. post. I 4, 73a34-73b24.66) Str. VIII 8,4-6.67) Str. VIII 8,5 (GCS 17: 84,27-30): ** , , . Following Eduard Schwartz, Sthlin

indicates a lacuna between and , but the phrase is probably correct. Compare the ollowing expressions in Galen: (Khn IX 729,2); (Khn III 308,10; c. 700,9-10); c. also Khn X 172,14-15: . Te dative could be linked to and reer to the researcher, last mentioned in 8,3 (GCS 17: 84,21: );hence my translation above. Te diculty o this solution might lie in the act that thenearest possible reerent o is placed rather ar in the text. Tis problem could beavoided by reading [scil. ] instead o (it is not well possibleto discover anything about it [i.e., the problem]). (My thanks are due to Proessor AlainLe Boulluec who, in a personal communication [August 2010], agrees that the Galenicparallels listed above allow us to avoid the supposition o a lacuna and suggests to retainthe dative .)


19/34


o its attributes, or example activities, aections or aculties. In othercases, we have knowledge o some o these attributes (or example desires

and aections o the soul), but do not know and investigate the essence(9,1-2). He returns to this division in 14,4, where he adduces a questionthat corresponds to the case when we all know the activities and aec-tions () but do not know the essence, namely in whichpart o the body is the ruling part o the soul. As noted by ieleman, thisquestion corresponds to the subject matter o the second book o GalensOn the Doctrines o Hippocrates and Plato.68 In the sixth book o the sametreatise Galen dedicates a lengthy discussion to the distinction betweenactivities and aections o the soul (VI 1,5-27), a distinction made, ina similar context, in Str. VIII 14,4. Moreover, the view that the essence othe soul is unknown is characteristically Galenic.69

For the sake o completeness, we may add that the doxographic sec-tion where Clement compares the opinions o Plato, Aristotle and theStoics concerning the dierence between animals and plants (10,3-11,1)contains several correspondences with Galen, PHPVI 3,7.70 In Str. VIII13,3 Clement introduces a defnition o the animal as that which is capa-ble o perceiving and moving by impulse (

).

71

Tis recalls a ormulation in Galens treatise OnMy Own Opinionswhere the author answers the question why animalsare superior to plants by pointing out two things, perception and move-ment by impulse ( . . . ).72

68) C. ieleman, Galen and Chrysippus, 30. However, as Manseld, Doxography, 3092-3108 et passim, shows, the question is a commonplace philosophical theme associated withthe doxographic tradition.69)

C. Prop. Plac. 3,1 (CMGV 3,2: 60,3.) and the reerence collected by Vivian Nuttonad loc. in Galen, On My Own Opinions(Berlin: Akademie Verlag, 1999), 60. C. also Pier-luigi Donini, in Hankinson (ed.), Te Cambridge Companion to Galen (Cambridge: CUP,2008) 185., 203, note 9.70) C. Strom. VIII 10,3-4 and PHPVI 3,7 (CMG V 4,1,2: 374,14-19). C. ieleman,Galen and Chrysippus, 24, note 67. Parallels in doxographic literature are discussed byManseld, Doxography, 3187-3190.71) Te defnition is already implied in 9,8.72) Prop. Plac. 13,7 (CMG V 3,2: 108,12-14). C. Nat. Fac. I 1 where Galen says that thespecifc eature o the animals is to perceive and move by choice ( ). Te latter passage is noted by ieleman, Galen andChrysippus, 30, note 93a, as closely similar to Str. VIII 9,8. C. also Ps.-Galen, Defnitio-nes medicae(Khn XIX 452); Congourdeau, Lembryon et son me, 291.


20/34


However, the most striking (and, to my knowledge, hitherto unno-ticed) parallel is the one related to Clements didactic example itsel,

namely to the embryo question. In Str. VIII 12,2, having discussed thevarious meanings o the word animal, Clement proceeds to ask his vir-tual opponent what he means by the word embryo, literally the thingconceived or the thing in the womb ():Does he take the expression the thing conceived or the thing in thewomb to signiy already that which is not ormed, and even the seeddeposited in the womb, or just that which is dierentiated and ormedalready? (), that is, theso-called ?73 In this passage Clement indicates that the word specifcally designates the developed oetus.74 wenty lines laterhe says: As regards the nature o the thing we are investigating, namely, we have clearly shown what it is like.75 Tis sentence apparentlyreers to the description o in 12,2, or perhaps to some moredetailed discussion not preserved in our excerpts. In any case it shows thatin order to solve the embryo question, Clement narrows the matter oinvestigation to the developed oetus.

