the JapanAssociation the Japan Association for forPhUosophy Philosophy Annals ofthe of Science Science of ,Japufi Associiitiollft)rPhiJosoph)r ofSciei]ce 57 Deeply, Contributing Originally Thinking An interview with [l'zamothy Williamson Timothvv WILLIAMsoN' Key (2009)57K87 Vol.18 Chen Bo" and Agnosticism, AnLi-realism, Epistemic Paradox, Philosophy words: ef Language 7L}LeeditoT' this ,iou･T'naZ asked of .t'or 'in Japan. article ,Japanese readeTs Tne to uJT'ite a short fbr'ward ofthis how its originat Engtish version goes back to rrby Oxford days in the summe'r' to ea;pZain It all pmblished met P7ofttssor' crLeTeBo foT' the fi･r'st time in thc fir'st lectuT'c o.f'the hi7ni'nthe gecttLre Tz}om on the tvheT- .Johrt a tectur'etheater Lock;eLectureserties secoTtcl was near inte7n)ieuJ came to be 2008. I St. Cross CoZlege, of delivered.Later,Ifoun[l .fiooT'at 10 Mar'torb wher'e Pr'ofkssoT'Ti7n,othy Hlilliamsonconducted his senLinar' on philosophy of tanguage, C7ben Bo anci I then began to discussvario'us kinds ofprobZems in phigosophyin various pgacesin OuifoT'ct. Sorrietim,es our discussiontook place in his o,tfice iTt. the base･rrbent o,f'the philosophy buigdiny. foculty On svmc other occasions, he invited rne to his house to continue discussionsover the dinner table. When we met and discussedphigosophy, J had occasions to lookthroztgh the drojZ's a'rticle. At first the drcoft ofthe present i･nte7"vie･w was ??,ot so long. [lihen,graduatly, i,tbecame gonger anci cleeper. PVIien the droftwas ･rrLc atmDst co77ipleted, Che･n Bo told that thc C7Linesetrunslatio･nuJas c'urrentZy being prepared, and pToposcd topublish the original English version i･nJa･pan.Through him, o･ur l obtained agreemenl/ I thank both ofthem fro7nPrcLfessor Wtlli･amson con,cerning itspubtication in Japan. fortheirkind offkr. Jt is 7ny pZea,sureto see th,eo7tiginal English puubtishedin thisp'ournal. This paper is a p'ointcollaboration versi,on structured as a guestion 7ihe reader and ansu,reT between tu)o phitosophers.77),eintervi,ew is session. CTLenBo asks guestions and U"itliamsoTe ho?llthe questions are carofully and skit4flLlly aTv'anged, so that the ansuJers sometimes go [leep, and becorTie .f'ar-reaching.Jt takes the foTTrt･ discussions uJhi･ch occ?tTy'ed betuJeen one phitosopheT' in OxfoTd an{l thc ofh,s/pothetical o,nswers. wilt see Jt consists o.f questions and anszc,ers whi･ch were curphilosopheT' in Beby'ing, with the tacit assumption that the readers are pramariZy students, Tesearchers otheT' ried out ' "* Oxibrd Universlt･y Peking IJniversit･y -57- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 58 and Timothy puhitosophers in VIJ' ILLpxMsoN and Chi･na.As a intended to be addressed contemporary Chen resuit, to Bo Vol. 18 there a fe7v phTases and in jmpan. However, are not paTagraphs which I thought it was oppropriate not to modalv any puart of the discu,ssions between the two philosophers, f hope tt will be u･sojhtl and intellectuatly sti7nutating forrea,ders in Japan as welg to read the interoiewarticle in its original form.it wiZZ he(p thern,to are their unders'tanding make ofcurrent readers issues i･nphilosophy wider and deeper. Koji Nakatogafvai TW] has been the Wykeham Professor of WILLIAMSON [hereafter I,ogic at Oxforclsin(te 2000, is a fellowof thc BritishAcaderr]y,a foreignhoiiorary alld etc. His main research meinber ef the AmericanAcadeiiryo{' Artsalld Seiences, interestsa,re iiiphilosophical logic,epistemology, metaphysics and philosophy of lan1990), ii}zgueness guage. He isthe auther of Identitgyand DiscriTrLination(Blac;kwell 1994), KnouJledgeand its LiTTLits 2000), 7"hePhilosophzlofPhiCRoutledge (Oxford ' 2007) and ovcr 120 articles. losophy (Blackwell CHEN BO [hereafter CB], PhD, Professorof Philosophy at･ PekiiigUniversity, August, 2008. China,;academie visitor at Oxford between August, 2007 and T'IMOTHY to CB: ProfessorTimothy Williamson, I'm very glad te ha\e this opportunity interviewyou. AsIknow, you arejust two years older than rne, but you are internationally known as a preeminent philosopher: I myself have few Englishpublications, not This isso,Ithink,, because of my laziiiessand stupidity, ... TW: Certainlynot! CB: . .,but because of a aseries of complicated reasons, Ijust meiitieii itin pass-t to m.y forrnal inter"view go questions, Chinese philosophers a-d logicialls kiiow only a little about you personally txnd works int･roduceyourselfbrieflyto Chinese readso far. Could y.ou first･ your academic ers?･ For example, career, your edueation, your academie your main rcsear ¢ h field$, illg,let me family, some your amateur hobbies, something ]ikethese? T"r: Although I'm Brit,ish, I was born in Sweden (in1955),becaiJsemy parent･s were teaching English there. Both of them went on to teach English Literature at OxfoTd University., so I grew up iiLan academie at･mosphere, I too was a student, at Oxford. My firstdegree was in mathematies and docphilosophy. My subsequent toral thesis was on Karl P.opper's idea of verisirnilitude, according te which seientifie theories can approximat,e better and better to the truth without ever being perfeetly t･rue.After that I taught phi]osoph,y at [Ilrrinity CollegeDublin f'rom1980 to 1988 and i Department of Philosophy, Hokkaido University ･--58 -- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of Thinking Deepl:, Contributiiig Origina]ly, 59 Oxfbrd University from 1988 to 1994. Then I moved to be Prefessoref Logie and Metaphysics at･ Edinburgh Universitybeforereturning to Oxford in my presentpositiou, I've been a visiting at )v･{IT, Princ¢ ton, the Austra]ian professor or researcher Hon.or Kong and elsewhere. If I had NatiorialUnjversity,the Chinese Universityof to pick out one central theme in my research, l'd say: the g'ap bet･weenwhat istruc many related and what can be known, That has led m.e to jnvest･iga,te questions, both technicalones in logic and broad philosophieal questions, such as the nature of knowledge, As forrny non-acadernic life,I'm married to Ana )･Iladenovic, a pianist from Belgrade in SerbiaiI was previousiy married t･oan Italian. I have a daughter of l5 and sons of 11 and 3. My work and family lifekeep me pretty busy, I used often to go wa]king in the mountains, but these days,alas, I rarely find time for that, and net even enough t･ime forthe poetry, iiovels and filmsI love,A,Iyphilosophy, has to visit count･ries in many, part･sof the world aiid to some given me the opportunity extent t,o see them not just･ as a t,ourist but from the inside, through the intcllig,eiit･ at t,hephilosophers liuve been hosts. eyes of also as a CB: Ybu have an aeademic background in mat･hematics alld philosophy, you are in a,n atomic energi.J reseaxch st･ation a logician, a,nd be/Eb/re universj,t/y you worked computer cont･ribution all this background programmer. I'd !iketo know what and training made "rho iity to your phi!osophical career? [I]W: Computcr programming provided useFul t･ra,ining in how to clarify initially vague ideasto the point where the.v can be rn,ade formallyprecise. It also gave me a concrete understanding of what･ was meant/ by models of the mjnd as a comput･er. From mathematics of problems, alld I learntthe importance of thinking about the abstract structure of discerningsuch st]ruct･ure by means of an elegantly suitable con'fra/mework. I learnt,such as probabilit･ytheory, is ceptual Some of t･hemat･hernatics directlyre!cvant to philosophy. Logic was strongly emphasized in my undergraduate bridge between philosoph/y degre¢ bccaus¢ it provides the widest and most natural mathema,tics and philosophy. pt･Iyphilosophical trainiiigwas firmlyiiithe tradition of analytic that there isa fbrmal aspect philosophy. Altogether,iVs not surprising work, even though most of it is iiot purely formal. At to most of my philosophical itssimplest, the fbrmai method involves giving inibrinal arguinents forsomc 'Fbxmally expressed formallyfrom those premises t･oa collclusion, premises, and then arguing These structural ways of thinking are useful in more ways than many philosophers realize. For inst,ance, they.make itmuch easier to firid counterexamples to philosophical t･heories,,hecause one can werk out in advance what･ sorts of struct}ire itmust have. I was lucky in the qualityof my training. Oxford was (audis) a wonder'ful place t,o my tutor for all hour once study I met philosophy. As an undergraduate philosophy a week, either by,myself or with one other student/; I had to write an essay for each on. Similarll,r, one has long individua] supervimeeting, whic:h the tuLer commented -59- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 60 WiLLIAMsoN Timothy sioiis on written one's knows that every werk sentenee as a graduate olle writ･es Chen Bo able VbL18 There isno place te hide: one student. bc must and to withstand scrutiny. Besides my fellow-students weTe very. able, and I learntmueh firomtalking them. I also eonfidence to fbllow through original ideas from philosophy gained working in a world centre of philosoph.v,where many of t,heteachers had international reputatlons ancl were makin.g new contrjbutioiis as it were in l'ro. nt ef my eyes, teachers, many of my with "Xhen I was a student, the leadersof Oxford philosophy and logicincludedMichael Dummett, Peter Str,awson,A.J. Ayer, Dana Scott,Gareth Evans, John )icDoweil, John Mackie, R,M. Hare, Derek Parfitand CrispinVLrright. I thought: I can do that too. CB: Whell you 'Ybu called. in 2005, said programme That stntement was coinmen.ts on studied pursued Oxford, yeu at "The struck exc me, ¢ met, the [`Davidsonic sses of the manner and still strike me, in which as an abuse boom" as it was the Davidsonian of formalization." is quite bold, directand irnpressive. I'd iiketo know your Darvridson's phi]osophy. TVSJ: Donald Davidson made irnpor'tantcontributions to current philosophy,In the 1960s he played a major role in stirnulatillg research of more const･ructive aDd systernatic kinds,and acted a,s a corrective to some aspects of t･heinfluence of ullfortunate Wittgensteinand ordinary language philo$ophy. In the philosophy of mind, Davidsen emphastzed the causal significance of explanatioiis of act･ioii in terms of the agellt's beliefsand desires.In the philosophy of lai!guage,he laid out the programme of truth-condit.ionalsemantics, by which the truth-conditions of all sentences of the Iancontributions of their censtituent guage can be derivedfrornthe seinantic phrases, which requires a systematic approach to the language as a whole. He also made many highlyinfluent･ial contributions on specific problems: for instance, his account of selfdecept･ion a,nd his t･heory of adverbs, according t･owhich most sentences are implicitly talking about events. so that itwas often to quest,ionsor objectiens, he clliptical, On the unclear negativc side, hisclaims what his style or arguments was too were. cryptic and In response dcfensive and guarded,doii)glittle to articulaie his ideasmore explieitly, Uiifbrtunately, some allowed him to people get away with that by treating him liko a guru, [I]hereNN[as also a tendency in the Davidsenian tradition to laydown founal constraints in a dogmatic rather than experim.ental spii'it,, without the justification forthem, and to use these const･raints to dismiss properly explaining legitimate theoretic:al alternatlvcs. It･didn't help t･hatsom,e of the people involved had a rather amateur knowledge of logie lit･tle Iearningisa dangerous thing"). was (L`A i,oofar towards a refusal of the scientifie spirit obstructed progress. Fortunately, that quasi-religiousattitude much less coTnmon now than it once was, tThis all went -- 60 of opei] debate, and tewards Davldson is - NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of Thin]{ing Deeply, Contributing Originally CB: Let 61 1990). (Blackwell go to your firstbook, ldentityand Discrimination Could you out･line the most importaiit ideas prese]t,ed in this book? us TVVi: I was oi' interested in situations identity for objects properties. For example, same species ifand only of can't up you might ifthey call a want･ to can suc ¢ essfully group Z - interbreed with large genet･icdifference.By say in which cases contrast, want･ed criterion successfuIIy are successfully to one kind, but the some doesn'twork, because there with group Y, and group Y where to give a