Dao(2011)10:359–375 DOI10.1007/s11712-011-9231-4 Holding an Aristotelian Mirror to Confucian Ethics? XIAO Yang Publishedonline:10July2011 #SpringerScience+BusinessMediaB.V.2011 Keywords Aristotle’sethics.Confucianethics.Structureofanethicaltheory.Virtue ethicsversustheoryofethics.Methodologyofcomparativephilosophy. Savingthephenomena ItisanhonortobepartofthisbooksymposiumonYUJiyuan’ssplendidbook.The bestwayofrepayingthehonor,Ibelieve,istoengageactivelyandcriticallywiththe issues Yu has raised. However, in this essay I will only be able to discuss selected aspectsofthisextremelyrich,ambitious,erudite,andinspiringbook.Ishallfocuson three of its most impressive achievements. The book’s first achievement is that Yu has redefined comparative philosophy by doing it “critically” and “philosophically.” He does it “critically” in the sense that both sides of the comparison are subjected to critical examination; neither is regarded as an unquestioned authority. It is one of the book’s strengths that the authorisanAristotlescholarbytraining,andthebookisasmuchaboutConfucians asaboutAristotle.1Infact,ashehasmadeConfucians seem moreAristotelian, Yu’s Aristotle has taken on some Confucian characteristics. Yu also does comparative philosophy “philosophically” in the sense that he tries to discover whether the two ethical systems are true. Yu thinks that it is not enough to have provided coherent interpretations of Confucian and Aristotle’s ethics; he also wants to know whether each of them is true. In other words, Yu is interested in not only getting the right interpretations of Confucian and Aristotle’s ethics but also whether they have got ethics right. In the end he reaches the conclusion that “neither Aristotle nor the Confucians are completely right or completely wrong” (Yu 2007: 6). Yu’s second achievement is that he has given us coherent interpretations of ConfucianandAristotle’sethics.Thefirstthingareaderwouldnoticeaboutthebook is that Yu’s style of thinking can be characterized in Aristotelian adjectives, such as “systematic,”“meticulous,”“patient,”and“perceptive.”ButYuisalsoanAristotelian in a deeper sense: he has mastered Aristotle’s method of “saving the phenomena,” as 1Yu’sfirstbookisamonographontheconceptofbeinginAristotle’smetaphysics(Yu2003). XIAOYang (*) PhilosophyDepartment,KenyonCollege,Gambier,OH43022,USA e-mail:[email protected] 360 XIAOYang wellastheAristotelianartofdistinction-making.ThereaderwouldfindYuconstantly making classifications and distinctions, and classifications within classifications and distinctionswithindistinctions. Since the Aristotelianart ofdistinction-makingisone of the most effective ways to clarify and explain away contradictions and inconsistencies, not surprisingly, as one reads Yu’s book, clear and coherent pictures of Confucian and Aristotle’s ethics with mirroring structures gradually emerge. Yu’s third achievement is that he has given us the most systematically articulated pictureofwhatConfucianethicswouldbelikeifitsbasicstructureswereparallelto the basic structures of Aristotle’s ethics (with various degrees of dissimilarities and contrasts on different levels). One way to summarize the book is to say that it consists of a series of what I shall call Yu’s “parallel structure theses” (PS theses), together with detailed and nuanced arguments and textual exegesis that support them.2 Yu also argues for a general thesis, which can be formulated as follows: (GT)ConfucianethicsisavirtueethicsthatisstructurallysimilartoAristotle’s virtue ethics. Here is a road map of this essay. It is divided into four sections: each of sections 1, 3, and 4 deals with one of the three achievements. In section 2, I subject two aspects of Yu’s book to critical and philosophical examination. 