Tis observation brings us back to Galen who in his writing On the

Utility o the Partsgives us the ollowing testimony: Tat the thing in thewomb is already an animal, at least when all o its members are ormed(), we said in the treatisesOn Demonstration and On the Doctrines o Hippocrates and Plato.76 Tispassage is interesting not only as evidence that Galen based his solutiono the problem on the same distinction between the levels o the develop-

73) Str. VIII 12,2 (GCS 17: 87,4-7): ,

, .74) Tis corresponds to a medical usage also attested by Galen; c. Vronique Boudon-Millot, La naissance de la vie dans la thorie mdicale et philosophique de Galien, inL. Brisson, M.-H. Congourdeau, J-L. Solre (edd.), Lembryon: ormation et animation(Paris: Vrin, 2008) 84-87. I ollow the authors suggestion to translate asembryo and as oetus (87.).75) Str. VIII 13,2 (GCS 17: 87,26.): , .76) Galen, UPXV 5 (Helmreich II 357,24-28 = Khn IV 238,19-239,4): , , .


21/34


ment o the embryo that we encounter in Clement.77 It is interestingespecially because it shows that Galen was occupied with the embryo

question in his treatise On Demonstration. Te act that this unexpecteditem is on the list o the contents o the treatise has, to my knowledge,never been suciently explained.78 However, the role that this questionplays in Clements discussion o the demonstrative method might give usa clue.

c. Four more parallels

More parallels may be added that indicate the proximity between the sec-

tions oStromataVIII discussed above and the writings o Galen. I willcomplete the list with the ollowing our.In Str. VIII 5,2, having made a distinction between two kinds o dem-

onstration and conviction, Clement also distinguishes two kinds oknowledge, but also oreknowledge, one scientifc and sure ( ), the other merely based on expectation ().79Now it is not very clear what Clement means by oreknowledge in thispassage. One immediately thinks o previous knowledge in Aristotlestheory o demonstration,80 but the specifcation o the defcient kind o

as suggests that Clement uses the word oreknowl-edge in the sense o the knowledge o uture events. Von Arnim thinks

77) C. Congourdeau, Lembryon et son me, 309. In Str. VIII 12,2 Clement describes theless developed phase o the thing in the womb by two expressions: a) , b) . Tese expressions seem to cor-respond, in a chronologically reverse order, to the frst two stages o development o thething conceived, as outlined in Galens writings On Semen and On the Formation o the Phoetus. C. esp. Sem. (Khn IV 542) where the initial stage called , in which the

orm o the seed is dominant (), is succeed by the stage o, in which the main organs (heart, brain and liver) are not dierentiated andormed yet (). Next is the stage o, which ur-ther develops into (543). C. Boudon-Millot, La naissance de la vie, 85-87.78) Von Mller, ber Galens Werk, 465, suggests that Galen elaborated the embryoquestion in connection with an inquiry concerning the essence o the soul, which in turnwas intended to show the limits o the demonstrative method (c. 460-64). However, it isunclear how the solution o the embryo question Galen proposed in De demonstrationewould have helped him convey his sceptical message concerning the souls essence.79) Str. VIII 5,2 (GCS 17: 82,14-16): , , , .80) C. Aristotle,An. post. I 1, 71a1-2; I,2, 72a34-36.


22/34


that Clement introduces a Christian element here,81 but it is not necessar-ily so. Te ancients already knew the dierence between layman and

expert opinion about uture events82 and Aristotle even (perhaps hal-mockingly) describes divination as .83 Te distinc-tion gained in importance or doctors who tried to base medicine (owhich prognosis is an important part) on frm scientifc oundations. Inhis Commentary on Hippocrates Prognostic, Galen distinguishes two mean-ings o the word , comparing them with two ways in which theword is normally used, one being sure (), as when we saythat ater winter there will be spring, then summer and then autumn,and the other not sure (), as when Aratus says that i the upperhorn o the moon leans orward, we should expect a storm rom thenorth, and i it inclines backwards, a storm rom the south (Phaen. I794.). Tough it does usually () turn out to be so, sometimesit does not.84 A ew lines later Galen says that people normally use theword oreknowledge to designate both the expectation that uture eventswill happen as usually, and the one that is certain.85 He adds that a gooddoctor should not care about such distinctions, but rather try to makepredictions that turn out to be correct as oten as possible.86 Nevertheless

on another occasion he does indicate that some medical predictions maybe regarded as sure (namely when they are concerned with necessaryconsequences)87 and he even uses the term as adescription o a sure prediction.88

