Ontological Alternatives

LOGIC AI\D RTALITY Bg the Same Author

T HE

ME TAPHYSICS

P E IL OS OP IIY

ME A N ING

OF

OF

AND

SCIEN

LOGICAL

P O SIT IVIS\{

CI]

EXISTENCI.]

CustaaBergnann

' I' IIE UNIVER SIT Y OF IVISCONSIN PRESS r\t.\r)tsox 1964

oNTolocrcArr ALTERNATTvES

Ontological Alternatiues* recent essayl Dr. Egidi, stating rvhat she takes to be Fregcr,s T* ^ ontology, sterts from and usesthroughout as a foil *-hat in a'otri.r. I J- s55ay2r ha'e said about Fregc. she holds that Irrege is an iclealis{, takes me to hold that lie is a nominarist. ,\s I usc ,nomir]alism,, I,'rcgt,is not a nominalist' Nor did I ever say that rre rvas o^e. r merely tried t, shorv tliat there is i' the very struct.rc of his o.torogy a i.r.,r.rr,,.. torvard nominalism. That makes thc title of my cssay,,,ri..g.'. rliclck,rr Nominalism," suggcstivc rather tlia' accurate. I' spitc ol ri-riat l)r.. Egidi says, I still believe that the nominalistic tcndency is unmi.striliably therc. Nor lras she cou'i'ced me that Frege is an icrearist.Srr. has, horvever, sho*-'very co'l'i'cingll'rvhat r, Ior on", had not sc.rr as clearly as shc does,namell', that thcre is also an urimistakable .strrrr.tural tendc'cy to*'ard idearism i' rrrege's ontorogl'. 1l1.,-s,if r *-cr. asked to choosefor her essaya titlc as suggesti'e and as exaggeratctrrrs mine, I rvould call it ,,Frege,s Hidclen Idealism.,, Those lvho are mercly clerrer sometimes cliscor-er tenclcnciesn.hir,lr are not thcre. Those rvho disce'r hidclen terde'cies rvhich are th.r.r, think structurally and, sometims.s,profoundly. In my judgment I)r.. Egidi's essay is of a profu'ctity thai dcse^.es high praise. 1-et I alr,, judge much of what she says to be radically mi.staken.Becauseof trics. mistakes she makcs an idealist out of l'r'ege. Hcr mistakes reflcct tlr. idealistic ambiencethat has'ourishecl L... bo she shoulcl'ot be bla.r,rl * r'his is the English original o! "Arternatiueontorogiche:Rispostaalra ])rl tores-sa-Egidi," which appearetr in. Giornarecritico delra Filosofiartaliana, 1,',

I : t t ; 3 . , r ' n , r O b y p e r m issio n . r" L a c o r r s i s t e n z a F ilo so fica d e lla L o g ica d i F r e ge,,, Gi ornare C ri ti co dti l ,t I t ilou t f t a I t a l i a n a , 1 6, 1 9 6 2 , r g 4 - 2 0 g . se e a r so h e r ,,trIatcmati ca, rogi ca e {;r. s . f i. . o l l ' o p e r a d i Go ttlo b F r e g e ,,, p h ysis,4 , 1 9 6 2 , 5_32. The essri ,s r.i l l t,,. r: ilc rl l s C F L F a n d \{ L F F , r e sp e ctivclv. ? "lrr.ge's Hidden p h ito so p h ica l No m in llism ,,, R eui ew , 67, l g5g, 4B T i \l S c r' :r l s o " I ) r o p o s i t i o n al tr u n ctio n s,' , An a lysis,1 7 , fg b 6, 43_4g.

t2,1

125

for them too scverely. Those rvho knorv horv difficult metaphysics is tlso knorv that, since its core is dialectical,there is ngthirlg paradoxical about judging an cssay in metaphysics to be both profound arrd profoundly mistaken. Dr. Egidi, I just said, makes an idealist out of l'regc. Tliis is not rlttite accurate and the inaccuracy is of a sort to rvhich one as comInitted as she is to accurate intellectual biography might rvell objcct. l,)xplicitly'she merely claims that Frege's renovation of logic and, inscparably from it, his anaiysis of the simple clause 'a is F' is implicitly i
126