part･tcularcriterion didn't have the right logical that two groups of interbreed with each group X interbreed sevei"a] can with sma]1 identitycan't animals other. successfully are the But that, illt･erbreed greup Z, but group X genetic differencescan add have tliatstructure; it has to be transitive,in the logicalsense: if U i$ idcnt･icalwith V, and V isidentical with W, then it Iogica]lyfo11ovtTs that U is identical with VLi.By set-theoretic reasoning, I showed that there is always a approximation' to the original criterion t･hat, Cbest unlike the of this problem (inthe te original say one, (ioeshave the logicalpruperties required foridentity. One can best thingi. I was part･icularlyinterestedi in instances fa,11 back to that therefore where cognitive sense, that two objects `next the problematic original criterion was one of indiscriminability not in the purely logical soiise), For instance, you might want have t･hesame apparen,t colour if and oiily ifthe subject can't That･doesn'twork because there are some cases in which t･he subject can't sce any difference in colour between A alld B, aiid can't see airy difference in colour between B and C, but cau see a difference in colour between A and C several invisiblc differcnces caii add up to a visible difference. Again, genuine identitycan't have that llon-trallsitive structure, I applied the saine st･ructural approach to those cases, but I also became interested in the nature of indiscriminability itself.I arguecl t･hat the indiscriiriinability. of t･wo t･hings is a matter of the subject not･ being ab]e to know that they are not identicaL I showed hew to form,alize that idea and cxplore itslogicalconsequences rigorously, within the setting of epistemic logic.A!though ldentityand Discrimination is similar to my later books in quality of argument, it doesn'treach tl.Lesort of clear, siinple conclusions they do. As a result, i,thas had lessimpact t,hallthey hEwe, although there has been a recent increaseof interestin itsaccount ef indiscriminability, see any difference between thcm in colour. CB: Leibniz'sprinciples includeboth the indiscriminabilityof identicalsand the `indiscriminability' ident･it,y of indiscriminables (where is understood in the legical sellse); in syinbels tos,ether: VmVy(x It seems to mathematics, his childhood, me = y e that these principles only and are youth, VF(Fm e to apply inapplicableto cross-tcmporal middle age, and old age: and Fy)) abstract object･s, are also objects like numbers in person across inapp}icable to physical such as a -- 61 -- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 62 Timothy objeets which exist X･VILLmMsoN Chen Bo XVhat do you think . cross-spatially. and the identityof t･hoseobjects in space to think that about; this? How to determine time? and TVSJ': The indiscrirninability of identieals applies tlon. It･is tempting child, the very same Vel. 18 to the old man although all objects without is ident･icalwith excep- the young' human being,thcy arc discrimjnablebecause the old rnall has IIowever,such apparent preperties that the young boy lacks, such as being wrinkled. The old man dissolveonce one isaccurate about･ time references. counter-exainples but he is identicalwith the person who was a y.oung boy; the old mall is wrinkled when he was a young boy, It js fa].se that the youiig boy st･ill is wasn't wrinkled unwrinkled, DaVid Wiggins has given an excellcnt dcfence of t･he indiscriminability these 1ines. The spatial case is similar. One and the same river, of identicals alollg the Thames, ruiis t･hroiigh0xforcland then L･onden;itiswide iiiLondon but not in Oxford. Thus different,parts uf the same river have incompatible properties -- the through London iswide, the st･retch of the Thames stretc;h of the Thames that･runs but that does not imply a count,erexample that ruiis through Oxford isnot wide The identity of indiscrirninablesalso applies to to the indiscriminabilityof identic;als. all objects without exception, provided t･hatthe preperty F cari be iion-qualitative, which applies for example the p,roperty of being one ancl t,hesame person as II/bd, to to Ted but iiot t･o his twin brother Fred. Leibniz himselfintended a restriction of two exactly sirnilar whic:h apply to both or neither purely qualitat･iveproperties, objects, but I assume t･hat y.uu havc in mincl the more liberalmodern iiiterpret･ation the modern interpret･ation, both are of the identity of indiscriminables, On principles iogical trut･hs. Of course, logicby itselfcan't t･ellus whether object･s we encounter identical;we need tu know something about in different times places at differenL the qre too! objects CB: Ip your view rance: non-heap, strange it? Why there really hut we and everi and T"": The straightfoi"ward. book, 1,ligw,eness 1994), you developan epistemic (Routledge on which vagueness is an epistemic phenom ¢ non, a kind of ignois a specific grain of salld whose reinova,1 turns thc heap into a secoi)d 'vagueness, of a cannot know which lit,tle bitwild, arguihent It,cembines it is. It Collldyou how do you develep centra] enc such clarify a, concept,ion fOr epistemicism standard souiLds quite counter-intnitive, your position and fer arguinents of vagueness? about vagueness priiiciples of classical is very logic and simple semairkics and with that･almost everyone accepts. For instance, 10,OOO grains make a heap but O graiiismake a heap, It fo11ows by quite ordillary logicalor mathis a natural llumber n, between 10,OOO and einatical rcasoni,ng that t,here soiiiewhere O, such t･hatn+1 gra･indoes not Tnake a 1]eap but ri grains do not make a hea,p.By a standard t,ruth,w･e ean acld that it is true to say of n+1 princ:iple about common seTise observations `That, 62 - NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of T'hinkingDeeply,ContributingOriginally grains many it is obvious that t･herelevant･ vague niake a heap and by heapT and iri,ake a a numeral of n tThat grains do many 63 not ma,ke a heap'. But That is, ' in idea how to findeut which number js n. sense of w･e arc in no position to know that n+]. grains n graiiis do not make a heap, where the number n is ident･ified have we no `heap:, such as Cunsequerit,ly, the `29'. extensioll of Lheap' has a boimdary that bozindaryis, This solves t･heancient position to know where sorites paradox: it appears that subt,racting a justone grain from a heap stil], ],eaves heapl therefore if you subtract grains one by one from a heap of 10,OOO grains there is st,ill a heap vLihen no gra,ills are Ieit. The principle i/ha,tsubtracting a grain from but we are no leavesa heap isfalse,but,we cannot know exactly to be a heap. Of course, many dislikethat eoncliision, philosophers it by ehanging classical or somehow principles of logicor msemantics application, That uilwillingness to apply. basicprinciples as soon a heap in alway, s encount･ered strikes the suggested ine inethodologically alternatives non-classica]. includinganalogues to the as as i""ie"r of the a result of I unsound. introduce all sorts also of there when ceased try to and restrict･iiig their diMcukies are as in the book that difficulties of their own, epistemicisrn. I came show their proponeiits use against Identity an(l Disc･r'iTrLination. writ･ing In explaining ones airoid the discriininabilit.v, I bad forrnula,ted what I call thc for error which says roughly t･hatin erder to know in a givell situatien princip],e', you must avoid faJseb ¢ liefin relevantly, similar situations, otherwise in the original your belieL' sit･uatioll is too unsafe to constitute knowledge. I reaiized tha,t the margin forerror how, also explain have sharp boundaries, principle would given that vague concepts we still cannot know wh,ere they arc, Such an explanation makes epistemicism a mueh mere attractive theoretica] optiell. Those ideascame to mc latein Lhe writing of Jdentityand Discrimination. My conception of vagueness was st･il] evolviiig as I I did r]ot arrive at a full.v･ wroi,e, and unt･il shortly after gencral t･heoryof epistemicisrn noll-t,rans/itivit･y it was `inargin of pubiished. CBi "F'hat･ isthe problem fuzzy logic and supervaluationism vagueness classical What diMcultiesdo many-valued logic, haye to fac:ewhen they arc used to deal with dees paradox (the paradox of the heap)? VLrhatadvant,ages and the sorites logic have whcn it is used T"J: The vague of vagueness? words central and problein c:oncepts, includingscientific wor'ds inelude concepts; thein? ishow logicand of vagueness which and to deal with virtually although semantics should all eur words and t･hey are ofteii inore treat concepts precise than perfectly precise. Does elassieal logicur need or restricted in some way for vaguelless? lf so, how? semantics to be modified for lii]guist･s Ifnot, why not? The problem arises not, enly fbr philosophers but･also and for computer scientists-who sNTant to buildcomput,ers that can communicate with thrce of the most･ popular human beings in vague language, Ybu have mentioncd every, da"･'words and cencept･s, they. are not 63 - NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 64 1]imothy WTLLIAMsoN noii-classical Many-valued approaches. inst･ance,they imply that ceunted a Vol. 18 fuz7y logicschange alld contradictions Chen Bo it says that heap:can be true a･iiexistential claim such as classical is a [There t･rueinstance -- having a which it is true without logie,Fbr they should be elassical be half-truc, whereas can definitely false.Supervaluationismcha,nges as stallce, and semantics, For in- point for being cut-off to supervaluation- according is a, cut-off point for being a heap', Classicallogicautoma,tically avoids A major these diMculties. difi,culty for most non-classical approaches ishigher-ordervagueness: bordeirline cases of borderline cases and so on. I argue in the book that they do not･ treat it in a satisfact･ory way. By contrast, epistemicism can easily handle higher-ordervagucness, which it interprets as limits on our knowledge about limits on our knowledge. The margin for error prillciplepredicts exactly such higher-order Iimitsen knowledge. C}assieallogic ists there is ne is standardly, point used of in logic, mathemat,ics Philosopherswbe suceessfuI, reject CThat to say it bas proved immensely the diMcultiesthat･such a rcvi- science, and it raTely where confront fbr science. In Englishthere isa saying "A bad workman blames his tools". Classlca] logic and semantics are amongst the philosophical workman's iools. sien would make CB: As Dumunett stat･e of the world that bivalencepresupposes says, makes some sentences some true and kind oth ¢ rs of realism: it is the false. But the intu- itionists and Dummett claim that truth issome kind of epistemic coricept concerned w'ith provability or knowability; it is ollly when we are able (at least in pr'inciple) to know t･he truth-values of a sentence that we can say it is either true or l'alse, prove gr about such claims? What is your eomment t･hattruth is any kind of epistemic collcept. There is no cont,radictiorL whatseever in the hypothcsis that there are truths about our world that nobody eould ever be in a pesit･iento know. The fimdamental nature of truth and falsity is eaptured in Aristotle's principle that to say of what isthat itis T"i: I cornpletely thc cla,im reject of what･ is not that it is,is fa,lse, while to say of what･ is that it is,or ot' what is llot that it is not, ist,rue. For instance,the proposition that itisraining istrue if and only ifitisrailling. That equivalence isstated iriterms none of which scems to be epistemic. The burden of proof lson thDse who claiiii that truth is an epistemic not･, or coiicept: iiisome alLer sense. all, on ･Wheii their own ene view, analyses if their view attempt･s is t/ruethen it should be provable for am epistemic conception Iiituitienistic mathema,tics, developed by t,o argue truth, they turn out te be ver}r･ weak. Brrouwer as an alternative to classical mathemaLies on the basis of' his rejection of bivalence,has remained utterly inarginal withiii mathematics. Attempts to extelld tihe intuitionisticprogrammc to the nat･ural sciences have had no serious success. of CB: It is said that yeur epistemic view picion by maiiy people, but has become a of vagiieness relativcly was mainstream at firstviewcd view, at vgTith sus- lca,stquite --64 - NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of Decply, ContributingOriginally 'lihinking infiuentialolle By the waty', what in an on research standards or vagueness criteria paradox. Is that true? the sorites and shou]d 65 we Ibllowto evaluate phi],osophical a work?