1 Doing Comparative Philosophy Critically and Philosophically Inthissection,IfirstdiscussapossibleobjectiontoYu’sproject,whichismotivated byconcernsovercomparativephilosophyingeneral.FollowingBenjaminSchwartz, Ishallcallitthe“appealingtothe‘latestword’fromtheWest”objection.Ishowthat Schwartzhasagoodsolutiontoit(whichIshallcallthe“Schwartzianstrategy”);as we shall see, such a strategy can also be found in Yu’s book. YuarguesthatitisnecessarytocompareConfucianandAristotle’sethics.Making use of what he calls “Aristotle’s friend-as-mirror thesis,” he states that Confucians andAristotleare“friends”inthesensethattheirethics“canbeviewedasmirrorsfor each other” (4). One reviewer of Yu’s book has rightly pointed out that Confucius andAristotlecannotliterallybe“friends”:“AristotleandConfuciuswerenotfriends. They never met and they never saw each other. They did not ‘spend time together’ and‘sharelives,’aswouldberequiredforbeingfriendsinAristotle’ssense”(Wenzel 2010: 306). However, this is how Yu states his thesis: One lives one’s own life, but still needs friends in various ways. Similarly, we mustreadtheoriginalworkofAristotleandConfuciustounderstandthem,but acomparisoncouldhelpthembebetterunderstood.Takingthemasmirrorsfor each other leads us to reflect upon the traditional roots of both ethics, to examine their otherwise unexamined presuppositions, and to generate alternative perspectives to determine why each proceeds in the way it does. (Yu 2007: 4; emphasis added) 2IfoneputstogetherallthemajorthesesYuhasdefendedinthebook,onewouldnoticethatalmostallof themare“parallelstructuretheses.” HoldinganAristotelianMirrortoConfucianEthics? 361 It is important to note that Yu clearly intends to speak metaphorically when he says that the Confucians and Aristotle are “friends,” by which he simply means that we should be “taking them as mirrors for each other.” Ifwemustdothecomparison,howshouldwedoit?Yuhasmade itveryclearat the beginning that both sides of the comparison should (and will) be treated equally and subject to critical examination: When comparison is used in the study of non-Western philosophy, Western philosophyisusuallytreatedassomeestablishedframeworkortoolofanalysis to be applied rather than as a subject matter that is itself subject to investigation. The focus of discussion has always been on the non-Western side. In contrast, in this book, although we appropriate Aristotle’s methodol- ogy,hisethicaldoctrinesarealsotheobjecttobestudied.Thisbooktreatsboth sides equally and aims at developing an interpretation of each side through comparison. (Yu 2007: 3)3 Whyisitimportantandnecessarytosubjectbothsidesofthecomparisontocritical examination? Why does Yu’s book have to include so many materials on how Aristotle’s ethics should be critically interpreted and constructed? I want to suggest that this is not because Yu happens to be an Aristotle expert, but rather it is the best waytorespondtooneofthemostpowerfulobjectionstocomparativephilosophyin general. Someone may praise Yu’s book by saying that Yu has validated Confucian ethics by invoking its structural resemblances to Aristotle’s virtue ethics, which is one of the most exciting rediscoveries or revivals in contemporary moral philosophy in the West. However, to other people’s ears, they may not hear such words as praise. Rather they may hear them as reasons to reject Yu’s approach. It is indeed the case that Aristotelian virtueethicsisall the rage at themoment.But why should we take the “latest word from the West” as the “unquestioned authority,” to put the point in Benjamin Schwartz’s terms? As we shall see, this objection is parallel to Schwartz’s objection to A. C. Graham’s similar praise of Herbert Fingarette’s book Confucius: Secular as Sacred(Fingarette1972),whichisthatFingarettehasvalidatedConfucius’theory of rituals by invoking its resemblances to Austin’s speech-act theory. This is how Schwartz characterizesGraham’spraise:“Graham, in fact, isdeeply impressed by the fact that Fingarette is able to relate the Analects to certain exciting tendencies of twentieth-century ‘serious philosophy’.