L o (i rc AN r) rrEA r ,rry

these several acts. Acts arc tnental cntitics. Irctcr,s lreinq blond is rr nonmental entity. 'I'hc i.te'tions of some acts ,,,.n ,,rc,,trl; those ,r' some others, no'nicrrtul. llcst.ting these bits of coni.r.n se.se sho,r,. ,errtity', ,act,, ,intentio',, ,tnc'tal, horv, commonscrrsir,lrll.y, a'd ,'.rrm ent al' s ' ill li o .s < ' ri . Irl e ,l i s m h o l d s th a t al l enti ti es arc r' c' tal ; m.terialism, that tlrr'.1'urcall noumental. Only realism: ,qitles n.ith commorr sense, assc.ts tlur,t (1) some entities arc me.tal, somc rro'mertll. Mate'ialism \\'(' mrry safcry dismiss as absurd. (r) b,v itserf is an empt' husk. Ilcalists2salsr must assert and justify that (i) minds ca' kro*. u'hat is rr.rrmerrtal.To justify (2) is to prese.t and to clcfc'd agai.st .ll dialcctit'rLlrlttlcks an ontological assayof acts and tlieir intentions tli.t fulfills t*'. co'ditions. (a) when I believe *-hat is falsc, the act is ilie.c, its irrtcrrtio. is not. The assaymust account for such acts. (b) The assar. must prtlvide a "conltection" betrvecnan act ancl its intentiori r-iricli is so "close" that it justifics (2), irrespecti'c of thc i.tentio, is (a) rnc.tal or (B) nonmental. (a) leads to the '-he.ther dialectic of thc carte,qiarr cogito; (0), to that of (2). An example rvill help. Accorcling to th. Aristotelian-Thomistic accou't of perception, whcn percci'e r a tret,, trvo substances,mine and the trec's, exemplify one unir-er..al. Clearl.v, this "connection" is sufficiently "close." The dialcctical difficulties of (a) and (b) are notorious. so I need not a.d shail not here co'sicrc. them except as they irnpinge upon that particurar picce of trre diarcctic I propose to unravel. (what has been said in the lasi'peragraph sufficesto unpack trrc apho.ism that epistemologyis merery theintotogy o;f the kttrncing situafe,tz. 'l'his usc of 'kno*ing' is of course generic, comprchcudi'g thc scvcrrLlspu'it:sof percei'ing, rcmembering, belie'ing, imagi'i'g, do,bting, c.tc.tai.irg, and so on. Henceforthlspecics, will be usccllnly for thc s o k ir r r lsof " k n o u .i rrq .,,) Yesterclty I pcrr,cir.cdthet peter is'blond. The act that ri.ould havc occurred if inst.:rd .f pcrceiving this fact I had remembercd it is clilferent from the ..c that a.tualry occurred. This no phirosopher has ever questioned. 'l'h.t sho*'s that they all take two thi'g-s for grantccl, namely, first, that a' rr.t rrrl its speciosarc both mcntai e.tities, ancl, second,that the latter is lr .rrstitue't of the former. If instead of perceiving that Peter is ]rl.'tl I h.
ON TOLOGIC A L A LTE R N A TIV E $