･ T"i: Before I.published iiZLyueness, epistcmicism was as too generally regarded crazy to be takcn seriously, even theugh a fow people had defended lessdeveloped versions of it. My initial hope for the book was that･itwould forcepeople to take the vicw seriously. It has certainly had t･hat cffect: episteinieism is now normally list･ed as one of the main theories of vagueness. In view of itssimple, classical nature, theory of vaguencss, yoimger philosophers regard it as the de/Fault be rejected oiily, on the basisof strong arguments. Illgeneral terins,the bc foIlowedin evaluating a philosophical work or not so differentfrom the many iiisome one to to standards evaluatiiig a Are therevalid arguments foritfroinpremises that we can indepenclent]y evaluate as true? Are there val,id arguments from it te con ¢ lusions '"Jhat that we can independer]tl,yevaluate as false? cxplanat･ory power does it have? Does ithave the in'Lrinsie virtues of simpiicity and elegance? work other science, CB: I'd also likete know why metaphor porary philosophy oE' language.Could you TP"i: Metaphor and porary sernantics the meanin.v. constituellt cxplanation encoulltered of a censtituLes phi,losophy o-anguage, linguistic expressien humans: iiitalking ability such a hoi/t,opi,c in eontern- that t･o me? tntriguing challenge to understand befbre (fbT instance, in reading of explaiii t･oa compositionality, central idea in contem- according is builtout coinplex (asI mentioned words aii has become about new of the meanings Davidson). '[I]hatis the sentepces which they out of standard have never. this interview), provided tha,tthcy are words, Metaphor doesn'twork previously encountered IfI tellyou that･ aii actor is a pea,cock,you can prebably work out wbat mind, evcii though itisisn'tsimply the result ot' combining t,hemeanings eonstructed t･owhich of its like t,ha,t. I have in my words had. Some people have postulateda special kiiidof iiietaphorical meaning to explain the phenomenon. Perhaps the most importa,iitsiiigle work on the t･opic is an article by Donald Davidson, whom you were asking about before,in whjch he argues that there isno such thing as metaphorical ineaning; there are justt･heliteral rnea-ings of the expressions, and the effects they ¢ ausc in hearers,Although hisview may be too simple, it provides a beautifu11}, simple and powerfu] basis for further discussien ef the topic, already CB: In 1997, you were e]ected as fellowof the BritishAcademy. I guess rnany Chinese readers arc TJot c].ear what kind of academic institut･ionthe British Academy is. XVha,t difference is there bet,ween the Royal Society Emd it? Could y,ou explain somethillg te them? -65- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of '1'irnothy "･'II,LmMsoNand 66 Chen Bo Vb].18 The Royal Societywas fbunded iitLondon in 1660 as a seientific society alld for t･henatural sciences and mathematics. became t･hena,t･ienal academy tt in Brit/ain Since it did llot cover the humaiiitiesand social sciences, the BritishAcademy was Philosophy, including for those axeas, founded in 1902 to be the national academy, philosophieal logic,fallswithin the domain of the British Academy rather than the a feIlowof the Royal Society.On ar"erage one British philosopher a year is elected BritishAeademy. It is not a job,justaii hoiiour.The BritishAcademy distributes represent･s the humanitiesaiid social sc]iences in llavarious serts of research a,cademies in ot･her on educat･i.on and research, cooperates with national tienal dcbat/･es 'IiW: .ffrant, countries, and se on. ' Professor of Logjc, Uiiiversity of OxCB: In 2000, you were appointed Wykehain is irriportant. XVho are your predecessoys, could you ford.I think this positioll quite briefiy. iiiti'oduce them te Cltinesereaders? What duties do you haMe on this position? who is best kllown for his My, immedlat･e predecessor was Dayid "J'iggins, and un ethies. Before "]iggills, the work on the logicand metaplrysics of iident･ity his famous books chair was hcld by Michael Dummctt, the defenderof anti-realism; Philosophy ofMathematics, 7}'?tth and include l7b'ege: PhiZosophy ofl]anyuage,Fleeger TW: and The Seas of Language, AlOtheT Enigm,as, 1ihe Logicai Basis ofAcletaphysics are much closer iny philosophical views though Dummett was one of riry supervisors, to those of VLriggills, but they, are both important figures,Dummet,t's predecessor was A,J,Aycr,whose book Language, 7V'uthand Logic played a major role in introducing The teaching dutiesof the chair mainly concern logicalpesitivisrrit･oBrit･aill. gradwit･h me; I try, to teach their doctoraldisserta･tions uate students. A･Iany are writing of rigour aiid precision by.criticizing what them how to work to the higheststandards they write Iilleby line and discussing with thern the best ways to developtheir ideas, 1 am very, proud of my st･udents, who have gone on to t･each philosoplty theinselves in rnany parts of the world. I also give classes, in which I sometiines present･my own well as the students recent research. The point of such classes isfor the professor as to learn from t･hediscussion.A,Iostof my artic].es ancl books were presented at olle tiine or another in graduate classes aild benefitedfrem that process.Akhough iiry, to plopr a, leadership pi'ofessorshipis not･ tm administrative position,I am expected but･ b}.rservillg on committ/ees role in the Philosophy Faculty,not justby example amd so on. Of eourse there are ot･her established ehairs of philosophy in Oxford too, Davics ls Professor CurrentlyJohn Broeme is T'refessorof Moral Philosophy, rvrart･in '[I]erence is of ]Vletaphysics and Irwin of Mental Philosophy, Hawtborne Professor is ProfessoT of the History of PhilosophLv. .lohn CB: Also in 2000, you published yoi.irthird book, KnoiL,ledge and itsLin'bits highly evaluat,ed It seenis that/ 1,his book has been warm].y wdcomed and (Oxford), . -66- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of Thinking by English philosophieal to epist,emology circles. out come ContributingOrt.ginally Deeply, since I read 1975". such [[It comrnents sets t,heagcnda decade ancl beyond." IX,1'v. question is the $imilar. t,hernost important ideas preser]ted iu this beok?/ next T"[: 'The of knowledge. slogan of the book is as a starting-poinL on as 67 it is "t,he best book in for epistenhology forthe Could you as abovc: first'. Itadvocates '`Knowlcdge the basis of which to summarize t,akillg t,he notion undorstand ot･her rather than vjce versa. Knowledge is a mental phenomena,, such as belief, its own right. Be]iefis a more state that･ aspires to the generalmental cognitive in state condition knowledge. Successfulbeliefconstitutes knowledgel belieft/hat･doesn't coiist･it･ut･e 'The knowle{ige is defective. llotjon of knowledge can't be analysed in inore basie terms, It is also used to explain the nat,uxc of justification, evidcnce and assertion, The view is a radical form of externalism, because it implies that･our mest basiccegnitive states are constituted by relatioiis to oiir external environmont. Fof'instance, the cat sat･ oii the mat only ifthe eat really did sit oii the mat .. ., yuu kitow t,hat, of CB: As far as I know, there isa tradition rooted-deeply. to which knowled.creimplics truth. Quincadvocates "the tic t,heory about web of belief,i', and a]so insistsoii a he even dees not accept knowledge as a legitimate concept, according What is your opillion t･herebetween your conceptien edge. TXNJ:Surel>,. the one claims to know common that there is iio c]aiming true beliefwarrantcdto vary. ing degrees,Although t･hingas kno"Tledge, but only I wil,1 not･ go that far,I have some such that knowledge implies truth, from Plato'sTheatetu,s an eiripirical and holisquiteradica,1 fallibilism; sympat,hy "Ji'th him. I strongly is against the commoii-sense about/ t･histopic? "]hat similariti/gs of knowledge and Quine's? which sense view suspect･ the thesis collcept･ of and kiiow].- (iiffbreiices are isthat knowledge does eutail trut･h.Ifsome- thaL Beijing is in Japan, his c;lai,rn is justfalse. He may t,hillk that･he knows that,]]eijing is in,Japaii,but he iswroug; he doesn'treally, know that Beijillgis in Japan, for Beijillgis rtot in Japan, He i's igriora,ntof his own ignorance as he just, is ignorantof geography. I don't think Quinewould disagreewith that. Even radical fallibilists can agree tha,tknowledge enLai]s t･ruth,since we are fallible about whet･her we know. Even ext,reme skept,ics agree that knowledge'enLa,i]s truth, and often nssume as much in t･heir argmnentsi suc/h skeptics are radical fallibilist･s. Actually,it is quite dificultt･ogiv･ean accura,te, precise defi,nition of but itisat least clear that the clairn that knowledge ent･ails trut･hisperfectly coinpatible `fallibilism', the claiin that there is no ge'neraltype of infallible.I acc;ept both claims. If yon are happy with subject with degrees,tl'teiiy,ou could define t,oa high degree,and knowledge in i,hatsense would ranted view t･ovarying I defend in Knowledge the concept･ about of wliich we are true beliefwar- t･ruebel{efwarranted autematically ent,ail truth. The itsLirllits is t･hat tlie concept of knowledge is at leasi/･ `knowledge' aTid mat･ter as - 67 - NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 68 Timothy basic as the concepts knowledge was based on as of WmT,TAMsoN belief and Chen Bo and distastefbr the Quine's warrant. uiiwarranted Vbl.18 reductionist opposing basic than knowledge other factors.I'm also tries to and them? against that takes beliefas epistemology knowledge in terms analyse in this book? /Eight rnore belief,truth of and doct･rinesof internalism, which give te internal stat･es of tbe subject as the st･arting-poiiit primacy purely. Another target is skepticism, the arguments for whieh presuppose episternological of epistemology. some long tradition in a of assumptions, CB: "[hat are the oppostte doctrinesxsrith whom you illtendto How do you object･ them? Could you outlinc >rour main arguments T"･i: I'm concept arguing against relat･ed internalism even when they appear not to. One argument･ against tbe to analyse knowleclge in tcrms of bcliefisthat Iit,erally hundreds of attempts sort atternpt of have been far alld all of them have fai]edl either there ar:e counterexamplcs it The view I defend explains why to the analysis or is circular, all such attempts must fail.I also argue that thc motivation foriiiterna]ismhas becn the assumptiou that there isa core of ment･al states t･hatare luminous in the sense that whenever one made so is in. For instance, internaliststend t･othink that･thc state of being in pain isluminous,in that whenever one isin pain, one is in aposition to know that one is in paiu Iargue that the only. Iuminous states is ii]. them, arc oue is in trivial,in the non-trivial state, to know ghat aposition isalways is in paiiiat that beeause one scnse one being in pain isa non-Iuminous state: to know that one is in pa,in. Thus the CB: The argument agaillst . Could yeu your epistemology. thaL epistemology cannot [I]NV:The anti-lurninesity lead you iiitoor one one in them. For instance,pain isa or never times some importan't functiooin forus? ",'hy do you claim to haye an this argumerit again be fullyoperationalized? argumeiit Could you considers a forthat briefi,y? ar.crue process by which slow the mental st･ate in questioll. For inst･ance, an comfortab!e. I gradually.diminishes unt,il you are coinpletely out o/F pain very of the margin forerror principle to show that if the state is not luminous; for instancc, at near the botmdary between being iiipain such a seme er not retajriing obeys one aecepts by that I epistemology, a]']d re,iccting t',herule one epiastemicism mea,n are use a versioii not about operational is in a poslt/ionto know t,hat one -68 excruciating being in pain, there must come a in a position te know that you are in pain. and vagueness. turi]irig it into the evaluation beliefisthat challges gradual transition can occur bitly int/ermediate stage in the precess, point at which you are in pain but are not Thus the argurnents of Knowtedge