… Confucius, we are told, ‘can be relevant to contemporary “professional” philosophy’” (Schwartz 1996: 144). Fromthispassage,asSchwartzpointsout,onewouldgetthe“uneasyimpression” that a new principle of authority is being invoked here by Graham (Schwartz 1996: 144): 3ItseemsthatYu’sapproachisrepresentativeofanemergingtrend,whichcanbefoundinanincreasing numberofworksinthefieldofcomparativephilosophyandChinesephilosophy.Ifwetakethistrendas forming a new tradition of doing comparative philosophy critically, its origin might be traced back to Benjamin Schwartz. Here I disagree with Yu’s reading and assessment of Schwartz. Relying on A. C. Graham’sreadingofSchwartz,YumisreadsSchwartz’sapproachasbelongingtooneofthetwouncritical approachesherejects(Yu2007:226,6n). 362 XIAOYang ThepracticeofvalidatingChinesethoughtbyinvokingresemblancestocertain Western philosophers such as Kant or Hegel without further discussing of the Westernthinkersthemselveshasofteninthepastbeenrecognized asakindof cultural-bound approach. It is not the comparison itself that is cultural-bound but the assumption that one side of the comparison represents unquestioned authority. (Schwartz 1996: 144; emphasis added) Note that Schwartz is not against comparison. In fact, Schwartz insists that he does not have an a priori view that ancient Chinese thought cannot be meaningfully compared to Western thought. He even acknowledges that “it may well be that Austin may in some of its tendencies be closer to ancient Chinese thought than Plato” (Schwartz 1996: 145). His worry is rather that a comparative studymaybe“cultural-bound”inthesensethatonesideisbeingtakenforgranted uncritically: “I would nevertheless submit that in order to validate the Analects by reference to Austin, one must critically examine the claims of Austin himself and even examine what critics of Austin in the West have had to say about him” (Schwartz 1996: 145).4 I believe the worry Schwartz talks about here is the same kind of worry that motivates Yu to adopt a critical attitude toward both sides of the comparison. Furthermore, Yu actually has two concerns here: first, he is keenly aware that there arevariousandoftenconflictinginterpretationsofeitherside,henceonealwayshas to defend one’s interpretation of either Confucius or Aristotle; second, one must be careful nottoautomatically take one side asrepresenting the truth. We shallcallthe first“Yu’sworryaboutinterpretation,”andthesecond“Yu’sworryabouttruth.”Yu’s worry about interpretation is explicitly stated in the following passage: Since one major mistake of doing comparative philosophy is the unreflective imposition of assumptions of one’s own tradition on the other, extra caution will be exerted not to fall into this trap. I will also take into account the rich scholarship on each side, and deal with relevant controversies on either side before a comparison is drawn. (Yu 2007: 6) YualsomakesthegeneralSchwartzianpointbysayingthat“itisdifficulttoseehow onecantakeWesternphilosophyasaready-to-useframework,forthereishardlyany conceptorissueinitthatisnotsubjecttocontroversy”(Yu2007:3).Heemphasizes that this is especially the case with Aristotle: [A]lmost each view of his has been subjected to different and even contradictory interpretations. The Nicomachean Ethics is by no means an exception.Therearenumerousongoingdisputesnotonlyaboutthecontentsof particular views presented in it, but even about the structure of Aristotle’s theory of eudaimonia and about whether the NE is a unified and consistent work. […] Our comparative approach, then, requires a defense of our own 4This emphasis on critical attitude toward both sides of the comparison is a common theme running through all of Schwartz’s work. As early as 1964, in his book on YAN Fu’s encounter with the West, Schwartz has already said, “In speaking of the encounter between the West and the ‘non-West’ we generallyassumethattheWestisaknownquantity.…Weundoubtedly‘know’infinitelymoreaboutthe West,buttheWestremainsasproblematicasever.Onemayevenhopethatthegroundofencountermay itselfprovideanewvantagepointfromwhichtotakeafreshlookatbothworlds”(Schwartz1964:1–2). HoldinganAristotelianMirrortoConfucianEthics? 