I27

I'rima facie that provides an ontological alternative. One may hold llrtt the "t$'o" acts are one and not tlvo. Or one may lirild thilt cnch Irt,t has a constituent such that in ttvo acts this constitucnt is thc 's:rmc il' aud only if they have the same intention. Call a constituent of this As one ordinarily speaks,'thought' is used in trvo u'ays. sort a th,ou,ghf. once thc u'ord stands for the act itself; once for a constituent of it w'hich varies rvith its intention. That makes our usc of 'thought' in this cssay technical, although only in the very limited sense that rve shall crnploy the rvord in otily one of the two rvays in rT'hich rve employ it *'hcn speaking as u'e ordinarily do. (Some philosophersuse 'thought', tct'hnically, as I use 'intention'. The dangersand inconveniencesof this usc are obvious.) Thoughts, as we use the n'ord, are constituents of ricts. Hence, i/ there are thoughts, they are mental entities. If there are rrone, thcn s'hat I am about to assert of them holds of acts, i'e., of those mental entities rvhich, speakiug as rve ordinarily do, we sometirnc-qalso call thoughts. That is rvhy I nccd not as yct commit myself lusto ii'hether or not there are thoughts. An ontology is au iuventory of lvhat cxists (is thcre). In his orvn pcculiar n'ay and for his orvn peculiar purposes,an ontologist, therefolc, describes the rvorld. I{is description is inadequate unless it accounts by what is for what is (phenomenologically)presentedto us. I shall next call attentiorr to trvo striking features of what is so prescnted. ,Vo ontology, therefore, is adequate unless it accounts for both of them. Tirat is r.hy I shall not make any specificorrtologicalcommitrncnt until these tu-o featureshave been stated and the dialecticsrvhich lrecauseof them euery ontology must face has been exhibited. Call a thought unitary if and only if it has no constituent which is itsclf a thought. All thoughts are unitary. This is onc of the trvo strikirrg featurcs. Call it the unity oJ thought. Replace 'unity' by 'One', 'tliought' by' 'mind' and you l-ill see that the first half of the apholism is already unpacked. Some hold that arvarenessis propositional, i.e., that there are no tlioughts g-hoseintcntion-sare not representedby sentences.If so, then lrc ulilill thoughts rvhoseintentions are representedby simple clau.qcs by is intcntion representcrl rvhose 'l)cter is trLrl'. Take the thought it lrtr tliose of of constituents could bc Irlorrd'.The only tlioughts that there are propcisitionlrl tlren is blond. If arvareness l'<'tcr',of is, and of or further l)ow sometlritrg unity thesis asserts rro such thoughts and the orrly for those thoughts rvlioseintentions are reprcs('Itteclby compound st'ntences.As it happcns, though, the argtrments fot' sut,h thoughts lrt'irrg unitary are the same as tliose for all thotrghts being unitary. 'l'hus, oncc more, we need uot commit oursclvcs.,

I 28

L o c l c A N D ttAr,rry

If the unity of thouglit is a striking (plrt{rrorncrrological) featurr', lvhat nccd is thcrc fol lLlgunicrrtsto support it, rvhlLtpoirrt in arguirrg agaiust it? Thc rlucstiottis rcasonableiudecd. 'l'lic lt'ast.nu,bleansn-err rs that the argumcnts irr support merely clcar up thc' niisunderstandiugs u.hich har.c bt'ctt crLttst'tlby tlie (phenomeuolcigically)inaccurate rvul' in ivhich marry plrilosolrlrcrsarrd psychologistshave dcscrilted tvhat is callcd introspcctiorr or iritrospcctive anal1,'sis.Intlospection presuniably "dcconrl)oscs" thoughts into their constituetrts. \\rhat actualll' occurs u'lrerronc irrtrospectsa thouglit is a seriesof further tlioughts thlt ftrlfills crrrtain conditions. Ts'o such conditions arc. first. that tlic irrtcntion of cach mcmber of the scriesis a constituelt of tlie intention o1 tlrc thought that is being "dccomposed," and, second,that the intcnt,iorrsof a,ll members of the sericsare all the constituents of thc itrtcntion of that thought. A thought's being unitary (a) and its intention liaving no constituents (b) arc tu'o propo-sitionsand not one. Tht' invaiid inference from (a) to (b) is onc of the pivots of the trriditiorurl dialectic. Of that more prescntly. The itn alid infercnce from the ncgation of (b) to thc negation of (a) is thc main sourcc of the misunderrstandings that may leakcn onc's grasp of the unit-v-of thought. Thc premiss as wcll as the allegedconclusionof an inr-alid infercrrcemay bc cithcr true or falsc. (a) is true. Horv about (b)? Onc's ans\\'crdepends on his grasp of thc sccondfeature. Unless somc intcntions rvcre mcntal, rve could not knon- that therc arc minds, just as rve could not knou'that thcre are nonmelltal entitics unlcss sorne intentions rvere nonmental. The differcnce betl'cen thc mcntal llnd tlie nonmental is itself (phenomenologically)prcsented to us. Call thc part of the l.orld rvhich is nonmental the truncated rvorld. The idcalisl,sclrirn that the truncated u'orld does not exist. The sccond featulrr strilics us most forcefully iu nonmenta,lintentious. As to n'hether it is tlso ri leature of minds rvc need not as yet commit ourselves.I shall tlrcrel'orccall attention to it by attcnding to some nollpresentedlo zs. And I mental entitics as thcrl arc (phenomenologically) shall save rvorrls by ornitting tlic italicized phrasc. That the omissiotr does not prejudgc arrvthirrg rvill soon be clear beyond doubt. Take a spot n'hich is rcd and round. The spot is an entity; its shapc (round) and its color (rcd) arc trvo others. The latter (rcd, round) arc constituentsof, or as I shall also say, they are t'in" thc former (the spot). An entity rvhich has constituents is com,pleror a complex. If you challenge any of this, I ansrver that thc example shoivs horv irr (complex' ontology I use 'coustituent' and and that therefore there is notliing to be challenged. Take tivo spots; one red and round, the other grcen and square.Rcd