Clements rare collocations (5,2) and (8,2) can also be ound in Galen.89 More importantly, in the latterpassage Clement recommends to the demonstrative man that he shouldnot worry about names by which premises are called whether peoplewish to call them axioms or propositions or assumptions (

81) Von Arnim, De octavo, 10.82) C. Plato, Tt. 178c-179a; Aristotle,Met. IV 1010b11-14.83) De mem. 449b12. I owe this reerence to Miroslav edina.84) Galen, Hipp. Prog. (Khn XVIIIb 12,14-13,13).85) Galen, Hipp. Prog. (Khn XVIIIb 14,10.): .86) Khn XVIIIb 14,12-15,5.87) C. CAM(Khn I 289,15-293,17).88) Hipp. O. Med. (Khn XVIIIb 636,14.).89) : Galen, Ut. Resp. (Khn IV 492,11); c. SM (Khn XI636,12-16); : Galen, Sem. (Khn IV 649,7); c. PHPII 3,17 (CMG V4,1,2: 112,30).


23/34


); this recommendation notonly corresponds to Galens sel-proessed tendency o using philosophical

terms he regards as synonyms interchangeably,90 but specifcally recalls theollowing passage in PHP II 3,12: Te ancients wrote about all thesekinds [o premises], whether you want to call them propositions or axi-oms or statements (), it makes no dierence or our present purpose . . .91

d. Dierences?

Can we detect any specifc dierences between the concept o demonstra-

tion recorded by Clement and Galens views? ieleman proposes two suchpoints o dispute, but on closer examination one proves disappointingand the other inconclusive.

When discussing Clements distinction, in Str. VIII 7,7-8, between twokinds o demonstration (the one based on evident premises and the otheron reputable ones), ieleman notes that in this context Clement alsoreers to the requirement that premises should be appropriate (),but quite unlike Galen presents this as a prerequisite or syllogistic(including endoxic) reasoningtout court.92 However, this supposed dier-

ence is based on mere oversight. It is true that in the passage concernedClement uses the word in relation to both kinds o deduction.But unlike in 8,4, where the word reers to the appropriate premises(precisely in the Galenic sense), in the passage discussed by ieleman theword reers to an appropriately drawn conclusion ( ).93 Clements point is that both kinds o deduction are stilldeductions (valid arguments), provided that their conclusions are appro-priate to their premises.94 In this sense the word is also used byAlexander o Aphrodisias and, indeed, by Galen.95

90) C. Ben Morison, in Cambridge Companion to Galen, 148.91) C. Ernst, 23, note 1; ieleman, Galen and Chrysippus, 24-25 and note 69.92) ieleman, Galen and Chrysippus, 20.93) Str. VIII 7,8 (GCS 17: 84,2-8): , , , , , , , ,

,

,

.94) Te point is already made in 6,4. C. Bergjan, Logic and Teology, 405.

95) C. Alexander, In op. (CAG II.2: 14,2); Galen, Hipp. Elem. (Khn I 445,14.); Ven.


24/34


ielemans second observation is more interesting. In a chapter dedi-cated to Galens scientifc method ieleman recalls a passage in Galens

treatise On the Diagnosis and Cure o the Errors o the Soul where theauthor argues or the utility o mathematical methods (particularly analy-sis) or healthy lie. In this context Galen elaborates his point by means oan example o the way these methods are used in architecture, namely inthe construction o a sundial.96 Having indicated how the logical methodso analysis and synthesis are employed in this case, Galen proceeds to dis-cuss ways by which the accuracy o the sundial can be tested empirically.97According to ieleman, this example illustrates a more general pointabout Galens demonstrative method, namely that the logical method oinquiry should be tested by experience: Te addition o empirical testingto the logical method (i.e. analysis/synthesis) reects a amiliar Galenicscheme, viz. the duo reason/experience. In this respect Galen diers romthe Platonist account o philosophical method oered in Clement, Str.VIII, which is closely similar to this and other Galenic passages in allother respects. ieleman notes that Clement agrees with Galen insoar asanalysis in concerned.98 Unlike Galen, however, Clement identifes dem-onstration with the downward route rom the axioms, which coincides

with what Galen calls synthesis.99

In Galens view as presented by iele-man, analysis and synthesis lead to some amount o clarifcation o thesubject matter o inquiry. Nevertheless, defnitive and clear confrmationthat what is looked or has indeed been ound is only provided by thesubsequent empirical test. Only when the later stage has been added dowe have proo or demonstration in its proper sense.100