363 understanding of Aristotle, just as it requires a defense of our own understanding of Confucius. (Yu 2007: 4) Yu has gone even further than Schwartz by suggesting an Aristotelian method to find out how much of Aristotle’s ethics is true (and how much of Confucian ethics is true). This is why Yu has made a major revision of Aristotle’s method of “saving the phenomena” when he applies it to comparative philosophy. He treats both Aristotle’s ethics and Confucian ethics equally as two of the endoxa(reputableopinions),whicharethestartingpointof his comparative inquiry.5 Starting from here, Yu then proceeds to find out how much of Aristotle’s and Confucian ethics is true. This is what he means when he says that he wants to “save the phenomena from both of them” (Yu 2007: 10). And the authority Yu appeals to when he critically assesses them is neither Aristotle nor Confucius, but rather what he (following Martha Nussbaum) calls “human grounding experience”: [H]uman beings live in the same world, possess the same psychic capacities, and share many of the same basic relationships and institutions such as father-son, brothers, family, community, politics, etc. Hence, there is a set of basic desires, feelings, beliefs, and needs which all human beings share and which are necessary for living a human life. This forms the common ground for comparative studies of different cultures. In Martha Nussbaum’s terminology, this common ground can be called “human grounding experience” and she rightly takes them as “reasonable starting point for cross cultural reflection.” The ultimate basis for us to take the ethics of Aristotle and the ethics of Confucius as mirrors of one another and to save the phenomena from both of them is the grounding human experiences. (Yu 2007: 9–10)6 In this sense, this book has gone beyond what is normally meant by “comparative philosophy,”alabelYuhimselfusestodescribehisproject.Ibelieveamoreaccurate descriptionisthatYuhasredefined “comparativephilosophy”byshowingushowit shouldbedonephilosophically.Ashehimselfputsit,hewantstoprovidea“defense of comparative philosophy as a philosophical enterprise” (Yu 2007: 1). Yu’s book should be seen as having given us a paradigm of a new way of doing comparative philosophy, which could be best described by the following slogan: philosophy 5Thisisaveryinnovativeandoriginalidea.Butwecanimaginethatsomemightobjecttoitbysaying thatAristotle’sethicsistheresultofapplyingthemethodofsavingthephenomenatoendoxa,whereas Confucius’ethicshasnotgonethroughsuchaprocedure.TheAnalectscanbeseenasacollectionof“raw endoxa,”namely,“theundemonstratedsayingsandopinionsofthosewhohaveexperienceandareold… whohavepracticalwisdom”(NE,1143b11-4;citedbyKraut[Kraut2006:78]).Soevenifweassumethat bothAristotle’sandConfucius’ethicscanbetakenasendoxaforourinquiry,theyarenot“equal”inthe sensethattheformerisnot“rawendoxon”becauseitiswhathasbeen“saved”attheendoftheprocedure andthelatterisjust“rawendoxon.”Therefore,accordingtothisobjection,itiswrongforYuto“treatboth sidesequally”(3).AnotherinterestingobjectiontoYu’sapplying“savingthephenomena”tocomparative philosophyismadefromtheperspectiveofthe“CambridgeSchool”inintellectualhistory(Lang2009). 6However, formanypeople,this viewof Yu’s(as wellas Nussbaum’s)raisesanother setof problems. Pleasesee,forexample,thedebatebetweenBernardWilliamsandMarthaNussbaum(Williams1985:30– 53;Nussbaum1995;Williams1995).TheseareextremelyimportantissuesforYu’sproject,butYudoes notaddresstheminthebook. 