ON TOLOGIC A I, A I,TE R N A TIY E S

I29

rind s(luarc llntl round are,,tiecl" togetherr.So are grecn and squarc.licd spots and two are round' That is I'hy there :rrc not. Nor are gr""n "nd are $'hat they are. That shou'strvo things. not four and rvhy these trvo .Ihere is a sort of entities rvhich are constituents of others and /,,irsl' the yet so "independent" that a complex is more than, as one sayst is of This things' S,,- oI. class of them. Entities of this sort I here call the more' something The ('oursea very specialuse of the \\'ord' /Second' ,,tie,, rvirich makes a complex out of the class,must have an ontologhave an ical ground. Everything except samenessand diversity must undernot does ontoiogical g.o.r.i. One rvho does not understand that are *'hich entitics str.il ihe ta,sk and nature of ontology. The entity or tlie ground of the "tie" I call sribsislenls' :].nncated, rvorld os it is (phenomenologically)presented'to us f n tne therearemari ythings. Thisist hesecondst r ikingf eat ur e'I f it r r 'er enot already have for the italicized rvord and the italicizecl phrase, rve rvould It rvill Nlany. is unpackeclthe secondhalf of thc apliorism:Thc rvorld comments fottr for be better if before finishing this job I iltcrnrpt yct help to rvhich, although they are badly in llced of cxpartsiotl,may avoid puzzlement' ..ties,,,but they are all .,depcndcnt'' in the 1. Not all subsistentsare ((independent." since the distinction corsense in rvhich things are and syncate' responds to the traditional one bet\1'een categorematic traditional the gorematic,I shall also use these t$'o \\'ords e\'en though entities the <listinction is among 'words rather than, like mine, among are there that rvhich I hold these rvords represent' That is not to say trvo(ormore)modesof exist ir zg. , Exist 'isunivocalindeed. ot her . r vise I,forone,donotk nor r , r vhat it m eanst oexist . But t her ear esever al or modes kinds of existenls and the differences among the highest kinds are\' erygreati rrdeed. Cat egor em r t icandsynclt egor em at iccr et lvo modes. (except' as the 2. A thing that has no cotrstituent rvhich is a thing or a simple' mathematiciatrs speak, trivially itself) is called sfm'ple or not whether What has been said so far does not at all depend on therearesi mpl e s. Am ongsubsist ent st hedist inct ionsim ple- com plex need fur
I30

