You are on page 1of 8

The Editors of the "Metaphysics" Author(s): Stephen Menn Source: Phronesis, Vol. 40, No. 2 (1995), pp.

202-208 Published by: BRILL Stable URL: . Accessed: 08/03/2014 19:37
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .

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

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

This content downloaded from on Sat, 8 Mar 2014 19:37:52 PM All use subject to JSTOR Terms and Conditions


The Editors of the Metaphysics


Christopher Kirwan writes as follows in his introductory note to Metaphysics F:

We are told that the fourteenbooks of Metaphysicswere broughtinto their present arrangementby editors after Aristotle's death. Book A, which they set at the beginning, describes the aim of philosophy as the removal of surpriseand perplexity by supplying "knowledgeof original causes," and assesses the work of Aristotle's predecessorsin thatfield. After the shortbook designateda, B outlines a set of most of which get examined, more or less directly, in the rest of the '"perplexities," treatise. F thus stands, by the traditionalordering,at the start of Aristotle's main in the first chapter;and discussion of metaphysics;it announcesits subject-matter its argumentis hardlymore dependenton what has precededthan on otherpartsof Aristotle's works.'

Kirwan thus claims that the Metaphysics as we have it has been heavily processed by post-Aristotelian editors, whose work extended even to placing Book A at the head of the treatise: presumably they had found it as an independent treatise, and decided that it would make best sense as an introduction to a collection of Aristotle's writings on metaphysical topics. This claim is importantfor Kirwan's interpretationof Metaphysics r, since it justifies him in taking F as effectively the beginning of Aristotle's discussion of metaphysics: so Kirwan takes F as defining this science in its opening chapter, and then as addressing problems in the science so defined, ratherthan as adding further specifications to a science described in A or a or B. This issue is important because, on Kirwan's interpretation, metaphysics would have to be by definition the science of being as being and its per se attributes. By contrast, if Aristotle himself intended AcaB to be read before F, then since these earlier books describe oowcia or i 6TL(YT,[r t714TOU[tvr not as a science of being but as a science of first principles, we
' Aristotle's Metaphysics, Books F, A, and E, translatedwith notes by Christopher Kirwan(Oxford, 1971), p.75. The second edition (Oxford, 1993) reproducesthe original edition with new supplementarymaterial. The passage here cited is not modified by anythingadded in the second edition.


Plironesis 1995. Vol XU2 (Accepted Februanr1995)

This content downloaded from on Sat, 8 Mar 2014 19:37:52 PM All use subject to JSTOR Terms and Conditions

would more naturally read Fl as making a new statement about the science already described: the statement, namely, that the first principles will be principles of being as such (and of its per se attributes such as unity and plurality), and that therefore the desired knowledge of principles will be found by a scientific investigation into being, i.e. by a search for causes of being as such (and of unity and plurality).2Since the claim (which is by no means Kirwan's alone)3 that the arrangement of the Metaphysics is postAristotelian has implications for the interpretation of F and subsequent books, it would be worth investigating the basis for this claim. Kirwan says, "we are told that the fourteen books of the Metaphysics were brought into their present arrangement by editors after Aristotle's death." He does not say who told us. A reader might easily draw the conclusion that there was some extant ancient source that said this. But there is not.4 We have in fact very few ancient testimonies about the history of the text of the Metaphysics: and the majority are merely attempts to explain why the