But is Clements description o proo (when we reach what we arelooking or rom the frst premises through all the middle terms) really a

96) Galen, Pecc. Dig. 4-5,16 (CMGV 4,1,1: 53,9-59,8).97) C. ieleman, Galen and Chrysippus, 34.98) C. Str. VIII 8,1 where Clement describes analysis as a procedure by which we ascend

rom demonstrable premises to that which is evident to sense perception and intellection.C. Galen,MM(Khn X 33,14-18), where the author provides a description o the proce-dure which is closely similar to Str. VIII 8,1. On Galens concept o analysis c. Barnes,Galen on Logic and Terapy, 67; ieleman, Galen and Chrysippus, 33.99) ieleman reers to Str. VIII 8,1: Demonstration takes place when we reach what we

are looking or rom the frst premises through all the middle terms (, ).

100) ieleman, Galen and Chrysippus, 35. On the importance o empirical verifcation orGalen c. Hankinson, in Cambridge Companion to Galen, 169-178.


25/34


point o a specifc dierence rom Galens theory? In his opusculumOn the Best Method o eachingGalen reers us to his treatise On Demon-

stration where it is written how, by starting rom the elements and prin-ciples in each case, a man may best demonstrate whatever can bedemonstrated.101 Tus the demonstrative method is described as the axi-omatic method,102 and the need or empirical testing is not a part o itsdefnition. It is true that in the writing discussed by ieleman Galen alsosays: When we fnd a demonstrative method that leads us to what we arelooking or and is clearly confrmed by the thing itsel, we have an excel-lent test o its truth.103 Here, as in Clement, the demonstrative methodleads us to what we are looking or. Te confrmation o the methodcomes rom the thing itsel, no doubt provided that the premises o theinquiry are appropriate to the things essence. But i our method is to beconfrmed by the thing itsel, such confrmation must necessarily bebased on something more evident than the premises on which we havebased our investigation so ar. What ieleman calls the addition oempirical testing seems to be an example o such confrmation. It pro-vides the researcher with new, more evident premises or his argumentabout the subject matter.104

Now as we know, Clement derives the frst premises rom that whichappears clearly to sense perception and intellection. Tis ormulationallows or the possibility that when it comes to matters that can only bedemonstrated on the basis o sensory evidence, Clements demonstrativeman will seek the same kind o empirical data as Galens. Tere is nodirect indication o that in StromataVIII. Nevertheless the whole discus-sion about the embryo culminates in the insight that in order to fnd outwhether the embryo is an animal, we must inquire whether it is actuallyable to move and perceive. But this question can hardly be dealt with by a

method other than empirical research.

101) Opt. Doct. 5,3 (CMG V 1,1: 106,19-108,1 = Khn I 52,12-14); I ollow the transla-tion o Barnes, Galen on Logic and Terapy, 66.102) C. Barnes, Galen on Logic and Terapy, 66.103) Galen, Pecc. Dig. 3,5 (CMGV 4,1,1: 47,4-7 = Khn V 68,4-7). Te passage is quotedby ieleman, 35, note 116. C. also Hankinson, in Cambridge Companion to Galen, 169.104)

A good example is the amous argument rom dissection described in Galen, PHPII3,4-8 (CMG V 4,1,2: 110,1-14). C. Donini, in Cambridge Companion to Galen, 190.


26/34


5. Te question of source

Te number and coherence o similarities with extant writings o Galenand especially the close correspondences with passages that reer to hislost treatise On Demonstration allow us, I believe, to identiy the source oStromataVIII, at least as ar as the passages discussed above are concerned,as a writing specifcally dedicated to the doctrine o demonstration whosephilosophical background, method and vocabulary is extremely close tothose o Galen. But then obviously the question arises whether theremight be any genetic relation between Clements source and the writingso Galen, and i so, o what kind. Here, equipped with the evidence rom

the previous section, we are entering the realm o the plausible. Withinthis realm, we may distinguish three lines o interpretation.

(1) Tere is no direct relation between Clements source and the writ-ings o Galen, but both draw rom the same scholastic traditionwhere the parallels noted above are common.

(2) Clements source is a source o Galen.(3) Clements source is Galen.