364 XIAOYang should be done “comparatively,” and comparative philosophy should be done “philosophically.”7 2 Critical Examination of Two Aspects of Yu’s Book In this section, I want to engage with Yu’s book in the same critical spirit embodied in the book. My main criticism is that Yu does not always live up to his own exemplar of doingcomparativephilosophycriticallyandphilosophically.Hesometimesdoesnot spread his scrutinizing net wide enough; several important assumptions of the book areuncriticallytakenforgranted.HereIwanttosubjecttwoaspectsofYu’sbookto criticalandphilosophicalexamination.ThefirstisYu’sadoptingacommontypology of ethical theory as the framework for his inquiry, and the second aspect is Yu’s assumptions about the unity of the Confucian texts. Yu’s book focuses on the comparison of Confucian and Aristotle’s ethics. Let us call it the “first pair of comparison”: (1a) “Confucian ethics” vs. (1b) “Aristotle’s ethics.” This is obviously the main focus of the book. Yu subjects both sides of the comparison, (1a) and (1b), to critical and philosophical examination, and tries to savethephenomenafrombothofthem(moreofthisinsection3).However,Yualso groups (1a) and (1b) together as belonging to a distinctive type of ethics, and compares and contrasts it with another supposedly radically different type of ethics, namely “modern Western moral philosophy” (Kantian deontology and consequenti- alism). Let us call it the “second pair of comparison”: (2a) “Confucian and Aristotle’s ethics” vs. (2b) “Western modern moral philosophy.” Yu gives only a very brief discussion of this comparison in the Introduction. Relying on a popular typology of ethical theories as his basic framework, Yu characterizes the second pair of comparison as a divide between two radically differenttypesofethicaltheory.Ishallcallithisassumptionaboutthe“bigdivide” (BD): (BD)Thereisabigdividebetween(2a)and(2b).Theformeris“virtueethics” or“character-basedethics,”whereasthelatteris“action-based,”or“rule-based or rights-based ethics” (see Yu 2007: 2).8 7ThisparadigmhasbeenanticipatedbyAlasdairMacIntyre,oneofthefirsttohavedoneacomparative studyofConfucianandAristotle’sethics(MacIntyre1991,2004a,2004b).AlthoughYuhasexpressedhis disagreementswithMacIntyreoncertainissues(Yu2007:6–10),theysharethefundamentalinsightthatit isimpossibletodophilosophywithoutinsomewaydoingcomparativephilosophy,andviceversa.As MacIntyreputsit,“allreflectiveethicsneedstodevelop,whetherexplicitlyorimplicitly,acomparative dimension”(MacIntyre2004b:152). 8Yudoesnotusetheterm“action-based”there,buthedoessaythat“modernethicsfocusesonmoral acts”(Yu2007:2). HoldinganAristotelianMirrortoConfucianEthics? 365 I wish Yu had not readily and uncritically accepted this popular typology of ethical theories,intermsofwhich(BD)isformulated.Inotherwords,weshouldnottakeit as a “ready-to-use framework,” to put it in Yu’s own terms used in a different but similar context (3). ThefirstthingwenoticeaboutthetypologyofethicaltheoryYureliesonisthatit only classifies ethical theories that can be described as a “x-based ethics,” x being “character,” “virtue,” “rule” or “act,” and so on. For example, a “virtue ethics” is a “virtue-based ethical theory” in the sense that character trait is a “primary concept,” in terms of which all the other concepts are defined; for example, the concept of “morally right actions” is defined in terms of what a virtuous agent would do (Watson 1997; Hursthouse 1999). The notion of “primary concept” is a central notion in this typology. The following is my formulation of the definition of “primary concept”: x is a primary concept in an ethical theory E if and only if (i) x is logically priori to, and independentof,alltheotherconceptsinE,and(ii)alltheotherconceptsinEcanbe defined in terms of x. When this is the case, we call E an “x ethics” (“x-based ethics”). Note that all the ethical theories classified according to this typology share one thing in common, which is that they all have a “hierarchical” structure: at the bottom there is a primary concept and on the upper levels there are other non- primary concepts. I shall call this kind of theories “ethical theories with a hierarchical structure.”9 We now can see that this typology is not a complete classification of ethical theories.Ithasleftoutthoseethicaltheoriesthatdonothaveaprimaryconcept(ora