For Kirwan, since metaphysics is defined as the science of being qua being, the question of the object of metaphysics is simply a matterof interpretingthe restrictive phrase "qua being"; and it is easy enough to dispose of Philip Merlan's view that this restrictsthe object of metaphysicsto certain special beings, those separatefrom matter (Kirwan,pp.77-8, togetherwith pp.201-3 in the second edition). But the real problemfor of metaphysics is that the majorityof texts, including all the ontological interpretation those outside the Metaphysicsand all those in the Metaphysicsbefore F, describe "wisdom" or "firstphilosophy"(or the like), not as a science of being, but either as a science of first causes or as a science of a special kind of beings (divine or eternal or immaterial); see the texts collected in Vianney D6carie, L'objet de la metaphysique selon Aristote (Montreal-Paris,1961). If [l does not define metaphysics, but merely adds a specification to what has already been defined, then metaphysics need not deal with everything that can be said about being as such, or even with all of its causes, but only with those causes of being that fall into the desired class. ' So, for instance, J.L. Ackrill says that the Metaphysics "consists of a number of treatises ... which were brought together by a later editor" (Aristotle the Philosopher, Oxford, 1981, p. 116); for a similar view see now also JonathanBarnes, "Metaphysics," in his CambridgeCompanionto Aristotle (Cambridge,1995). 4 It is sometimes thoughtthat Werner Jaeger's researches,in his Aristotle: Fundamentals of the History of His Development (English translationby Richard Robinson, second edition, Oxford, 1948), had the effect of showing that the Metaphysicsas we have it was the result of post-Aristotelian editors, who put togethera numberof originally independent Aristotelian treatises (and perhaps even some non-Aristoteliantreatises, since the authenticityof a and K is controversial).But Jaeger himself did not believe this, except inasmuchas he thought that cc,A, K and A had been added by later editors:he thought that the otherbooks had been writtenat differenttimes, and as partsof differentprojects, but that they had been sewn together (not wholly successfully) by Aristotle himself. CertainlyJaegercannotbe cited as an authorityfor the claim that the work of the editors went as far as puttingA in its currentposition. 203

This content downloaded from on Sat, 8 Mar 2014 19:37:52 PM All use subject to JSTOR Terms and Conditions

containsa book called A and a book called a.5 But there are Metaphysics two texts of Asclepius (circa 500 AD) and one of the pseudo-Alexander (probablyMichael of Ephesus,circa 1100 AD), which seem to promise more.As far as I have been able to tell, it is the two texts of Asclepius- as or so I shall argue,by scholarsincludingRoss - that have misinterpreted, testimonyaboutthe editorsof thatwe have authentic led to the impression
the Metaphysics.6

that"with to his editionof the Metaphysics, Ross says, in the introduction regardto the time at which the varioustreatiseswere put togetherto form
the Metaphysics we have very little to go upon."7 He then cites, without

andone of Asclepius: of the pseudo-Alexander one testimonium endorsing,

passages were "placed Alexander(515.20) expresses the opinion thattwo particular together by Aristotle but separatedby Eudemus."Asclepius (4.9) has a different story, that Aristotle sent the whole work to Eudemus,who thoughtit unfitting"that so great a work should be published";and that after his death, and the loss of parts of the book, later scholars filled up the gaps by drawing upon Aristotle's other works and piecing the whole togetheras best they could.8

of AscleRoss cites the othertestimonium In the body of his commentary, on A2, Ross notes thatthe chapteris "almostword for pius. Commenting He then addsthat"Asclepius wordidenticalwith Physics 194b23-195b21." 305.19 tells us that 'they' (the editorsof the Metaphysics)'said that some partsof A had been lost, andthey suppliedthe deficiencyout of Aristotle's own writings"'- i.e. that they restoredthe missing passage by simply copyinga passagefromthe Physics.9
s These texts reporta traditionthat either A or a is not by Aristotle but by a nephew of Aristotle's disciple Eudemus of Rhodes, named either Pasicles or Pasicrates;see discussion below, with referencesin n.10. None of the extant ancient writerswho transmit this story endorse it themselves. 6 Kirwan and others may also have been influenced by Porphyry'scomparisonof his (Vita own work in editing Plotinus to Andronicus'work on Aristotle and Theophrastus But there Plotini 24), althoughPorphyrydoes not mentionthe Metaphysicsin particular. is no sign that Porphyrythinks Andronicus'work involved creatingnew large treatises by stitching small ones together. Porphyrysays that Andronicus"collected"or brought together different works of Aristotle, as Porphyrycollected Plotinus' treatises into the six topically linked Enneads:this is quite differentfrom what an editor would have had to do in constitutingthe Metaphysicsout of smallertreatises,which would at a minimum involve inserting a great many forward and backwardreferences, especially between Metaphysics B and subsequent books (there is nothing like this in any of Porphyry's Enneads). 7 Aristotle's Metaphysics, a revised text with introductionand commentaryby W.D. Ross, 2 vols. (Oxford, 1924), v.1, p.xxxi. 8 Ibid. Ross cites the Alexander and Asclepius commentarieson the Metaphysics by page and line from the editions in the Commentariain AristotelemGraeca (in vols. 1 204