and itsLimits are closely 'l/?zguaness. og.v However, I show of t･hatone can accept, the whether 1]herefore Tiot･ at･ others. be in pain without･ being in a position foriiiternalismis unsound. ca,n inotivatioii luminesityseems clarify and in the isobeyirig to the epistemol- related anti-luminosity As for of sciise argument `operationalixirig' for rules that acquiring, whenever iL.Of coursc olle that "[ould - NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of Thinki]]g Deep]y, Contributing Origiiia]ly make the obeyiiig a,nti-luminosity of the sort be can't a luminous argument, that state, it is not a trivialstate, But, by the are ]uminous. Thus there are no st･at･es requires. Conscquently,epistemo]ogy and trivialstates only operat･ionalized episteinology operationalized. CB: Ybu bols,E kind of rule 69 =: identit'}i a person:s total evidence K. Such evidcnce a concept of evidence still objectix,e TNV: 1 regard E K = or seerns subjective? as a very natural, wiLh the person knows, tn sym- what to be quite counter-intuit･ive. Could it still serve us t,oreach common sense view Is this trut･h? It does of evidence. independent of your persenal situation, for what yeur evidence you know -depeiidson your personal situat･ion forinstance, on what events you happened to turn y.our head and noticc yesterday, But equally it,does not iiiake evidence purely ± subjective, forsince knowledge requires truth, E K ensures that all evidence conmake not (although all Thus evidellce eaii serve us t･o truths are esiidence). reach truths. The probabilit･y. of truths we don't know can be evaluated on t･hebasis of the truths we do kno"r.Of course we are not always in a position to know whether 'al,wai}Js something const･itutcs in a position to part of our ev･idencc, since we are not kllow whether we know it,but the anti-luminosity argument shows that that is not sists an oE t,ruths net to E objection = K, since whatever evidence is,the state of having a specified wil] not･ be luminous. Therefore,one can haye pr:opositionas part of one's evidence that propositioll as part of o'ne's evidence without being in a positioii t,o kncfiN that itispart of one's evidcnce. VLJemust simpl}r learnto livewit･h that fact. CB: Does conceptien new of knowledge have logieV.Wliat is y.our coniments log'ic?Could you predict tlieprospccts of velopment of epistcmic epistemic TW: of your On the new conception epist･emic logic known as of `positive kiiowledge, we implications for the de- any tb,ccurreiit･ logic? about epistemic should introspection' and reject Lnegative situat･ion of two of the axioms introspection'.ex- in drast･ically, idealizedsit･uatiolls. Positiveintrospect･ion says that ifyou know semethin.ff, you kllow that you know it. Negative introspection says that ifyou don't know tiomething, know it.The fouiiderof formal epistemic },/euknow that you don't･ legic,Jaakko Hintikka, already gave decisive reasons ibr rejecting negative introspectioll,but he hold on to positive intrespection, I argue iiiKno･wlectyeand its L'ivLits that pesitlve iiitrospectioufails, by a special case of the anti-luminosity argumeiit. Much of the i'ecent development,of epistemic logichas been in cornputer science and knowledge,where theoretical econoTnics, both of which are concerned with eommon knows that everybody, kllows it,and so everybody knows something, and everybedy, on. Epistemic ]ogicturns out to be a good setting fbrthe study of comm, on knowlcept edge. [I]'he discussion there tends to be teclinically ]'La'ive. I'm trying t,oencourage epistemologists to but philosophiea].ly sophisLicated make more use of epist･emic logjc -69- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 70 Timothv X･V'IT,T,IAMsoN and Chan Bo 1loL,18 The techniciallshave much to learii problems in which they are interested. from philosophers and vice versa, but it,is hard to overcome the barriers to commurtication t･hatari$e from differences in t･rainingand goals. to model CB: Ybu are that the gap between trut･h and knowabi!iLyis pervasivelthci'e truths. Could you sum,merize forthem here? Could your arguments about the paradox of the surprise and cxamination the your vicws claim unknowable you also outline paradox knowability, ? of rl-VLJ' Epjstemicisrn : about･ t,heaiiti-luminosity ueness, vas, argumellt the rejecbe known, and tien of posj,tivcand negative introspectionall involve truths that cannot at least in the relevant circumstances. ",Iy treatrrientof the surprise exarnination is similar, An example isa teacher who infbrms her pupils t･hat on justone paradox day in the }rearthe.v wi,]]be given an e.xaminat･ion, and that,t,heywill did kiiow oii the morning of the exainination that it will be that day. The argumcrit, ist･hatit camiot be on t]ie Iast day, sinee t･hat morning they will know that it is the only day. Iefl, fbr thc Thcreforethey can rule out an examination on the lastday in advance. [f]hereforet,he day before the last is the last possibility.But･ t･hen they can ,ruleout an examination on that da:}r too, by an analogous axgument. By conCinuing in advance the ar.ffument for each day of the year, the pupils can that there eannot But clearly, ifthe teacheris trustworthy the pupils can be such an argument. know in advaiice t･hat･there will be a surprise examination. [Vhe solution is to see t･hateven if one knows something t,oday, it･ does not, fbllowthat one knows toda}rthat one will still know ittomorrow. [[ihis issimpiy to reject a diachronicversion of thc positiveintrespection principle. In Knowledge and ･itsLimit,s I also defend against recent criticisms a very differellt sort of argument. for the existei]ee of unknoxyable examination. tprove' The devised by the logician Alonzo Church. It goes "ke this. There are uuknown trut･hs.Fer instance, either t･he number of books in my oMce on 1,stJalluary 2{}08 was edd or it,was even. Since I did not count･ thcn, and too much 'which it was. has changcd sillce,, nobody will ever know Thus either it is all always truths. argument was unknown truth t･hat the riumber that the number ways a,re unknowll, known one unknown books ef them by counting t･rut･hthat the rnimbeJ' alway. always after odd al] unknown know always sincc unknown that the number an knowledge requires (since odd alwa>･.s or caTL allew itis an always unkno"rn truth t･hatthose truths, alt･hough al- 1,stJanuary 2008 en the books in of books was that the nnmber that it was was We t,ruth t･hat the s unknown knew t,hatitxnras an would thercby know books evell. unknowable, noa/ of that it is aii be of wa,s my･ However, if it is all then it is an unknowable of books was truth that the number of of books books ullknown t･ruLh;the wholc was odd, which case trut･ht･hat the number odd, argument uses was 7'eductio t,ruth odd, they it would case iiiwhich was always fbrif any. body odd, beoks of in have could room. odd, number semebody they of ad would books not not was abs?srd?tm). ･--7C) NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of Thinking Deeply, Coiitributing Originally Simi,larly, ifitis an always trut,ht･hatthe unlcnowii number of 71 books vLras everi, t,hen it is an unknowable truth that it is an alwa}･rs uiiknown truth that the llumber of books was even. [I]hus,eitherway, there is an miknowable trTith,Anti-realists often call the argument the paradox of knowabj]ity, because thcy don't likethe conclusion. In m;vr view it, isnot a paradox, justa surprisingly, neat argument from true premises to a true conclusion, CB: I'm that your epistemicism about vagueness and your claiin that there are unknowable truths will imply semc kinds of agnosticism, which means that we are imable i,nprineiple t･o know some parts or some aspects of the world, c,g,, in bor'What derlille cases. de you think? Is agnosticism acceptable or not in episternology? If vcs, how to draw out the boulldarvbetween the knowablc and t,heunkiiowable? afraid -tt o My TW: cept we t･hatthere can know - priiiciples I see do views are entail truths some even ]imit･ ¢ d a about te iio objectioii such In some very clear introspect･ionjustmeans cisrn, can't cases sorne us and we cases when we kllow,we don't know, Arguments unknowabi,lity, ac- truths mally epistemologica} know that "re know; ever know t･hat･ we know. Similarly, that･ when we don't know, we can't can't that always when we don't know, we can't I have been explaining give the locatioiiof the boundary bet"Jecii knowability show t,ha,t we cannot have coinplete knowledge of about but thoy its location. That7s what also must provided that it does not turn into skept,iknow that we kno"i things. The failureol' positive we that knowledge considerable we agnosticism, mean that when we know, we can:t the failurcof ne.crative !ntrospection justmeans always know that we don't know; it does not mea,n know that which the possibilityuf knowledge in others, and it does rLot ever to aceording know. Howes,er, there are there is a god. Thc vcry same we whether ignorance in explain agnosticism, also of the sorts lifeisIike, CB: In 2007, Lyou were elect･ed as foreignhonorary member of the Amerieaii Academy, of Artsand Sciences,Congratulat･ions! I'dliketo kllow,how many members from philosophy are in that academy? Take some eminent, philosophcrs forexample? TW: The AAAS tion, although combines philosophers seem Philosophy and ReligiousStudiesinto a singl ¢ secto be in the ma.jority, . The section has almost 150 probably include all the inost distiiiguishedsenior American members, who would American philosophers. Tbe sect･ion has 28 foreignhoncn'arymembers ft'omeutside t･heUSA, as a separate category, of whom a quarter are Oxford philosophers.The other six are Michael Dummett, Anthony Kenny, Leszek Kolakowsld, Dcrek Parfit, David Peanrsand David Wiggins, CB: Recent].y I read y,our new 2e07),Ithink it is quite important, book original The Philosophy and also o,1'Phitosophy provocative. ButIalso (BIackwell find it is -- 71 NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 72 Tiinothvv WILLIAMsoN aiid Cheii Bo Vol. 18 bccause itstopies and ideasa,rc quit,encw, Could you summarize quitecomplica,ted. importa]t･ ideas presented in this book? not to so easy most read and TVLJ': Olle inain chair understand, some quitetechiiical, its arguments chapters method likethose of tho book is to defend the lcgitirn acy philosophical of of aiin the natural reasoning the traditionalarm- ph!losophers don't do most but scicnces of deingso by without the experiments the subjeet restricting I regard philosophy as part of our philosophy to our own words or concepts, overall atteTnpt to llndcrstand the world in gencral and our place in itin particulai. Thus the question arises: how can armchair met･hods give us knowledge of the world is encouraging, outside our armchairs?･ The example of mathematics because ituses matter of knowledge of the world. I argue that the armchair methods of philosophy are developinents eut of quite ordinary methods of of the werld. For instance, the thought experirnents that give tts gainingknowled.ffe armc;hair knowledge special but still givesus methods about cases and possibilit,y ability to evaluate metaphysical ofour ordinaJry metaphysieal uecessity counterfactual arc just really (such conditionals a$ tl'C' to the party, itwould have been a feLilure;) by,a disciplined use of the irnagination. It is wroiig to thiiik of thc imaginat･ion as primarily a rcalm of role. log,ie capriciious fantasy. Naturally, also plaiysan ; itp}ays a significant cogni.tive ten people oiily important role had valuable come in armchair In defendingarmchair reasening, that the resu].ts of scicntific experiments are nQver relcvant I don't mean methods, te philosophical questions, justthat in principle does not undermine the use uf arrnchair meththeirrelevance ods too. After all, even though the resuks of experiments are in principle re]evant･ to mathematical methodology is still usually t･he most questionstoo, an armchair eMcient way of dec:iding a inathemat･ic:al question. For seme philosophiqal questioiis -- fbr instance, abeut time - experimental perception or the phllosophy of space results are important, For others - for iristance,about - they do not help. between possibilityand existence relation the philosophy of and the turn,' are past and CB: Ybn assert that t･he turn' and the should be leftbehind,because they are wrong, Could y･ou explain these claims more '"Jhy c:learly?