Te frst line o interpretation is chosen by Solmsen who thinks that thedoctrines common to Clement and Galen reect the philosophical syn-cretism (alias eclecticism) which prevailed in logic as much as elsewhereand describes Galen as representative o a trend.105 In a similar vein,ieleman regards StromataVIII as a document o the Platonist scholastictradition and as an indicator o traditional elements in Galens concepto demonstration.106 ieleman agrees with Jaap Manseld, according towhom the source or sources o Stromata VIII belong with the tradi-tions o the Middle Platonist scholastic literature.107 Manseld bases thisassessment mainly on his elaboration oStr. VIII 17-21, a chapter dedi-cated to the problem o division and defnition that contains parallelswith the Middle Platonist sources, especially the fth chapter o AlcinousDidascalicus.108 Some o these parallels were already noted by R.E. Witt

105) Solmsen, Early Christian Interest, 286. As a possible source o the trend representedby both Galen and StromataVIII Solmsen suggests Gaius (290, note 36).106) C. ieleman, Galen and Chrysippus, 20, note 47; 24; 104.107) Manseld, Doxography, 3184. C. also Manseld, Heresiography in Context(Leidenetc.: Brill, 1992) 62.108) C. Manseld, Heresiography, 78-109, esp. 80-84.


27/34


who traces them, as all the philosophical material in Stromata VIII, toAntiochus o Ascalon, or rather to some Peripatetic authority inuenced

by Antiochus, or example Aristocles o Messene.109However, these explanations underestimate the act that the source (or

one o the sources) o Stromata VIII is specifcally concerned with theproblem o demonstration. Moreover, neither Witt nor Manseld takeinto consideration the Galenic parallels noted above. According to Mans-eld, Clements possible sources are a manual or more than one and pos-sibly (. . .) include the logico-epistemological section o a substantial workdealing with systematic philosophy in the manner o Alcinous.110 But wehave seen that the logico-epistemological section discussed above can becharacterized more precisely as an introduction to the doctrine o demon-stration. We may add that the chapter on division and defnition (Str.VIII 17-21) neatly coheres with this theme. It is impossible to analyze thischapter in detail within the ramework o the present study. Suce to saythat the text abounds with allusions to the Organon,especially the Poste-rior Analytics, which apart rom its critique o division also includes chap-ters where division is presented as a useul scientifc method, notably as atool or fnding defnitions.111 Another point o convergence with the two

earlier sections oStromataVIII is the act that the chapter on divisionand defnition covers topics that are oten discussed by Galen and werecertainly elaborated in his treatise On Demonstration.112 Tus the sourceo the logico-epistemological material in StromataVIII may be more ade-quately characterized (to paraphrase Manseld) as a substantial workdealing with the demonstrative method in the manner o Galen.

Having described the topic and style o our source more narrowly wecan ocus our inquiry on the ollowing question. Is it possible that the

109)

Witt,Albinus, 31-41.110) Doxography, 3184.111)An. post. II 13-14. By emphasizing the Aristotelian content I do not mean to denythat the ramework o the discussion is probably Platonist; c. Witt,Albinus36-38; Mans-eld, Heresiography, 80-84. C. also J. ManseldD.. Runia, Atiana. Te Method andIntellectual Context o a Doxographer, Vol. III (Leiden/Boston: Brill, 2010) 63.112) C. Barnes, Galen on Logic and Terapy, 68; 72-76; Chiaradonna, Le trait deGalien, 45. For Galens interest in division and defnition c. also ieleman, in CambridgeCompanion to Galen, 59.; Hankinson, ibid., 167. C. also Manseld, Heresiography, 330,who quotes Galen, PHPIX 9,43-46 (CMG V 4,1,2: 608), as another parallel to Alcinous,Did. 5. Te passage recalls Clements distinction o three kinds o division in Str. VIII19,3-8. However, in this instance Galen is closer to Alcinous than to Clement, as he intro-duces not three, but fve kinds o division. C. Manseld, Heresiography, 81.


28/34


parallels we have detected in the method and vocabulary oStromataVIIIon the one hand and Galens treatise On the Terapeutic Methodon the

other reect a scholastic tradition where this particular method andvocabulary was common? Possible it is. Ater all, the topic and much othe vocabulary is Aristotelian, though elaborated in a distinctly eclecticmanner, and many isolated motis can be ound elsewhere. But as ar as Iam aware, we have no indication o a similar treatment o the demonstra-tive method, any testimony or ragment to support this explanation. Tiscould still be ascribed to an optical illusion created by the exceptionalnumber o Galens writings that have been preserved in contrast to thenumber o philosophical documents o the same period that have beenlost. But the curious detail that both Clement and Galen employ thesame research question, whether the embryo is an animal, in the contexto an exposition o the demonstrative method, and solve it in a similarashion, using similar words, supports a dierent view, namely that therelation between the two texts is somehow more exclusive.