hierarchicalstructure).FromanAristotelianpointofview,thisisobviouslyaserious flaw. Here I mention only two possible configurations of ethical theories with different structures. First, there can be an ethical theory that has a “flat” structure: thereisnoprimaryconceptintermsofwhichotherconceptsaredefined.Second,we canalsoimagineanethicaltheorythathasaglobal“flat”structurebutmultiplelocal “hierarchical structures” with different primary concepts in different spheres of life. Inthissecondtypeofethicaltheories,somelocalstructuresincertainspheresoflife may have a hierarchical structure with a primary concept, even though the global structure is still a non-hierarchical one. CHEN Lai has recently argued that in the Analects, li 禮 (rituals) is not a virtue, ratheritishaoli好禮(lovingrituals)thatisregardedasavirtue(Chen2010).Toputhis point in our terms, we may say that in the sphere of rituals, the concept of “rituals” is a primaryconcept,intermsofwhichtheconceptof“lovingrituals”asavirtueisdefined.10 Note that even though the concept of rituals is a primary concept within the sphere of rituals, it is not a primary concept in the global structure of the ethical theory in the Analects.ThisisbecausenotalltheotherconceptsintheAnalectsaredefinedintermsof theconceptofrituals.SotheglobalstructureoftheethicaltheoryintheAnalectsisstill “flat,” even though there are local structures within it that are “hierarchical.” 9I borrow the term “hierarchical structure” from Julia Annas (Annas 1993). Similar ideas can also be foundinSusanHurley(Hurley1989),whousestheterm“centralism”torefertoanethicaltheorythathas a“central”concept,whichissimilartowhatwehavecalled“primary”concept. 10Thetermli禮(rituals)intheAnalectscouldmeaneither“ritualrules”or“rituallyappropriateactions.” Sotheprimaryconceptherecouldbeeitherritualruleorrituallyappropriateact. 366 XIAOYang Here itmightbehelpfultomake useofa distinction made bysome philosophers recently, namely the distinction between “virtue ethics” and “theory of virtue” (see Baxley 2007). Virtue ethics is a “virtue-based” ethical theory, in which one takes charactertraitsasaprimaryconcept,intermsofwhichtheconceptofrightactionsis then defined. However, we may find a “theory of virtue” within an “act-based” ethical theory. Since this is an act-based ethics, the concept of right actions is the primary concept, and the concept of virtue is then defined as the reliable character traitstodorightactions.Onewillthenhavea“theoryofvirtue”withinanact-based ethical theory.11 To put CHEN Lai’s point in these terms, we may say that there is a “rituals-based”ethicsinthesphereofritualsintheAnalects,butonecanalsofinda “theory of virtue” within it, which defines the virtue of “loving rituals” as the reliable character trait that reliably takes pleasure in following ritual rules (doing ritually appropriate actions).12 HowdowechoosebetweenCHENLai’sandYu’sinterpretationofthestructureof ethicsintheAnalects?Whosereadingisthevalidone?AsIshallargueinsection4, tosettleissueslikethisthereisnosubstitutefornuancedargumentsbasedontextual evidence and philosophical considerations. Until we examine closely their concrete arguments,wedonotknowwhohasgotitright.Fortunately,thepointIammaking here is not about the validity of CHEN Lai’s alternative reading. The point is rather that the possibility of this reading is ruled out from the very beginning when Yu adopts an incomplete typology of ethical theory as the framework of his inquiry. The second aspect of Yu’s book I want to subject to critical examination is Yu’s two assumptions about the nature of Confucian texts. The first is the assumption about the “unity of the Analects”: (UA) The Analects has a coherent unity, which is Confucius’ unified moral vision. The second is the assumption about the unity of the Four Books (the Analects, the Mencius, the Daxue, and the Zhongyong): (UFB)TheFourBookshaveacoherentunity,whichisConfucianvirtueethics. Here I shall only discuss (UA). Most of what I say about it is applicable to (UFB). Yu thinks that (UA) is supported by Confucius’ own self-understanding. After quoting several relevant passages from the Analects, Yu concludes: Apparently, he [Confucius] thinks that his work forms a coherent ethical worldvieworoutlook.Hence,whenwereadtheAnalects,wemustassumethat there is a coherent moral vision that is implicit in the fragmented sayings. A good reader of the Analects must grasp this moral vision. Needless to say, this 11Forexample,inthesecondpartoftheMetaphysicsofMorals,entitledtheDoctrineofVirtue,Kanthas spelledouthis“theoryofvirtue”(or“doctrineofvirtue”tousehisownterms),whichdefinesvirtueasa charactertrait(Baxley2007). 12Ofcourse,itispossiblethatinotherspheresoflife,wemayfindtheconceptofcharactertraits,notthe conceptofrituals,beingtakenasprimary.ItisalsopossiblethatalthoughthisisthecaseintheAnalects,it isnotthecaseintheMencius.Wehavetomakeourjudgmentsonacase-by-casebasis.Inotherwords, theremightbenounityoftheFourBooksintermsofthestructuresoftheirethicaltheories(moreofthis later).Itshouldbepointedoutthatthisismyreading(andreformulation)ofCHENLai,andhemightnot necessarilyagreewithit. HoldinganAristotelianMirrortoConfucianEthics? 367 vision is open to different interpretations, but if we fail to see its unity, we fail to understand Confucius.” (Yu 2007: 14; emphasis added) These are very strong claims. Yu would allow only different interpretations of Confucius’unified moral vision.Thepossibilitythattheideas intheAnalectsmight contain, for example, a hybrid of heterogeneous and inconsistent components or visions, is ruled out a priori. However, Yu does not offer an argument for the transition from Confucius’ self-description before the word “hence” to the conclusion after it. Of course, I am not saying that Confucius’self-description must befalse;Iamonlydenyingthatitmustbetrue.ItmayturnoutthatConfucius’self- description is true, and it may turn out that it is false. Ibelieveitcanbearguedthatitisprobablyfalse.Thecentralpartoftheargument shouldbebasedonnuancedanddetailedtextualexegesis.Inaddition,thereareother considerations;Imentionfourofthemhere.First,theAnalectsisnotasingletextbya singleauthorintermsofitscomposition.Individual“books”(pian篇)oftheAnalects, orsetsofthem,mighthavebeencirculatedbeforetheywereputtogetherasabook.13The receivedtextoftheAnalectsisaproductofalongprocesswithmanyhandsinvolved,and it is not clear how many of Confucius’ sayings in the received text could be safely attributed to the historical Confucius. This fact about the composition of the Analects, of course,doesnotnecessarilyimplythatConfucius’moralvisionmusthavenounity.Butit ispossible (or even probable) that itmight not.14 Second,manycommentatorsandscholarshavenoticednumerousinconsistencies and contradictions in the received text. Third, in the long history of the exegesis of the Analects, commentators and scholars have come up with many interpretations of what Confucius’ moral outlook is, and they are astonishingly pluralistic and heterogeneous, and often in conflict with one another. And furthermore, they all have textual evidence to back up their interpretations.Oneofthewaysto“save”thesescholars’interpretationsistoassume that each of them captures one component of Confucius’ moral vision. As a consequence,wemayconcludethatConfucius’moralvisionhasnounity,butrather it is a complex vision that contains multitudes. The fourthreason why Yu shouldnot have accepted (UA) and(UFB) isthat they arenotconsistentwithYu’scriticalandphilosophicalstyleofinquiry.15Itseemsthat we should not rule out a priori and uncritically the possibility that the Four Books, as well as the ideas embodied in them, might be more heterogeneous and diverse than we have imagined. We do not want to rule out a wide range of possibilities before our inquiry starts. I mention only two of them here: (1) Itispossiblethatitmakesthebestsensetointerpret(ororganize)certainpassagesin theAnalects in terms of Aristotelian structures, but it makes little sense regarding someoftheotherpassagesintheAnalects.