Locrc AND nEALrry

ordinarily do. Thc cartseof tltc rnvkn'linlrrcssis tlrlrt lirr.vr,lassificatiorr ol com plex es ,c . g. , i u t' f' l t(,ts "l r.n d ,,c c l rn p l c xtl rf rgs,' , I.crl ui l cs .qomc()1. those specific ontologirrrrlrrolllniilpptrts rr,hicli l ls \.r,t ir.ish to ar.oitl. In an ont ology t h ri t l rrl rrri tss i mp l e s , fo r i .s t a.cr' ,.rc coul d rcs(,r.\.(, 'thi'g' for sirnpk's urrrl clir.idc all ,,inclepc.cic't,, crrtities irto t*, kinds, things arrtl r,onrplt'xcs. 4. l'{inrls arc irr tlrrr rv.rlcl, of course. Are t}rey, too, }Ia'r,? Lcss aphoristictllr', r[rc rLctsr,omplex?r har.e not as I'et comnritted m1.scl1. Rut rve rrrrrl.r'stlr,rrd already n-hy, e/ acts are complex, this feature irrr_ posesitst'lf k'ss forrrcfully in their case.Thouglits are unitary, i.e., thcr. h A l' c r r , r ' or r s t itu e n tsrrh i c h a re th o u g h ts . That i s not thc sarn.:,* bei.g tlrirrgsand ircing simple. Yct the ideas are closc.\-or is tller. arrr. dorrlrt thrt cr.en if its thought is merely olle among the co,.stituenrs()l lu' li,t,t,it imposes itself so forcefully that tiic others a'e casil;. or-cr._ l o ok et l. 'llic task of ontology is to acco.'t for r-hat is (phcnomerrorogicaill.) prcscnted to us. If one holds that an-arcrcssis propositionll, he must also hold that all ('onme'tal) i'tcntions arc conipicxes. Er-er)-orr. agreesthat somc are. lfvery o'tologist must account for this manifoltl (complexity). The realist2accounts for it by a correspondins manifoltl i n ilr c t r ur r c r t c d n -o rk l o i * ' h i c h , rrn l i k c tl ro i tl erl i .t,' r," ,.r" i n,, i n,,t i r exists.At lcast, that is for him -qtructurallythc obvious t-ay to account f'r llr . s oc onds t ri k i n g fc l tu re . -\o r d o I i ' o * of nrrr.rrti crrl rtl orcal i ._ tic2 ..tologist rT-hohas fou'd a'other rvav. Trrat f'lly unpacks the second half of tho aphorism. Thc s'orlrl is liarro,. 'l'hc lrrsi. rlirilcctic that co'trols the reali-"m-idealism issrrehas three centers.'l'ho irter:lctions among these centers clctermine the ontological altcrrlll,ivcs rvliich arc or secm available to us. trIost orrtologies,rvlrcther rcalisticz or icloalistic,recognizethat thcr.r, a re t hir r gs ,lr r ' 1lr cy rtrr' rrtaol r n o n me n ta l o r e i t her. \-crv fel .recogni ze th e ont r ilc igir r : sr lt u trrs .f th e s u b s i s tc n ts ,i .c ., they do not recog.i ze that the syrrcritt'go.errrrticrr:rnisroprese't entities.The o.11,tri-o rccerrt exccptions from tJris rrlrrost rrnivcr-qalneglcet arc Fregc ancl the early Husserl' i'e., the arrtlror of the Logische (.tntersttchLtmgyeiz. That is indeed one vcry major roaso' for my admiration of these tu-o thinkers. Characteristically they arc both rcalist-q:.Ontologics rvhich do not reeognize the syncategorcmatic cntities r eall reistic. For an or.rtology to be realisticr is one thing; to be reistic is quite another thing. A reistic ontology, $'e kno\r., eannot be adeq'ate. It eannot er-er acco.rrt for the truncated rvorld. This impo-ssibirityis the.frsr center.The unity of thought must be adequately accounted for. That task is the second center. To spot the third, remember the old idea of acrequation, orlc-