･ For example, what is the linguistic turll? Wha,t is the concept･ual turll? `Iinguistic are of, t,hey wrong? For wliat, `conceptual Especially. , coul(l reasons? you explain the rnore aboutness language and theuglit･? TW: The linguisticturn tury phllosopl]y.It wa,s the by,which languagebecame was one ofthc in both movemeiit, the central most important features ef twentieth analytic largcly of applying the competence one has structure -- the as itssubject of our cen- traditions, non-analytic fbcusof philosophy,eithcr -- the idea that philosephy studies the llature and or of artificial formal languages --or as itsmethod consists and actual mat･ter language idea that philosophizing as a speaker of such a languas.e. -72- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of Thinking Deeply, Conti'lbuting Originally Philosopherssuch as Ludwig Witt･g'enstein, Rudolf Carnap 'tutediverse t･urn on the analytic paradigms of the lillguistic regarded as subst･itutes an exainple the on non-analyt･ic thought for language and guisti¢ turn, and of thought over the turn is similar in the central conceptual concepts who much as the priority of the advocat･ed language in philosophy of analysis J.L. Austineonstiside; Derrida might be and forwerds the conceptual turn as mental P,eter Strawson represeRted these likeGareth Evans In his later work, The side. 73 but role. the linanalysis further, Both the linguis- went･ be interpretedas defencesof the armchair methodolos.y of philosophyby restricting thc ambitions of philosophyto what･ we still have in the armchair: our ewn language and tbought, The linguistic turn alld t,he conceptual turn are elosely related because a very pertillellt foatureof both words and concepts is their illtentionality, their aboutness, in t･hesense of their c:apacity to refer. For t･icand turns cor]eeptual instance, both the One of `water' for writing motivation inuch word can and the book t,othe substance water. that it has become increasingly clear that the concept was water philosophy doesn'tfiteither :space' primarily study the words t･urn. For instanee,philosophers `tiine' and or the concepts space analytic recent physic:sdo iiot refer themi they primarlly,study space and t･imethemselves, Nor do t,hcy priinarily rely on their eompetencc simply, as 'They speakers of a language or possessors of the concepts in their methodology, rely the results of thought I on both experiments and of physicists'experimeiit･s. argue and tlme, or eveii what･ words or concepts shoulcl through detailedcase-study that t]'iere isa concepts even in parts of philosophy where replacc similar shift of fbcusaway from words and do not, play sucl] a physicalcxperiinent･s role. CB: 1'dliketo know what herit･age Lhe last･ century Anglo-American philosophy, left/ us. XVhat advantages and disadvantageshas aiialytic: phi]osophy hadteVLrhatcan we learnfiromit,whet･her good or bad? XNrouldyou still likeLo cal] yourself an analytjc philosopher, sin ¢ e you are a philosopher and at thc same time a ]ogician? T"i: I'm happy Lo call rx'i:y'selfan analytie philosopher,,be ¢ ause lt is obviDus that I beloiigto t･hattraditioii.But･I den't･regard the labelas very iiifbrmative,since t,here is se rnuch both counL other. Of as variety analytic course, the tradition, For iiistanee,"[ittgeiisteinand Quine firom each philesophers, even though thcy are very Ctiffbrent, within not･ all analyLic American philosophers philosophers are Anglo-American and not all Anglo- Severalgiants of early analytic philosophy, such as Gottlob FTege and Carnap, were Geririall.Some Anglo-Ainer'ican philesophers are not tee many). One post-modernists rather than analytic philosophers (fortunatc]y,, of the inost iinportEmt,developinents jn recellt/ decades has beeiithe rapid gi:owth of are anal}･'tic. Eng!ish-spcaking philosophers in China. I It seeins to mc t･hatt･hereare t･ensionbetween being a, phi]osophilosophy have been logicians. analytic philosophy outside countries, now see ph,er many and analytic being a logieian,Many leadersof no analytic 73 NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 74 s･tseN 1]tmothy VgJiLLrix Chen and Bo Vbl. 18 Their logicis just･ the morc Nrege,Russell,Quineand Kripke are obvious exalliples. formal part of their philesophy. At one time rriaiiy analytic philosophers regarded the of analyt,ic As I've said, I regard linguistic turn as the eentra,1 achievement philosophy. it as a mistake. Nevertheless, when one looksback over the past century, one sees that analytic philosophy over that period has lefta huge positive heritage,and not one that･requires us to scaie down the arnbitioiis of phllosophy in the wa}r the linguistic turn the and but･I'11single properly, rize First,the turn did, T.he achievement conceptual refinement of la,nguage with ordinary fu11ypresented spirit Second, t･heapp!ieation and to sys.tematically the evaluate clearly validity few of itsmost that of the semanties semantic articulated counter-exampleas encourages the the stylc sense of and to varied surnmaaspects, in plajn arguments, care- imaginat/ive theughr discussionof disagreements. lallguages to refleet more deeply, reasoned natural philosophical theses, of structure and important methodological reasoning' of philosophical theses,,explieitly stat･ed common of on a a st･raightforward evidence, in a experimcnts, out is too huge informalphilosophical arguments more and accurately thereby. is (this in the linguistic instrumentalrole for semantics, not･ the primary, role envisaged t･urn).Third, the use ef formal languagesand logicsto provide more precise versiolls an somet･imes under judiciously philusophical theses and to elicit their conscquences, idealized assumptions, Pburth,t･heintegrationof scientific knowledge -- forinstanee, refiectioii from mathematics, physics, biology aud psychology - into philosophical on the relevant dornains,None of thesc trends originated wit･h analytic philosophy - of they have a!l been taken findearly versions of a]1 of them in Aristotle birt･ significantly furtherin anal,v. tic philosophy than evei' befbre.Of course, we also have specific knowledge from the past hundred years ef analyt･ic a huge Iegacy of more but also of the variety of theoret,ica]opphilosophy, not, justof logicand semantics in respect various t･hedificultiesthose tions available of philosophical problems and options face. one can CB: Ybu subject-matter, sophical some with wor!d, and methodolegy, exceptionalism, shadows of that science, Quine's about philosophy 'Yt)u are as a,11 the isn't reducible over other and explain the relation right; that with, exception criteria et also detect that phi]bsophy is eontinuous belongs to the totality of our knewledge of the similarity and between philosophy view or even of difference bet･weenyour position and seiences? in regarding philosophy is similar to Quine's overlapping, other branches of human knowl- dlflerence is t･hat Quineprivileges natuTal science, branches of lmma,n knowledge. He tends to to I can his cenception e.g, what my fl]omt･he other sciences, illits al. Ybu fiercely criticize philo- that it is false. Er'om your arguments argue philosophy continuous [[iherriain evaluation views, Quine's Could you etc. TW: edge. that philosophy is no argue fundamental physics. I'm much physics, ]-eject aiiyt･hiiig that lessreduct/ionist, than Quine.R/}r especialiy -74- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of ThinkiTig Deep]y, Contributing' Original]}r instance, he thinks that the concepts ultima,tely 75 beliei' and desireare of not scient,ig ically respectable, because they can't be expressed irithe language of phy.sies, But the lan,guage of phy･sics was never intended to be a tmiversal languagc, It rna;yr be the best language for the questions of physies, and physicists are the best pcople to ask about the answers to these questions, but not all good quest･ioiisare questions of physics; other words there are things of a are for other lleeded questions, If physics does kind (such as inusical not ent･ail that･ itdoes not followthat physics eiitails that there are not things of that kiiid,I take our kllowledgeof history and matheinatics justas seriously as I take our knowledge of physics, Philosophytoo has sorne limitedautonomy: itcan't ignoreconsidera,tions frornphysics, but･it･inust also pay duc rcspect CB: Ybu seem cert,ain to armchair is subst.`mLial, Could you different･ ways. TX4r: According to form one point forphilosophy, consisLs we ask must physically one Lhat/meaning means if it is true analytic of simply rnetaphysical four cate.vo' ries: or Ytm also truths are true in not cl.early? 1,helinguistic or conceptual truths. To or conceptual in thcm. Roughly speaktng, in v･irtue of its meanillg. analyt･icity; metaphysical, that, analyt,icity argue synthetic tihisrnore explain of ana3ytic `analytic' what of clariE'ying way reasoiiing. to divideanalyticit,y intothree and modal or Fregean analyticities. not insubstantial; analytic truths and epistemological, radically syinphollies) a sentence turn, the starting such c]aims assess a sentence is meta",lodal aiia'ly, ticityis ismodal-analytic ifl)aving' beiiig true, I allow that sorne t,mit,hs, such as truths of logic -analytic, rriathemat･ics, and are modally, but deny that this shows that the}r cxprcss something epistemolo.ffica]].y or metaphysicall}, obvious or iiisubst･antial, for it is not gi,venthat mologicall}, argue necessitat･es analytic t･hatno ieal error isepistemologically grounds;even sentence di$seiit on aiways themselvcs are in any if assenting to it is a necessary truths necessary, does amount or are because a of understanding eompetent is episteit. I speaker can ifthosc ground$ are mistakcn,, a theorett,olingui$ticincompetence. A scntence isFtegc-analyt･ic logicaltruth. I allow that some truths, such as truths of if it is synonymous "rith a logic,are Frege-analytic, but again logicaltr'uths colldition sentenee theoretical not epistemologically ana,lytic, A wab･r obvious. metaphysically in any way deny that ebvious obvious, that they,express irisubst･antial, for it is not this shows or' In genera], ideasof of philosophy in th ¢ analy, ticor something given that cunceptual t,ruth do not i,1]uminatethe epistemology way that proponents of the linguistic or concep. tual turn had hoped, They do not deflat･e or explain philosopbic:al knowledge. [[b argue that supposedly analytic and supposedly synthetic trut,hsare way, I show true in the same that their trut,h is handled in a uniform way by logic and seinantics, the two discipliiies most conce:ned with the concept of truth, CB: I'd like t,o know your view of t,he nature of logica]tr'uth:Vklhatis a･ logical -- 75 ･ NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 76 Tirnot,hy WILmAMsoN trutli? Are logicaltruths a priorl and Ifso, how to revise it? Do you agree logic? XVhy? TW: In my and In necessary? with Chen Bo what revisionism lvbl. 18 Is IQgic revisablc? sense? fallibilism in philosophy or of the PolishIogician Alft'ed Tarski gave the best accou]t of Iogical truth, It,is ene that has proved immensely fruitfu1 in the developinentof logicas a discipline ever the past sevei!ty of the years, First we dividet-hebasic vocabulary `non-logical'. ClogicaY ]anguage into the `logical' and the The vocabulary, typically consists of words specific structural every such words `is', as in a is st･ructural, meaning more view, such fbatures.Then `not', that sense as `or:, be can iu ;know' `all', made epistemic Csome', m,ore meanirig of that is,words precise;wc !ogic if we isa legicaltruth ifalld a sentence interpretation that keeps the the interpretation of the [and', t･he logicalwords arc only may also whosc include interestediiitheir if it is true fixed but may urider vary Ibr installce, grass is green then grass is `grass' `green' truth because it is true no rn a,tter what and refer to, green; isa logical `if'. `Gra,ss of isgreen' it･self is a truth but not a logical truth, s.iventhe fixedmeaning `grass' `green' since itisfalse under the interpretation on which refers to water and t･o ot･her words. `If dry things. The factthat t,hisdefinition is freeof epistemological or modal elements is an advanLage, l)ecauseit makes the concept of logical truth pure alld simple, as a basic coiiceptual instrument should be. The inclusion of sueh epistemelogical or modal elements in thc definition of logicaltruth would have severely obstructed the dev¢ lopment ef modern logie.It is not aut･ornatic that logicalt,ruthsin Tarski's sellse are a priori or necessary. Some people don't li]{e that consequence but I see nothing wrong with it.Logic is in some sense c:ollcerned with formal or structural matters; why shou].d it be t･aken for granted that･ the answers to such questions are a priori or necessary?