This content downloaded from on Sat, 8 Mar 2014 19:37:52 PM All use subject to JSTOR Terms and Conditions

These two passages of Asclepius are the only texts that offer any hope of giving us an authentic chain of testimony going back to the original postAristotelian editors of the Metaphysics. The second passage is particularly suggestive, because (as Ross interprets it) it seems to imply that Asclepius knew (at first or second hand) of a declaration by the editors themselves, made perhaps in a preface to their edition. Unfortunately, Ross has misinterpreted Asclepius. The verb "they said" [RXryov]at Asclepius 305.20 has no antecedent, and Ross supplies the deficiency by supposing that the unnamed subjects of eXt-yov are the same as the unnamed subjects of "supplied the deficiency" [&p1QtoOGav] at 305.21, who must be the editors of the Metaphysics. This is initially plausible; but a closer examination will show that Asclepius is not claiming any knowledge of anything the editors of the Metaphysics may have said, and that he is simply reporting a schooltradition designed to exculpate Aristotle from the apparent faults of the received text of the Metaphysics. This becomes clear if we look first at the passage from Asclepius' prolegomena to the Metaphysics, which includes the story of Aristotle's sending the work to Eudemus. Among the various topics to be discussed before beginning his line-by-line exposition of the text, Asclepius notes "the manner of composition" of the work (Asclepius 4.4). He says:
The present treatise has not been fitted together in the same way as Aristotle's others, nor does it seem well-orderedand continuous; rather,it seems that some things are missing for the continuityof the argument,that others have been brought in [[tevTvqxOat] bodily from other treatises, and that he says the same things several times. But they say in his defense, defending him nicely [;taoXoyorVTCnt U t6;Q TOuOT xCL xakXig 6MoWoyoUvTat], that after he wrote the present treatisehe sent it to his companion Eudemusof Rhodes, but then, as it turnedout, he [Eudemus]decided that it would not be right to publish to the many so great [so exalted? so long?] a treatise.Meanwhile he [Eudemus]died, and some partsof the book were destroyed;and the men of later generations,not daring to add anything of their own, since they fell far short of that man's [Aristotle's] thought, supplied the missing parts [6t(?6octv-te; rT XlaovToaJ by bringing them in [[LePTyaCyov] from his other treatises, as far as was possible. Nonetheless you will find that the sequence of the argumentis preservedin these [partsof the text] too. (4.4-16)

Asclepius does not suggest that there is any direct evidence for this story: it is simply one story that has been proposed in the school, whose attractionis that it explains the current state of the text in such a way as to deny Aristotle's responsibility for its objectionable features. But what does Asand 6 respectively), and so will I. As Ross knows, the passage cited from the Alexander commentaryis from the spuriouspart. Ross, v.1, p.292. Ross himself does not accept this explanation, and thinks (surely correctly)that Aristotle himself used the same passage in two differentcontexts. 205