oN ToLoc IC A L

A LTE IIN A TIv E s

131

rltrctiorei et intellectus.If mind (thought) is one and the (truncatcd) ir-orld is Xlany, horv can a unitary thought be adequate to a complex intention? This clifficulty, or apparent difficulty, of reconciling thc L)ne ancl the llany in the knorvledge situation is the th'ird center' A realisticzontologycannot be adequateunlessit fulfills threeconditions. It mtLstnot be reistic. It must accountfor the unity of thought. It must resolt'ethe ctifficutty of adequatiorl' Or, to hark back to rvhat has been saiclearlier, the realistz must shorv that in his 1vorld the "connection" betrveen a unitary thought and the complex it intends is so "close" that hc can defend against all dialectical attacks the proposition $'hich he must liold lest his realism: remain an empty husk, namely, that mincls may knorv tvhat is nonmental. That unpacks and thereby gets ricl of tliat olcl phrase, ctderltatiorei et intellecltrs. (Lct me point out, in pareuthesis,sincc at this point I neither need nor rvish to make any specificoltolrtgical commitmcttt, thtt as far as the third center is cotrcemed it rnakesno dialcr,ticaldill'crcnccu'hcthcr thc complex intentirln that tlie crucial is rneptal or lonmcntal. 't'hat is u'liy I trn t'olivitt<,tttl tlic rcalismtliat arrcl act tlrt-9f assay task is an aderluate oltologit'al it dtlcs rlot lie that lclst, at or, shallou'. as szrchis rather idcalism i-q-quc years or hundrcd last threc thc lie during as decp as it ri-asthought to to be has still rvhich idealism tou-ard sinee the structural drift .o, """I' got its start for reasonsthat rvill be touched briefly at the very -stopped end of tliis essa5'.) Let us look at the alternatives open to one $'ho does not kno$'holv to meet all three conditions. An ontology u.hich is reistic &s, alas, almost all are, cannot even adequately account for the truncated rvorld. This inadequacy and its source nced not bc clearly seen in order to be morc or less strongly felt as a difficulty. Its source, r'veknorv, is the .sccondfeature, l.hich, as we also knol, imposes itself more forcefully in the truncated trorld. That shorvs horv one may be temptcd to choosethe idcalistic alternative. If one shrinks arvay from the absurdity of idealism he still has t\vo alternatives left. He may relieYe the pressure from the third ccntcr by opting for materialism. If one also shrinks arvay from the allsurdity of materialism he has ouly one alternative left. He may relicvc the pressure from the secondcenter by doing violence to the fcaturc that is its source, making thought complex. That also removcs the prcssure from the ihird center (adequation). Thus, if he remairrs inscnsitive to the subtler pressurc from the first center (inadequacy of all reistic ontologies), he may be content. This is the choice or, at least, it is the tendu. I'ell as the basic rveakness of the Rriiish succession from ".r.y Locke to Russell. In Locke and Berkeley it remains a tendency; they at

132

Locrc A\D rtr,tAr,rr.l.

least still recognize thc act. Ilrrrnc n'rries the .hoicc. Iror him, all mental entities are mosaicsof scnscrllta. Litt'nrlly, the phenomenalisrrr u'hich so frequently also appcars in this su<,ct'syion is of coursca kincl of idealism. structurall-v, tlrough, as u'cll as in 1lur.or,it is nrere often tharr not materialistic. Ilusst'll oscillatcd betlecrr tlicsc trvo cclually uuat_ tractivc alter.ati'cs, phc'omenalism and matcrialis'r, all through his career.6 I am ready to statc r'y o\I,n ontological commitments and to sho*, that they mect tlie tlrrcc conditions every adequate realistic2ontolog-vmust mc.t. ll.t it rvill be better if first I say rvhat in ontology I mea' (and rvhat I bclicve most philosophersha'e meant) by ,unirrcrsal, ancl 'individual', arrd, also, hon I use the tivo labels ,realisml, and ,nomi_ nalisnt'. ll'al
ONTOI,OGICAL

ALTERNATIYES

133

numerically difthey hold, differ qualitatively; individuals are merely are callcd bore individuals such ferent. In the Anglo-Saxon tradition rccogimplicitly particulars bare partictLlars. Ontologists rvho accept they and they that i'e'' primary, are nize that samenessand cliversity explicitly_recogx'ho An ontologist ground. alone neeclno ontological first faced rvhen r.rizesthat f"ces ,qu"r"ly tlie striking feature Aristotle that notion are of The difficulties he introciucecl his notion of matter. terms'with the to come must notorious. Yet, every adequate ontology feature. are the Bare particulars are one extreme, individual substances Aristothe of those particularly other. In most classical ontologies, and Substanccs substances' are telian-Thomistic variety, inclividuals (as l used both are "independent" universals are both things, i.e., they ontological differthis ambiguous rvord), but there is nevcrthelcss &n $'ord) makes the the of sense (in anotlicr ence bet$ een thcm ,thi.n ,,more inclepenclent"than thc lattcr. clcarly that depressesthe former individuals' ontological status of univcrsals as conlparctl rvith prtlblcm of accounting tlic ignored Feriontotogists have complctely trvo spots of of thc that as cases for "samenessin diversity" in such rather problem the took though, exactly the same color. Quite a ferv, not epistemology' I'ith concerned lightly. Perhaps they l-ere too psychological tell these of Some concerned with ontology' ",io.tgttrvhich may or may not be more or Iess true but are completely stories patently inadeirrelevant to the ontological problem. To call only such rn'ord' I rather good a of rvaste a quate assaysnominalistic seemsto me speak in such casesof dead-end nominalism' Ical l arrontologyr ealist iqif andont yif it sindividualsandit suniontological ver-qalsare both things and there are only two fundamental that indiother' the one; differencesbetrvecn In"-;one, the obvious vi dual sareonlynum er icallydif f er ent ( bar e) 'TTocallallot her ont oloprobably is idiogiesnominalistic may seem and, as things norv stand, as I do, that insist, to Jyncratic. But it is anything but idiosyncratic as universals of status any ontology rvhich dep.essesthe ontological f:r'milo'rc Evcry tendency' .o|put",l ,tittt ir-tairria,,u1.i. nomincrlistic'in that' In iar n ith the strttctural history of the dialectic rvill apprcc'iatc ontol ogi esrvhichm akeuniver salssyncat egor em at icent it iest hcnom ijust one step nalistic tendency is as pronounced as it can be, stops short of the dead end' TThereis the prinrulaciepossibilityof realisticrontologicsin q'hich the only difficultiesthat lead differenceis the obviousone.But they encounterstructrtral qui ckl y to catas troP he'