Tis view is urther supported by the ollowing consideration. I ourinterpretation o the word prognosis(5,2) is correct, we have a reason tobelieve that the context o Clements discussion o the theory o demon-

stration is specifcally medical. Tere is another medical term used in thesame section, namely (6,1).113 It is worth mentioning in this con-nection that the source rom which Clement draws in the last chapter on

113) was a technical term in the Methodical school o medicine; c. MichaelFrede, Te Method o the So-Called Methodical School o Medicine, in Essays in AncientPhilosophy(Minneapolis 1987) 263-66. It was extensively used by Galen in both medicaland logical contexts; c. Barnes, Galen on Logic and Terapy, 98-100; Fridol Kudlien, Endeixis as a Scientifc erm, in F. KudlienR.J. Durling (edd), Galens Method o

Healing(Leiden etc.: Brill, 1991) 103-111; Hankinson, Galen On the Terapeutic Method,202-206; Philip J. van der Eijk, in Cambridge Companion to Galen, 292. Galens use o is oten compared with the concept o an indicative sign ()known to Sextus (PHII 99-101; M. VIII 151-6), a concept that is probably o medicalorigin itsel; c. David Sedley, On Signs, in J. Barnes, J. Brunschwig, M. Burnyeat,M. Schofeld (edd.), Science and Speculation (Cambridge: CUP, 1982) 241. and note 8;James Allen, Inerence rom Signs(Oxord: OUP, 2001) 87-146, esp. 106-122. Tere doesnot seem to be any evidence o a similar use o outside the medical literature. Is itnoteworthy that in Inst. log. 11,1 Galen makes a distinction between and on the grounds that the ormer does not proceed by the rules o syllogism (c.Phillip de Lacy, Galen on Te Doctrines o Hippocrates and Plato [CMG V,4,1,2, Berlin:Akademie Verlag, 1984, 3 vols.] 658). Eectively the same distinction is made by Clem-ent in Str. VIII 6,1.


29/34


causes (25-33) is most probably medical.114 It does not seem unthinkablethat a substantial work dedicated to the problem o demonstration would

include a section on causes.115 Considering how much attention Galenpays to the problem o causes in his extant writings, we might entertainthe possibility that he discussed it also in his treatise On Demonstration.Be it as it may, the indications mentioned above suggest that the doctrinalramework o Clements source could be specifcally medical. But i this isa sound conclusion, the hypothesis o a scholastic tradition so narrowlyspecifed, o which there is no trace in our sources, would seem ratherarbitrary.

I we reject the idea o a scholastic tradition in which the similaritiesbetween StromataVIII and the writings o Galen were common, we arelet with two options. Either Galen draws his theory o demonstrationrom a non-Galenic source which is also the source o Clement, or Clem-ent draws his theory o demonstration rom a lost writing o Galen.Again, the problem with the frst explanation is that it posits somethingo which we have no evidence. As or Galen, he does not appeal to any

114) C. Jean-Jol Duhot, La conception stoicienne de la causalit (Paris: Vrin, 1989) 211-

235, who argues that the character o some passages in this amalgam o Stoic and Aristo-telian elements is indisputably medical. C. Str. VIII 28,7; 30,1; 31,4-5; 32,7; 33,1-9;Duhot, La conception stoicienne, 221, 224, 226, 232-4, 235. We may add two observa-tions: (1) In Str. VIII 25,2, having explained that procatarctic causes provide the occasionor something to happen, Clement adduces the ollowing example: beauty, when seen byan incontinent person, creates in him the erotic condition () butdoes not necessitate its ulflment. Tis is oten explained as a Stoic description. However,the word is not used in the sense attested or the Stoics, namely as an enduringstate which additionally does not admit o degrees (Anthony Long, David Sedley, TeHellenistic Philosophers, vol. I [Cambridge: CUP, 1987] 376), but rather describes a tran-

sient inner condition. Tis usage might be labelled as Aristotelian, but the description othe procatarctic cause as an external actor that provides or some has closer par-allels in medical literature, particularly Galen; c.MMX 242-9; Caus. Puls. IX 2-3; Hank-inson, Galens Teory o Causation,ANRWII 37.2 (1994), 1766. (2) In Str. VIII 32,4,Clement makes a distinction between pre-evident and non-evident causes and adds thatwhile the ormer are grasped , the latter are grasped . Again, thedistinction between and , and the way it is applied here, closelycorresponds to the medical usage attested by Galen; c. esp. SI11 (Khn I 77,14-78,6);or the context c. Fredes introduction to Galen, Tree reatises on the Nature o Science(Indianapolis 1985) ix-xxxiv. For the history o the concept o epilogismos c. MalcolmSchofeld, Epilogismos: An Appraisal, in M. Frede, G. Striker (edd.), Rationality inGreek Tought(Oxord: OUP, 1996) 221-237.115) C. Aristotle, An. post. II 11; Clem. Str. VIII 18,1. C. Le Boulluec, Extraits, 116.