This content downloaded from on Sat, 8 Mar 2014 19:37:52 PM All use subject to JSTOR Terms and Conditions

clepius find objectionable? I am not sureexactlywherehe thinksthingsare "missingfor the continuity of the argument," but the otherobjectionscorrespondto specific featuresof the text: the objectionthatsome things "have been broughtin bodily from other treatises" refersto the duplications between Metaphysics A2 and Physics 11,3, and between Metaphysics K8-12

anda seriesof chapters fromthe Physics;andthe objectionthat"hesays the same thing severaltimes"within the boundsof the presenttreatiserefers, similarly,to the duplications betweenMetaphysicsK1-7 and Metaphysics BFE, and betweenMetaphysics M4-5 andMetaphysics A9. For let us look backat the laterpassagefromAsclepius,the one Ross took as referring to a declaration by the editorsof the Metaphysics:
These remarks[i.e., Metaphysics A2] have been brought in [RETEv'vExntELI here from the Physics [#x Tiq (puoatxi dtxQotorw;]; for they said [9Xcyov]that some things had been lost, and being unable to imitate them, they suppliedthem [(NpfpLoactv] out of his own works. Anyway [dtFqXrL TOL], the whole letter K as well las the presenttext] is from his other lectures [dtxQodaowv].(305.19-22)

This interprets the first passage, by makingit clear that the evidence for in the form of copying other Aristotelian editorialintervention, texts into
the Metaphysics, comes from Metaphysics A2 and Metaphysics K: a Greek

commentator's "X, &[LtEXEt tOL Y" is a sure-firesign that Y is his only evidence for X. When Asclepius says that "the whole letter K is from his otherlectures," it seems thathe mustbe thinkingnot only of K8-12, which are from another Aristoteliantreatise,but also of KI-7, which are from anotherpart of the same treatise. So this passage explains not only the in bodilyfromothertreatiscomplaint that"somethingshave been brought es," but also the complaintthat "he says the same thing several times" within the Metaphysics; and both of these complaintscan be answeredby did not have these faults,and thatthe saying thatthe originalMetaphysics editorsintroduced themby copying, whetherfrom the Physics or fromthe that had been destroyed. Metaphysics,to replace parts of the manuscript (And whereverAsclepiusthinkssome thingsare missingfor the continuity of the argument, this can be explainedthe same way: a partof the manutext scriptwas missing,andthe editorscould not find any otherAristotelian to patchit up with.)The subjectof "theysaid"is not any historical witness, butthe sameanonymous certainlynot the editorsof the Metaphysics, "they" who offeredthe "nicedefense"of Aristotlein the earlierpassage;and their in the text. only evidenceis the fact of duplication is not that Asclepiushas no directevidence What is perhapssurprising about the editors of the Metaphysics,but that he imaginestheir work as simplyfilling in holes in the text:he assumesthattherewas an originalA2, either an original A9 or an original and an original K (and presumably

This content downloaded from on Sat, 8 Mar 2014 19:37:52 PM All use subject to JSTOR Terms and Conditions