･ XVhen we restrict ourselves to the most elernentary branches of logic,such as propositional logi ¢ we ean a:gue on more specific grounds that all logical truths in Tarski'ssense are a priori,in the scnse of provable, and metaphysically necessary. For lesselementary branches of logic,the picture changes. Ybr instance,second-order loglc which allows us to generalize intopredicate position as well as into namc position is esseiriially ineomplete; it is impossible to have a fbrmal systcm o/E proofs in which all and only t,helogical trutlisof sec:ond-order logicare provable, Thus one cannot expect all iogicaltruths of second-order logicto be knowable a priori. In -- the logicofpossibility modaHogic and iiecessity one can inc].udea special operator analogous to in tcmse logic,and doing so results in Iogicaltruths that don't express necessary truths. An example is is raining if and enly if it is actually raining', which is a lo.ifical truth because it ist･ruello matter hew is lnterpretedi but it is iaot a nec:essary trut･h becailse it cou]d have rained even if it is not raining in the actual woi'ld. Tarski's conception of logical truth nicely fits fa11ibilism about logic,since it makes it possible in priiicipleto think t･hata seiLtence , `actually', `now; `It `raining' -･ 76 - NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of [IihinkingDeep]}r,Coiitributing Origirially 77 is a logicaltruth when in factit isn't. In practice we know what the logiealtrut,hs are iiielementary loglc,justas know what the arithmetical truths are iiielemcntary arithmetic. Iiilesswell understood branches of logic, wc may wel! be in error about whether some our opinparticular sentollce is a logica}trut･h,and be forcedto revise ion, wkhout any chaiige in what the sent,ence means. Such changes wottld not be radically dissimilarte changes of theory in other sciences. CB: By the way, what･ du you think about Quinc's paper and its influeiice in analytic )i,Ioreover, what pi]'i'cisin" philosophy? "Two evaluation of TXN: Quiiie's philosephy, especially criticism Quine's has been immensely and lesscompelling. his naturalized the analytic-synthetic infiuent･ial, but his arguments of epistemology? distinctionin a,gainst Doglliasof Emis your general [CTwe Dogmas" t･he distinction look less Alt･houghhe showed that attempts to defineit invelve other he didn'tshow what iswrong with semantic concepts such as --- in general. Aft･er. s.v, nonyrny all, they are used iiiseinantics as a branch of linHis objection to them seems Lo bc that thcy are t･oounclear because guistic science, they,c:aii't･be reduced to purely behaviouralterms, but few philesephers these days would accept his implicit rcductionist or behEw!ouristpremise. Nevert･heless, what surprised me in writing on analyt･icity was t･osee how well Quine's conclusions about the analytic-synthetic distinetion stand up even ifone rejects his skepticism about rneanillg. No such dist･i]ctioncan do the work that philosopherssuch as Carnap exor metaphy, sics of a bunch of truths ¢ alled pectedi itcari't trivializethe epistemology There are indications in Quine:s owii writings of that aiternative, lessideological critique of an analytic-synt･hetic distinction. Other aspect･s of, Quine's work seem very dat･ed,foriiistancehis behaviourism: fifty.vearsago Noarn Chomsky demolished the pretensions of behaviourismt,obe the scientific approach to mind. Quine;s reductionism looks far t･oonarrow ancl some But of his comments on preconc:eived. logi¢ still come as a breath of fresh air. As fbr Quine's naturalized epistemology: I agree with him that the a,cquisitioii of knowledge is a process wit･hin the natura] world, but t,hat･point by itselfhas comparative}y few methodological consequences, if one doesn;t･ combine it with his extreme reductioiiisin. Fbr instanee,it cloes not forceone to try to tran$late the werds and into the language of fundamental physics. That language probably can;t express any interesting episternological naturalized episteinolo.cry questionsat all. However, on.e can see Quine's as an extreme manifestation of a more away from a general shift in epistemology, Cartesianinternalistapproach, on which the start,ing-point is restricted to what the semantic concepts, "anal"rtic'. `knowledge: subject knows. In studyj.ng someone's knowledge of the world, that･know]edge. For instance, we all our psychology of perception. This even can niore `belief'; t･heoristsc:an use if the person we are st･udy, ing doesn't share use knowledge gained frornexperiments on the has been externalist approach to epistemology epistemi,c situation, we as -77- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of a]d Chen TirnothvttXny･TILLI,NMsoN 78 Bo Vbl. 18 importalltdevelopments in the period since 1960. 0ne sees it in episternolo.ffists such as Alvin Goldman, GilbertHarman, Robert Nozick and Hilary and it is at least par't of what Kornblith and in Knowledge and d,tsL･imi,ts, Quine for by naturalized episteniology. I use it in Th,e Phitosophy o,f I'hilosophy too, meant of reasoning instance whell I a,ppeal te t}'teresults of experimeiits on the psychology of logic. in discussing the epistemology one the of most CB: Ybu that･ like all philosphyalso needs thought experi6aid, will benefitfrom an attitude that looks more acmellt･s. Ybu also how to apply thought tively foropportunit･ies forinodel-building." Could you explain and Tnodel-building inet,hodology in phi!osophy? Pleasegive some examexpcriments argue sciences other `CPhilosophy plc(s). is ]3ertrand Russell'simagiT",':A famous example of a t･hought experiment clock that by luck shews the uary example of Iooking for a moment at a stopped correct time. The person acquires a true belief, even a justified true belief, about what timc itis,but the beliefdees llot constitute kiiowledge.It followsthat iieithcr true belicfis sufficient lbrkiiowiedge,In irnagilling thc t･ruebeliefnor evell justified not･ know the time one is perexample and judgingthat the person involvedwould thought･ experiment. One learnssomething about what weuld be fbrrnjng' a simple from which we caii infcrsomething about the case in a specific possib]e situation, in thought experiynent in t･hiscasc, a,bout the nature the properties involved th,e of knowledge. Epistemologists alld other philosophcrs use such thought experiments from thought experiment･s, Typically,models use continually, . )･Iodels are difft)rent of a, forma] desc:ription a possisimplifications and idealizations in order to achieve by forma,1reasoning rather than by bilitywhose consequences can then be elicited imagination-based judged.For instance, we can set up a model in epistcmic ]ogiea6 is a subset fo].lows. The worlds' are the numbers O, 1 and 2, A `possible of {O,1, 2}. the set member `propesition' A proposit･ioiiis `true' at the pToposition. A Telatioll ca]led if the numerical difference betwecm them of and only a vLrorld isat ifand most only if the world iasa holds betweeiiworlds if 1. The proposition that a givenworld only ifevery world accessible propositien X contains brom that oitc is a niember of X. Thus if X is the propesition 1}, the proposition aecessible from O are O itselfand 1, t･hat oiie knows X is just{O}.IIbr the worlds one knews if and [aecessibility: a {0, both does of whicli not belong to belor;g to {O,1}, wherea･s {O,1}. One can world then 2 isaccessible conc]ude from worlds the proposit･ioni,hatone which is the empt･y {O}, 'Thus knows X isthe prepositioiitl]at is ne world from which only worlcl O is acccssible, knows X without knowing tliat, olle knows X. In other words, that sinee one there fbrmal axiom one to the positive in'trospectionaxiem. counter-exainple failsin the knows model is purely Tnat,hemat,ical. 1 and 2 and [I]hatis an at we The advantage knows set O world {}, one have constructed tha,t the argument over a a thought ---78 - NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 1'hinking Deeply, Contributiiig Origiiially Aiiotheradvant･a,gc experimellt. from the in way it was which 79 of the ist･hatthe formal consistency specified. A disadvantageis that the scenario is clear model builclsin of idealizedassumption that ene knows all the logical consequences what one knows. One has to argue that the idealizationis irrelevant to the point we is surncicntly are currently interestedin,pesitive introspection, and that the cxample realistic in rnore relevallt respects. Those sorts of issue arise for models in nat,ural science too and c;an ofteii be resolved by good judgment wit,hout destroyingthe value the um'ealist/ic, Sueh a of the model. is applicable methodology to philosophy Loo. CB: Ybu argue tha,tphilosophy needs and uses evideiice as all intelleetualdiscicensists of factsor true propositions. I want to know, plines do, amd that evidence on earth is a fact? Is a fact ob.iective or subjective?･ How te in your opinion, what individuate facts,in other words, how to couiit them, e,s,, are t･alkingabout your alld tongues are making soupds;', are ncw book':, are sittiiig in armchairs", ef a saine factor different facts?How to discriminateth ¢ m? they different versions `:"i'e "Vtt'e "()ur [I]"J': [I]hediscussioi] of in The PhiZosopuhy of Philosophy ls oiie of the of Kno?vZedgeand its LiTnits, cmphasizing pla ¢ es in whi ¢ h I;m using t･heepistemology is part of our evithat･we are not always in a position to know whet･her somethin.cr dence. The equation E K applies to philosophy too. As forthe notion of a, fact,I don't appeal t･oa hjghlymet,aphysical conceptioii of facts. It suffices formy purposes treat facts as true Since truth is an ob.iective mat･ter, so is being a to proposit,ions. my new book iiithe normal way while sltt･ing in armfact.If we are t･alkingabout chairs then the proposition that we are talking about my, new book, the proposition evidence = that we sounds are are none of making armchairs t･rue,so they is neeessarily all them book new in sitting without sollnds: sitting are t･he proposition that eur tongues are making facts. They. arc tln'eedistii]ctpropositions because and all equivalent to in armchairsl we t,ongues could our make any of con]d sounds t･alk about rriy t･beothers. "･'ecould sit in arnnchairs "dthout our tQngues without our talking about iny iiew book. Thus they are three dist/jnct･ faet･s.Although more questions can be asked of facts, the epistemology of 7ihePhilosophy o,t'Philosophy about the individuat･ioii is compatible CB: Ybu and enabli,ng. with any I'ast experiellce mathematical has t･hatexperieiice claim the habitsof imaginati,on, our answer reasonablc could abilit･ies intuition,has some leasttwo different･ fuiietions: evideiitial have the functionof mouldiiig, res,ulating our of judg,ement, so all knowledge, cvcii including with The distantconnect･ions past experiences. thc cat･egories of the a priori and produces littleinsight,in fact,it is wrong. iise of you an euipiricist to them. at a our knowledge posLerori to characterize Could you explain further?Could I call iiTepistemology? T"J: In playing an evidential role, experiencc supplies us with new evidence: for - 79 -- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 80 Tamothy Chen Bo W'II.I.IAMsoN and V6I.18 instance, the faetthat a man in a red shirt iswalking past is added to my. I that a man in a red shirt is walking when see past. In play,ing an enabling suppl: ing us with new pericnce ena,bles us to acquire new coiicepts without foTjnstanee,we by others to visibly distinctionbet･ween priori'and of `a kiiowledge in no evidential role. but role A posteriori kiiewledgeis knowledge in whi'ch R)r instance, on that coiiception, one can know role. experience of evidence: the applicatioll red Ca plays experience which through red role, ex- things. Oll the standard way of drawing posterioriTkiiowledge, a