30/34


authority on the demonstrative method later than Aristotle and Teo-phrastus. He regards their Posterior Analyticsas the best accounts written

about this topic so ar, and claims to have clarifed some o their ratherunclear and brie statements in his work.116 It is true, he tell us that romhis young age he was interested in what philosophers, all the amous Sto-ics and Peripatetics o his time, had to say that could be used or demon-strations. But he claims to have ound very little (), and eventhis was controversial, i not downright absurd.117 He gives credit to hisather or introducing him to geometry, mathematic and arithmetic.Tese disciplines saved him rom scepticism by providing the geometricalmethod o proo. He presents his own workOn Demonstration as a devel-opment o this method, to be used by those already trained in the linearproos o geometry.118 Galen might have exaggerated his achievement. ButI wonder i he could have claimed so much originality or his work ondemonstration i there were in circulation an earlier treatise on the samesubject, elaborated, at least to some extent, in a similar manner, and pos-sibly written by a doctor, a work wide-spread enough to reach Clement oAlexandria sometime around the turn o the 2nd and 3rd century. At anyrate, neither Galen nor anyone else indicates that such treatise ever

existed.

6. Te Galen hypothesis

It remains to explore the possibility that Galen himsel is the source o thepassages in StromataVIII discussed above. Surprising as it may sound, thisexplanation should not be dismissed too lightly.119 Chronologically it is

116)

C. PHPII 2,4; II 3,1 (CMG V 4,1,2: 104,3-5; 108,22-25).117) Lib. Prop. (Khn XIX 39,17-40,4).118) Khn XIX 40,5-41,12. On geometrical (linear) proos in Galen c. Barnes, Proosand Syllogisms in Galen, in Galien et la philosophie (Genve: Fondation Hardt, 2003)1-24.119) Solmsen, the only scholar known to me who considers the possibility that Galenmight be Clements source, quickly dismisses it as dramatization, preerring to lookupon Galen as representative o a trend. According to Solmsen, to think o Clement asworking his way through 15 technical books about [demonstration] puts a strain on theimagination. (Early Christian Interest, 286). However, Solmsen is not aware o the ullscale o parallels, including some characteristic details that render the hypothesis o atrend unlikely. O course, the explanation that Galen is Clements source does not nec-essarily imply that Clement worked his way through all the fteen books o Galens trea-


31/34


possible. It is true that there are not many dates in Clements career thatcan be fxed with certainty. But we know that he wrote the frst book o

Stromata ater the death o Commodus (192)120 and it is reasonable tosuppose that the terminus post quem or the text known as StromataVIII isnot earlier than that.121 As ar as Galen is concerned, his voluminous workOn Demonstration was written beore 162.122 We know that he later com-posed numerous treatises in which he elaborated various aspects o histheory o demonstration and two o themOn the Demonstrative Discov-eryand Summary o the Teory o Demonstrationseem to have been occu-pied with the theory o demonstration in general.123 Te dates o theselater writings are unknown, but the date oOn Demonstration suces toallow or the chronological possibility o Clements acquaintance withGalens method o demonstration.

Do we have any evidence that Clement was acquainted with Galen?Not to my knowledge. But there are reasons to believe (1) that Clementcouldhave heard about Galen, and (2) that he couldhave been interestedto acquire Galens work on the demonstrative method. Tat Galen wasamous among his contemporaries is attested by Alexander o Aphrodisiaswho ourished in approximately the same time as Clement. In an ot-

quoted passage Alexander mentions Galen as an example o a man orepute (endoxos) alongside with Plato and Aristotle.124 An important pas-sage in Eusebius, quoting an anonymous heresiological source, indicatesthat sometime at the beginning o the 3rd century there were Christiansin Rome who held Galen in such high esteem that, according to Eusebius