M4-5 differentfrom the presentones), and that the editorswere forced to patch them up with other Aristoteliantexts because the originalchapters were lost. If it has occurredto AsclepiusthatAristotlehimself mighthave left the text in an unfinishedstate, with repetitionsand perhapswith older of of the same material, thenhe rejectsthe attribution and newertreatments to him thatthe suchhumanweaknessto Aristotle.And it has neveroccurred featuresas the orderof the editorsmightbe responsiblefor such structural books of the Metaphysics (beyondpossibly insertinga spuriousA or a, a chargeAsclepiusrejects);or whatcould lead him to assumea lost Ur-K? Thereis one point, though,on which Asclepius gives us some valuable of his day. The "nice defense" evidence,at least aboutthe school-tradition involvedAristotle'ssendingthe Metaphysics to Eudemus,Eudemus'refusimpliedthathe had the only copy), andthe ing to publishit (it is apparently of the text while in the custodyof Eudemus'school. corruption subsequent alludedto above: This fits well with the commentof the pseudo-Alexander commentingon Z 11 1036b32-1037b5,and pointing out that it treats the same subject as ZIO 1034b24-1035al7, he says "and I think that these [words]oughtto be put next to those, and most likely [tow;] they were put togetherby Aristotle(for in none of his other treatiseswill you find him doing the kind of thing that appearshere [i.e. writingin such a disorderly way]), but they were separatedby Eudemus"(515.19-22). It is not clear by whatexactlyEudemusis supposedto have done (if he was not motivated fell out sheermalice to mutilatethe text, perhapsa piece of the manuscript and he put it back in the wrong place); what is clear is that the pseudoAlexander,like Asclepius, thinks that the Metaphysics is more disorderly thatwhatthanAristotle'susualwriting,and knows the scholastictradition ever is wrongwith the currenttext of the Metaphysics can be blamedon a time when it was in the custody of Eudemusand his school. This same after (by Asclepiusimmediately mustunderliethe storiesreported tradition the "nice defense"passage, by the pseudo-Philoponus, and in two anonyfact of a book called A and a mous scholia) to explain the embarrassing book called a, by attributing eithera or (as Asclepiushas it) A to a nephew has it) Pasof Eudemuscalled eitherPasiclesor (as the pseudo-Philoponus it is unclearhow such a book would have been slipped into the icrates:'?
'0 Asclepius discusses the problemabout A at 4.17-35 (Pasicles); the pseudo-Philoponus briefly mentions the controversyabout a (Pasicrates)at loannis Philoponi enarratio in omnes Aristotelis libros quos metaphysicos appellant, Latin translationby Francesco Patrizi (Ferrara, 1583; the Greek original is lost), p.7, reprinted in Commentariain AristotelemGraeca, versiones latinae temporisresuscitatarum litterarum,v.2 (StuttgartBad Cannstatt,1991). The scholia are cited by Paul Moraux,Der Aristotelismusbei den Griechen, v.1 (Berlin-New York, 1973), p.83, n.89.


This content downloaded from on Sat, 8 Mar 2014 19:37:52 PM All use subject to JSTOR Terms and Conditions

Metaphysics,but it must have been done by Eudemus'school at a time when no copies of the text were availableelsewhere. The tradition, then, that the text of the Metaphysics comes to us, edited and in some way damaged,from Eudemusand his school, is widespread amongthe Byzantine commentators andmustgo backat least to Ammonius the son of Hermias(whose lectureswere the main source for Asclepius' commentary); it may be mucholder.This tradition is likely to be connected with whatever lies behindthe title "Eudemian Ethics,"andit is possiblethat it reflects some historicalreality,but also possible that it reflects nothing morethanschool-polemics. To conclude:we have not been "told,"as Kirwansays we have, that editorsafterAristotle'sdeathbrought the fourteen booksof the Metaphysics into theirpresentarrangement. We have been told thatthe editorsreceived from Aristotlefourteenbooks of Metaphysics in theircurrent order(except possibly for a or A), and that, perhapsto repairsome damage,they made we might local changeswhich did not affect the overallstructure (although considerit a majorchangeif the originalK has been completelyreplaced). Apartfromthe possiblycorrectgeneraltradition thatthe presenttext comes from the school of Eudemus,those who "tell"us aboutthe historyof the text know no morethanwhatwe know,namely,thattherearedoublets,and that sometimeswe have difficultyconstruing the argument. Of course,we may very well concludethatthe orderof the text as we have it fails in some we mighteven concludethatwe majorway to reflectAristotle'sintentions; have before us fourteenindependent treatiseson metaphysical topics; but we cannotdrawsuch conclusionsfrom authority. We can only drawthem if, afterseriouseffort,we areunableto makesense of the text as we have it. F "announces its subjectKirwanmay possibly be right that Metaphysics matterin the first chapter" and that it and laterbooks of the Metaphysics more dependent proceedto explorethe science first definedthere,"hardly on whathas precededthanon otherpartsof Aristotle'sworks,"rather than specifyingfurtherthe science of first principlesdescribedin AaB and resolving problemsin that science; but the claim cannotbe allowed to pass withoutargument.
McGill University


This content downloaded from on Sat, 8 Mar 2014 19:37:52 PM All use subject to JSTOR Terms and Conditions