priorlknowledg'eis `red' of th ¢ word a the concept acquire may evidence may play an plays an experience ena,bling ei[idential priori that all red t,hingsare because to acquire the concept of red, kno"r a post･eriorithat a man in th,red a even though one needed experienee is a purely eiiablillg role, One can only shirt is walking isthat experience of past, Thc problem I raise for this distinction krigwledge t,hat is neit,her purely enablillg nor strictly ten play,s a role in armchair coloured, that In doing thought evidential. relevant clock, concepts our ski]1 oMine, in properly we use our skill in apply. ing the in the imagination - forinstance, in the case of the st,opped applying exper;iments the c;oncept･ of knowledge. We are net really experiencing lookinga,t a stopped clock. Experience is playing a lessthan evidential role here. But even if the concept･ has an inna,tebasis,our skill iiia,pplying i-t - which - "Tas devcluped through to possess the coneept exceeds what is required merely online experience in perceptionof a ¢ tual cases of knowledge and ignorance or relat･ed and actual uses of the word That isa more than enabling role for phenomena experience, I argue in T7be Philosophy ofPhilosophy that ho"iever one triesto fitthe a priori f a posterieri distinctiQn te such examples, itdoes not work in the intended and obscures underlying similai'ities and diff'ercllces between cases. I den'tt･hink way, of rnyself as an empiricist, but rather as someoiie trying to finda middle way between empiricisin alld rationalism. Ernpiricists emphasiz･e a posteriori knowledgel rationalists emphasize a priori knowledge. I'm emphasizing knowledge that doesll't into . Anoth r respeet in which I'm not･ particularly fit･ comfortably･ either category. ¢ einpiricist isthat I believeon thc basisof evidence from linguisties and psychology that there is a large innate element in human cognition. someone Cknow'. CB: Could you bricflyexplain epinion aboiit philosophy of mind, the role of iirtuition in pltilosophy? "ihat isyour the especially studics on intentionalityin cont,em- porary philosophy? TXK,' : Some philosophers talk as though ii!t･uit･ion i'sa separate type cognition, eiie nced to de thought experimems, for instance.When I ca,me te analyse such supposed examples of intuition, I realized that what was going on was lessstrange tha,twe aBd ment distinctivethan that. IJiorinstanee, in whose if certain eontellt conditiuns is a coullterfact･ual obt･ained: "If a thought conditional those conditions expcriu]ent about obtained. onc what･ ' he makes would wollldn't･ be ajudg- the case know the L80- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of Thinking Deeply,CentributingOrtginall.y t･ime".I re,ieet･ the are intuitlonas misleading, Ybiiare right that these issues Few philosephers of mind usc thc concept of philesophy of mind, concept to the related 81 of lntuition.Parts of Knowledge and itsLimits concerii the philosophy I discuss how an agent's knowledge and other environment-involving can causally why they environment as explain in the external states of affairs in the external ternalist,success-oriented as falsebeliefs and ernpty a central mind, where mental states they did. Just as I take reference to objects case of intentionality, so I talkeknowlcdge of environment as a central in which phi!osophy of mind, names) ls understood inrelation malfunctioning as aeted of to l'unctioning. By in case of t･hought. It is an unsuccessfu1 relatiell cogn!tion to suecessful ex- (such cognition, internalists take as basicmenfailure, su ¢ h as belief (which be may contrast, between success and true or false)and apparent referenee. From a methodological point of view, the philosophy of mirid raises fascillating questions. In the l980s many, people thought that it was taking over from the philesephy of language in the drivillg-seat of philosopliy, in some cases because they thought that the coiieeptual turn vgias more fundamental than the linguistic turn, in ot･her cases hecause they thought t･hat breakthroughs in cxperimental would transform philosophy. Philosephy psychology and neuroscience tal states that ol' mind are neutral then hasn't since rcally lh,ed up to Lhat promise, Indeed, no branch ene of even ifsome are more peripheral than philosophy iselearl,v, centr,al to the subject, others: instead, ],ikeinost･ oLher acadeinic discipliries, philosoptiy has become more specialized. Experimcritalwork has ccrtaii)ly, contributed to our philosophical understanding of the mind, and will continuc to do so, but the idea that we could just read our philosophy of mind off the experimentalists' conclusions was always iiadl've. CB: be 'Ybu said, is a branch of bot･h phtlosophy i`Logic iiitumutually 6eparatcd exclusive contributions and from it cannot matheinatics; mathematics and phjloso- Recelltlyyou finisheda new paper and philosophy Chinese translat･ion will be on The Jts published ofPeking University.Could }rou claxiCy' the rclationship betweeu logicand philosophy in 21t･h century? X･V'hat ean logicdo forphilosophy, and what caii philosopl]y do for logicin I agree ph.y" ill21th with "Logic y,ou. century" .JowirbaZ this eentury? TVKr:Logic isnecessary but not forsolving suMcient to use long chains progress in philosophy we need it is very hard to distinguishvalid from inva].idreasoiiing, make more to Logic hclps us to do that logic enables For installce, the developriient of quantified modal us far more securcly than belbre about possibility,necessit･y and existence. reliably. reason Secondly,logic plays a talking abeut before t/ributesspecific of philosophical problems. To of reasoning iiiareas whei'c rrLat･herxiatics. central I gave role an illthe exauiple sort of from activity inudel-building epistemic logic. rl"hirdly, that I was legic con- to many branches of philosophy, not only to the philosophy, ]Jbr inst/･ance, oiie of the inost active and fl]uitful areas of researc:h results 81 - NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 82 on and Chen Bo WILLIi",Isorv Timothy t･ruth coiicerlls solutiolls isnot truc?' saying iswhat important too, with In the science, t･othe paradox of the liar,who says he issaying true? Most of t･hatwork froiiimathernatical t,echniques solve. concerns but then happen in other phcrs possible examp]e can sometimes epistemic of action is the logic of completely diMcult and teehnical autonomous now to computer supply ncw problems for logicians logicerr]crged from philosophical generality,which unrest･ricted logicof generalizations about, absolute],y everythin.cr. Thc but many of the problems are tricky teehnical elles. CB: Ybu requires development,the saniLe Logician$are on the look-outfbrnew problems, which see because they are lookingfrom new, unexpected underwent areas. logicand moda,] will I am C`What is becoiiiing illcreasing!y action both to t･hephilosophy direction, philosophy Just as to try to logie. The logic of applicatiens reverse V61.]8 thing will philoso- A an.gles. co]'Lcerns the isphilosophical motivation formalmethods extensively in your philosophical research. But you also e]aiin, should be clea,r t･hatinformal methods ai'e also esselltial to philosephy, forforinalmetheds, likeall others, easily mislead wheii applied witheut good informal I agrec wit,h you strongly. Could you explain more about these cla,ims? judgement.'; use "it T"i Most philosophical problems initiallypresent thernselves in vague informa] terms, To apply forinal rnethods, one must formaliz･e t･heproblem, express it in precise fbrmal terrns. One needs to determine whether the fbrmalgood informal judgment : is most･ important about t･heorigillal problem, For instance, in the process of formalizin.ff. the sirnplifications or idealizations made harrnless, do thcy omit the heart of t/hedifficulty?Judging the plausibilityof foriiialized izatien eaptures are or premises is also whose solut･ion skcptical of what an inforrr]al mattcr, fbrmal method$ have And nothing se T'herernay on. at all to contribute, eveii be problems t･o although I'ni a bit that･. CB: Ceuld you listteu iniportant acafleinic far`[' Could you describe t･heir main coiLtributions papers from justin one all of sent･ence your papers so separately? TW: I revised many of my best-kiiownarticles forincorporat･ioninto my books. It will be more interestingfor nie to describeten t,hathave not undergone that proThey he!p cess. will te give a better sense of the variety of m}r work. My. very firstpvtblished art･icle was Disproved?; (Analysis, 1982), a short piece `Iiituitionisrn in which I showed t･hat the argument of the `paradox' of knowability against the knowable doesn't･ work by itselfagainst an anti-realist who fo1,lows rvIichael Dummett's proposal that intuitionism is the right logic fbr anti-rea],ism; of course, I reject the anti-realist princip]e, but I think more argumeiit against･ it isnceded. In Relations'(Phitosophical Review, 1985) I argued tha,t pa,irssuch as kicking and being kicked by, er above alld below correspond to t/hesa,me language-independent feat･ure of reality and explained some of anti-realist principle thaL a]1 truths arc `Cenverse -82- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of DcepLy, Coi]tributing OriginaUy Thin{dng the logicaland paraLive problems seinaiitie Siml]arity. Damc (IVotre by that idea. `Flrst Order Legics forCornoffuT'malLogtc,1988) discusscsthe logic raised ' of sentences such as .rournal is more LAima 83 t･o Betty than similar Chris is te David', L'[I]wo ofSymbolic Logic,1990) (Journat Incomplete Anti-RcalistMedal Epistcmic Logics' explores what the logjc of an anti-rea]ist theory in possibilityof the theory not oTy some specified to Iigentthan cemputers, conception that the have t･oleok likein would the Knowledge co-authored a colleague allows to one between of for order Turing Machines' (The- of with mine defend the Barcan ferrnula,a principle cont･roversial that great!y simplifies quantified tiie Structureof IIigher-OrderVagueiiess'(Mind, 1999) takes the relation logic.`On existence p,ossibility and iiim},, book Vkegucncssfurtherb}･r showillg of vagueness analysis to the the theoretical economist Hyun Song in Oxford; it concerns the way in vThich be constrained by t･heassumption that agents are not more intel[Bare Possibilia?(Erkenntnis, 1998) develeps a metaphysical logiccan epistemic modal collapse. then was status [Representing' Decision,1994) was and Shin, who about epistemic truth is equlvalent which borderlinecases article is of and so on have obey hew borderlineeases logicalprinciples.My unexpected most cited How' (,Journal o.fPILilosophy, 2001),co-authored with t･heAmerican philosopher Jaseil Stanley (whomI firstmet･ whell he was assigned to be my teaching assistant at･ MI[I]); ituses evidence firomt･hesemantics of Englishindirect･ly to ar.vue that krtowing how to do sornething is a form of knowing that something is t･he `Kiiowi-g GilbertRy.Iea,nd many other authors bad claiined sorncone said that people react･ed to tbe paper as though they had just diseovercdthat Santa Claus does not exist. Existents'(in a volume edited by Anthony O:Hear, Logic, V'lhoughtand Lang7tage, CambridgeUniver$it･y Press, 2002) is a fo11ow-up to Possibilia'; I show liow simple, plausible principles about existence, proposicase, to contrary what `Necessary [Bare tions, truth and necessity given iiidividualexist,s -- a entail the (:enc:lusioii t･hatit is not conclusion contingent whether first but iuakes sense that itinvolvesfroiii inure t･hat souiids crazy at･ the purely logicalnotion of existeiice you distillguish cret･e iiotions of existerice, such as beirigin space or tiinc. a oiice coii- (PhiZosophical CEverythiiig' Perspectives,20e3) is a defence of the int･elligibility ot' generalizing simultaneoiisly about absolutely everything to geiieralizewithout paradox of agaillst w･hatsoever, restrictions who 'that the argue leadsto paradoxes in set theory. su ¢ h , the set of all sets that are not itself?); I show critics members how dangerous such paradoxes are, of themse]vcs it a (is as attempt Russellis meinber but/also how t･oavoid them of at t･he last rriinutc. CB: Could you tell me something hope to aehieve in the near future? TXV: I recelltly its Limits; the fillished writin.cr articles and my replies about replies will your current research? to