120) C. Str. I 144,3-5.121) In order to date StromataVIII more precisely, we must examine the question o its

chronological relation to other books o Stromata(especially those where parallel motisare ound), as well as the role o StromataVIII in Clements overall project. Both thesequestions are controversial and cannot be addressed here. Modern discussion o the dateso Clements writings is well summarized by Carl P. Cosaert, Te ext o the Gospels inClement o Alexandria(Atlanta 2008) 15-18.122) C. von Mller, ber Galens Werk, 411-14.123) Lib. Prop. (Khn XIX 44,17; 45,4). For the list o Galens books useul or demonstra-tions c. Morison, in Cambridge Companion to Galen, 66.124) Alexander, In op. 8,5 (CAG II.2: 549,24). For Alexanders acquaintance with Galenc. Nutton, Galen in the Eyes o His Contemporaries, in From Democedes to Harvey(London: Variorum, 1988) III, esp. 318-320. Alexanders date is determined on the basis oa dedication to the emperors Septimius Severus and Caracalla in the frst lines o his De ato(CAG suppl. II.2: 164,3); the dedication must have been written sometime between 198


32/34


source, some o them perhaps even worshipped him.125 Eusebiusdescribes them as the ollowers o Teodotus the Shoemaker who was

excommunicated by the Roman church in the last decade o the 2nd cen-tury.126 One o the main errors o this group, as perceived by the churchauthorities, was their tendency to interpret the divine Scriptures by meanso syllogistic fgures. In order to improve their exegetical abilities theystudied geometry, especially Euclid, and the writings o Aristotle, Teo-phrastus and Galen.127 As Nutton observes, modern scholars are agreedthat what these Christians were doing was erecting a deence o Christi-anity (. . .) by the use o scientifc, demonstrative logic.128 We can saelyassume that Galen was popular among them because o his logical writ-ings, especially the writings on the demonstrative method. Teir choice oGalen may have been motivated by the act that he wrote the most well-known, most elaborate and most accessible account o the topic that wasavailable at that time. In addition, it may also have been provoked by theamous doctors interest in Christianity and his criticism o its demonstra-tive impotence.129 We have seen that Clement, too, tried to employ thedemonstrative method as a tool o biblical exegesis.130 And he responds tosimilar objections against Christian aith as were those ormulated by

Galen.131

Had Clement ever heard o Galens works on demonstration, hewould have been intrigued to read them or precisely the same reason asthe Teodotians were. But i around Clements time Galens reputation

125) Eusebius, Eccl. hist. V 28,14 (Bardy).126) 28,9.127) 28,13.128) Nutton, Galen in the Eyes o His Contemporaries, 316.129) C. Hermann Schne, Ein Einbruch der antiken Logik und extkritik in die altchrist-

liche Teologie, in . KlauserA. Rcker (edd.), Pisciculi: FS F. J. Dlger(Mnster 1939),262: Die Achtung des Galenos vor der im Leben bewhrten berzeugungstreue der Chri-sten au der einen Seite und au der anderen Seite seine missbilligende Verwunderungber ihre Unzugnglichkeit r Deduktionen machen es begreiich, dass gebildete Chri-sten im Anang des 3. Jh. die Mngel auszugleichen versucht haben, die der berhmteLogiker an ihnen empunden hatte. For Galens attitude to Christianity, c. the classicstudy by Richard Walzer, Galen on Jews and Christians(London 1949).130) Str. VIII 2,1.4. C. Apostolopoulou, Dialektik, 93: So ist der Beweis, wie ihn Kle-mens versteht, die Methode der Interpretation der Bibel. At least on one occasion (Str. VI121,2) Clement detects a demonstrative syllogism in the Bible itsel; c. Apostolopoulou,Dialektik, 87.131) C. e.g. Str. II 8,4-9,6, where Clement tries to explain to the Greeks in their terms whyaith cannot be proved. C. also II 24,2-3; V 18,3, and reerences listed above, note 38.


33/34


was already o such good standing as to be compared with that o Platoand Aristotle, we have a reason to think that Clement could have heard

about him too.Tese considerations, I believe, allow us to propose the ollowing the-

sis: the parallels between Clements account on demonstration and thewritings o Galen are due to the act that StromataVIII 3,1-15,1 drawrom a lost writing o Galen about the doctrine o demonstration. I leaveopen the question whether this lost writing is a part (presumably the frstbook or more) o Galens treatise On Demonstration, or whether it is somelater compendium concerned with the same topic. Its solution partlydepends on another question, which I leave open too, namely whetherand to what extent the remaining chapters o Stromata VIII could bealigned to the same source. Tis as well as other problems, with which thepuzzling bequest o the Alexandrian teacher conronts us, must bereserved or another inquiry.


34/34

Copyright of Vigiliae Christianae is the property of Brill Academic Publishers and its content may not be

copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.

Documents

Galenus Stromata VIII