(Thispossibilityisruledoutby[UA].) 13This specific hypothesis is based on our general knowledge of how books were circulated in early China(Yu1985).WedonothavedirectknowledgeabouthowtheAnalectswascirculated. 14ItseemsthatYumightallowsuchapossibility,namelyitispossiblethatthefactthattheAnalectsisnot asingletextmighthavepartlycausedsomeinconsistenciesinConfucius’moraloutlook.Heallowssucha possibilityinthecaseofAristotle’sPolitics:“PartlybecausethePoliticsisnotasingleandcoherenttext, Aristotle’sdescriptionofthebestconstitutionisnotalwaysconsistent”(Yu2007:134). 15NorisitconsistentwiththespiritoftheAristotelianmethodofsavingthephenomena.Isaymoreabout thisinthenextsection. 368 XIAOYang (2) It is possible that it makes the best sense to interpret (or organize) certain passages in the Analects in terms of Aristotelian structures, but not passages from other texts of the Four Books. (This possibility is ruled out by [UFB].) It should be emphasized that my point here is a modest one. I am not claiming that these two scenarios are actually the case; I am only saying that these possibilities should not be ruled out a priori before the inquiry begins. I believe Yu could recast (UA)and(UFB)asworkinghypothesesthatwillhavetobeconfirmed,modified,or evenfalsifiedbytestingitagainsttextualevidence.Thenthepossibilitiesmentioned above would not be ruled out a priori; rather they would be determined and tested through arguments based on textual interpretations. They will then become an indispensable part of the critical enterprise of comparative philosophy. I offer more nuanced arguments for this approach in section 4. 3 Yu’s Uses of the Aristotelian Method of Saving the Phenomena In this section, I focus on two of the main methods Yu uses to explain away the contradictions and inconsistencies, as well as varieties and multiplicities, in his interpretationofAristotleandConfucianphilosophers.Ialsoarguethatthemethods might also be used to reconcile scholars’ various interpretations of a text. When one reaches the end of Yu’s book, one will get a coherent picture of both ethical systems, with apparent contradictions explained away, and different interpretations reconciled. How does he do it? The answer is: Aristotle’s method of saving the phenomena. This is Yu’s summary of it: Aristotle’smethodof“savingthephenomena”consistsofthefollowingprocedures: (1) collecting and establishing the phenomena; (2) discussing and analyzing the conflictsofthesephenomenaandthedifficultiestowhichtheygiverise;and(3) savingthetruthcontainedinallreputableopinions(endoxa). (Yu 2007: 5) Thesecondstepturnsouttoincludenotonly“discussingandanalyzing”aporiaibut also“puttinganendtodifficultiesandcontradictions(aporiai)”(EE,1235b14;cited by Yu on 5). Aristotle gives the following description of the result of the procedure: “Such a view [that we arrive at in the end] will be most in harmony with the phenomena; and both the contradictory statements will in the end stand, if what is saidistrueinonesensebutuntrueinanother”(EE,1235b15-7;citedinYu2007:5). ObviouslyAristotleisnotreallytryingtoholdtwogenuinecontradictorystatements atthesametime. Heissayingthatit ispossiblethatastatementcanbe“true inone sensebutuntrueinanother.”Asaresult,thetwoapparentlycontradictorystatements willturnouttobeconsistent.Aristotlehasbasicallytwostrategiestoaccomplishthis goal of reconciling apparently contradictory claims. The first is to assume the ambiguity of the semantic meaning of the terms. The second is to assume the plurality of the content of the concepts.16 16Thereisanambiguityofthemeaningoftheterm“themeaningofA.”Itcouldmeaninanarrowsense just the semantic meaning of A, or it could mean in a broad sense, which includes both the semantic meaningofAandthecontentoftheconceptrepresentedbytheword“A”.Ifweusetheterm“meaning” initsbroadsense,wewouldthensaythatthereisonlyonestrategytosavethephenomena.