fifteen articles soon on be published in "rhat do y,ou Knowtedge a volume and called -83- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of 84 Witliamson and on University (Oxford Knou;ledge Duncan Pritchard.Now I'vehad to ofPhilesophy, Zosophy fbr symposia on in doing somet･hing iiew about in articles possibility, existence and and - and ";'im,JAts･fsoN Chen Bo Timothy Press), sta,rt writ･ing that･ book. My eonstructive. of iiext inine edited repljes XJOI,18 by Patrick Greenough to articles on The Phi- I'm rnore interested Of course, boek will be based on the ideas such Pessibilia;and `Bare as CNeces- Existent･s;, which Ijustmelltioned. I hope to strengthen the a]cguments by using formal sy, second-erder modal logic. Parts of stem, the book wili be powerfu1 rather techiiical,but I hope that some cent･ral parts will be accessible to ull analytic ii].ctaphysicians. The working .Rigidity;it will be published by titleis Ontologi･cat 0xforclUlliversityI'ress,After t･hat who knows? sary a rnore CB: AsItalked with you befbre,although Iadmit that Kripkc is a v ¢ ry. great objections to hisbook Naming and Necessity. philosepher I still hatresome systematic AsI know, you are a strong Kripkean. Could you say something about Kripke and his works? Saul Kripke give the prestigious John Locke lecturesin Oxford; my tutor had describedhim as Lhe brilliant was t･hen in his early, 30s). AlthoughI yoimg llsan of Amerieaiiphi!osophy (Kripke didn'tunderstand everything he said, l was hugely impressed -- by his clarity, acTW: As a first-yearundergraduate, I went to hcar discrimination, logical power, ima.ffinationand curate sense, common all combined huinour, Evey siuce he has beella uiedel te me ef how phi!o$ophy should be done, He probably influencedme more than any of the Oxtbrd teachers did. I especially liked,and like, his insistence oit keeping metaphysical questions scparate fr'omepisteinological isthe basisof hisfamous examples of questioiis,which the contingent a priori and the neces$ary a posteriori. Although some of the deta,ils in Naming and Necessitymay not be quit･eright･, the plcture he presents there strikes a with me as touch of basically correct books. Of - which i.snot something I;d say about inany phi}osophical hisear]y t･echnicalwork on moda,] logichas also beeu of fLindamental importance. Even if one disagrees with seme aspect･s of it (as I de, for instance on the Barcaiiforinula ment･iolled al)ove), jt'sstill the bcst starting point forthinking about the metaphysical issues raised by quantified modal logic. Hi'sarticle en tT.'ut･h has illspired inuch of the leadiiig techiiical work on ti'ut/h over the past thirty Kripke's later book Wittgenstcin ort RttZes and Private Language was rejeeted years. by most foIlowers of XVittgenstein as a misinterpret･at･ion ef him, Mayh.e it was, as a mat･ter of historical scholarship, but what they rnay not have realized is that Kripke was effer'ing VL[ittgenstein a sort of lifeline. In the 1970s t･he positiollwas roughly this. VK･)ittgenst,ein's followers held that X]Vitt･geiistein's central contribution to philosophy his was argurnent against the possibilit･}r of a private language.The private language 'appeared argument of selltcnccs course to ascribillg rest on a verificationist mental states consisted premise, to the in their method effect th4t the of verification. meaning Since -84- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of Thinking Deeplb,, CentribuLingOt'iginally forsome (except philosophers ited, it was widely most anti-realists) regarded premise, it was held t,hat it did indeed rest, on a verificationist unsound and discreddid rest oll Ant･i-"iittg'ensteinians verifieationism if the privatelanguage that agreed 85 argument uninterestiiig. as really premise. "['ittgensteiniansinsist･ed to expla,in coiiviiicingly. hew it could work without way of iiitcrpretingthe argument that explained a verificatienist that･itdidn't,but were never able such a premise. Kripke offk]red a how it could work without a verifieationist premise, which ,younger generation "VLrit･t･genstein's made work for philosophers of a who wercn't iinpressedby "J'ittgensteinian defencesof the private language argument, Once it was agreed that Kripke's interpretation didn'tfitWitt/genstein's intentiolls,, the private language arlost most of its int･erest, except to committed "J'ittgcnsteinians, As a result, gument "Tittgensteiii's work has become illcTeasingly marginal te the development･of ana,lytic much of Kripke's irnportant work remains unpublished, philosophy. Unfortuiiately, foriiistaiice en identityover time and on some epis'temolog/i.cal questions, but many of the ideashave been absorbed by the philosophicalcomrn,unity froinhis lectures and samizdat transcripts of them. It isa great pity that he hasn'tpublished more. relevant again CB: There is a familyof :realit,y', phers Cactuality', et,c., sometimes be',`to `to as exist', `being(s)?, `exist･ellce;, Chinese philosodelicatediffe:ences bctwcen them in philesophical tell the cannot such is very which Could yoii c:larify context,s. concept,s import･ant in met,aphysics. them for us? T",':Nat,iveEnglish speakers gct often eonfused about the differencesbetween them in philosophical context･s t･oo!Each of those terms is itself used iiia vaxiet･y of diflerentways. Often,whcn subtle distinctions with them, they philosophers make are not simp]y, relying en a clear but subtle differcn ¢ e in t･hcir usual meaningsi rather, they are harnessingthem as technical terms fortheir own theoretiicalpurpo6es. For illstance, ifa philesopher says entities have being but not existence" itishard to knew what distinction is being made until one hears the rest of t,he philosopher's [CSome , discourse. In `exist･; papeT my is axnbiguous }ogicalrcadiiig, `cxisti `Bare between a logicalreadiug aiid a is equivaleiit to soinethiiig' tEverythiiis, to t,helogicalti'uth alent more is obviously exists: is soinethiiig', `Everythillg I suggest coiicrete (inthe Cbe in t･hissense, [soinethiiig')i Possibilla'(ment･ioned above), On that the readiiig. word Oll the general sense tr'ue, because it is equivmost the inore coitcrete of reading, like in spa ¢ e or time'; in thissense, to something morc erything existst makes the highlycontroversial claim that･everything is in space er is false set is somethillg that is not in time arguably, it, a cla,im because the empty and in philosspac:c er time. Tb bi'ing out･ the difference between ophy, it is helpfulto notice that is typieally contrasted with or is equivalent iexist' `be LEv- ;reality' `actualit/y' `reality' `pretence' one would whereas naturally. `actua]ity' say. CIf `appearance; is typical]y contrasted with I had eaten much less,I wuul.d - ]Jbr instance, much less `possibility:. have weighed 8i]- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of Ttmothv v NNilT,uAMsoNand 86 I actually do' ra,ther than I had I really. de': tli¢ comparisoll is between `If than not love scei)e in love with a other', appears each other' or rcaders they when the nuances see of are possibility,but Converse]y,if one js watching `Those to say not actors actually that it is,while are not really iiilove with cacb with the fictionthat they are in loverather in love.But in many very centexts it matters `Did say either )Iacbeth actually murder his murder Englishusage, a hi$predecessor?'. My text is not philosophical but rather to t･ryto work out to Chinese tDo much advice to worry this part･icular what the terms, CB: How do you dcfineyour ewn philosophica,lposition? In beliefsL? your majn philosophical T"r: I st･art･ with not really these terms in philosopher' isdoing wlth less than a non-a,ct･ual case. are have weighed would and a,ctors VbL 1.8 the eontrast Maebeth CDid predecessor?' about than rather `Those thc possibilitythat they. are term ene uses; one could little whieh actuality or sillce the fbcus is on thaii with less,I eaten is pretended to be the in thc theatre, it would be natural that one Cheii Bo classical other words, what logic and truth is saying is the principle that falsityis sarsrillg of what what isthat it is.What istrue depends little on of world, with no special humans. "Je form just, a tiny bit ef the natural privileges. But, despit,ehuge areas of ignorance, some accidental, some necessary, we haNe signifi systematic inquiry. icant knowledge of' t･heworld, from bet･hcommon scnse and more everything is what it is and not another thing. I reI'm suspicious of reductionism; jectall religious beliefas consisting largelyof falsehoodshuiuanityshould have growiN out of but can"t. We should regulate our belief by the evidence, however diMcu]t it/ isto do so, and not give way to wishful thinking, Aii anti-realist･ philosopher once us described me as a `rottweiler realist'. I'm happy that description, with CB: In your opinion, what are the most impertant cha.racteristics ef a suceessful philosopher?"rhat kinds of basictraining, abilities and knowledge should he have? XVhen a young scholar intends to involve himself in a research project･in philosophy, what elements has he to consider' carefully? Ibr example, how sheu!d he choose his topic? Hew should he do hisresearch? TW: It;shard to give general no one and other size Nevertheless,olle forsuccess in .iustabout wonit succeed unless you other. that suits other one foxNr obvious diings. Some qualjties are needed any academic field.Naturally,hewever clever you are, you work hard. YOu need the boldnessand imaginationto think t,hepatience, persistence and aecuracy to developthem properly. can say thoughts, and If you have neither the power from others, wi,11 never your wo]'k new young philosophers because in temperament･, skills philosopher doesn't suit an- to ambitious differfrom each advice fitsali: good philosophers of working qualities. A way of a self-criticism improve. Ybu nor must to the modesty be able to accept recognize eriticism when your -- 86 -- NII-Electronic Library Service the JapanAssociation the Japan Association for forPhUosophy Philosophy of Science Science of Deeply, Contributimg OriginalLy FI]hinkin.[.r teacher has made 87 but ifyou are too willing to attribute mistakes you will not learnenough from them, It'sdiMcult･to findthe iniddle way,. Y6u must learn to write clearly, precisely and unpretentiously,, Don;t･worry about whether what is deepF,worry abeut whether it;strue, I Iike philosophers who rnove you arc saying slow].y and carefully, one step at a time, but of course they will not･ get･allywhere unless have a sensc of where A they they are g.oing. good basictraining in philosophy, includessome logicaiid some hist･ory of philosophy, and lotsof pra,cticeiiiwriting abeut and critieiz･irigalteriiative problems in you,rown words, analyzing philosophical solutions amd having your analyses and criticisms analyzed and criticized illdetailby your teaehers, Choose a research topic tha,t jnterestsyou and on whieh you feelthat te say; it should be small enough to be manageab]e is you might have something (it much more a mistake, to choose common It･doesn;t･ matter topic that is t･oolarge than M)u you say intelligible to them, but it is boring conversatioxL, must one that is too sn]all). the topic is fashionable the most fashionab!eareas are should have been some on it te definethe previous work Contributingto the published discussionof starting-poinL. sa'icl, and whether but there over-crowded a make what relevant uiiless to you a wh,at add topic is likejoining a thc others somcthirig o' have just f your own. t,o t,alk with m, e. I CB: Professor Timethy "['illiamson, tha,nk you for agreeing work guess Chinese rcaders have got a letof inforinationabout you and your academic from this talk. I personally appreciat･e your supervision, advice, help and t'riendship forme while I have been an academic visitor at Oxford Facultyof Philosophybetween Aug,ust,2007 and August, 2C)08.]N,Iay. to succccd in yeur philosophical you continue endeavors! T"･J: Alld mabg ing':,our of and analytic you teo, Dr Chen! It･has been a pleasure to get to kllow you durst･ay in Oxfoi'd. Ybur work is an important contribution to the development between Chinese philosophers philosophy in Chiria,and to cerninunication philesophers i'romthe rest of the world. 2009.4.27; Aecepted 2009.6.14) (Received -87- NII-Electronic Library Service