134

Locrc AND REALrry

In my *'orld there are i.dir.icl.als a^cl uni'crsals (characters). All individuals are simplc ltttd blrt'c. cliuractcr is either sinple or (,orr'\ plcx. In all other rcspccts,cxt,r'ptfor i,lrcolrr.iousdiffcrelce, ildir-iduals and charactcrs ,.re lLlikt'.'l'lrat rnllkcs rrrc ir lclrlistl. Iri rnr,'g-orld the1. a re als o s ubs is t c rrts\o . t l rl i rrg r.r' i s ti r., rrrv ri n tol ogy fLrl fi l l stl i e secol d o f the t hr ee c or it l i ti o l ts< :r' t' r' :rtl .1 ' t' rl rrrL ltrr r' :rl i s ti < ,,/ontokrgy must ful fi l l . fts f undam c r r t ul l i rr i s t:l c 4 1 tL i l i r:a l i tttL . )r,tt ,r i s bl o1d, i t i s repre_ l \,1 sented by (ir'. I)cl,r'r"slrt'irrg brorrd is u .omplcx of the kind callecl a fact. ('l'lv: .tlrc. liirrrl rL.c thc co'rplex cliaracters. .rheir ontcilogicai status, tli..glr, is dcrivative. Thus rve can safell, ig^ore thcr.' 'r.r.l.y in this t'ss:r,.y.) "Lr" tlic fact representedby ,peter is blonJa'd lrary is tall' thc'. arc tn'. othcrs, peter's bcing blond ancl rlary,s being tall, as .''cll a,s thc subsistent, represented by 'arid', rvhich iics tliem tog ctlic r . , \ r r d s o or r . T h e i d e a i s c l c a r, I tru s t. \ry * -orl d i s' ot..ato' r_ i sti ." ; it is ' ot jus t a c l a s so r c o l l e c ti o ' o f d i s j oi ' ed enti ti es.Il ather, it is completely structured. (A Begrifsschrift or ideai schema (language) reveals the expricit or implicit ontology of its author. Although I shail presently i'troduce a ferv abbreviations, r shall not in this essay usc a schema of my orv'. But it rvill hclp to bri'g out an important poirrt if rve consider *.hat i' such a schema the transcription of 'Tiiis spot is red aril rourid, rvould bo. If 'n', trcd', and 'round' are macie to represent trie indir-idual and the trvo characters "in" the spot, respectir.el],,then thc tlanscription reads 'o is recland a is round'. Thus it rooks very much like the scntence rvliich it trarrscribeseven though ,o,'does not rcprcsent the spot but, r,tirer, thc i'dividual "in" it. The spot itserf, in this rrorid, is ihe fact repres.rrtcclby the scntence!That is the point r-hic,hthe schema help-r to b r ing or r t . ) In my rv'rltl ri.ts a.e very similar to spots. Take a case of my pel.ceiving tlic Iu,r:t/), rrlrrncly,that Peter is biond. In my rrorld there are thoughts, e.g., tlrr'-th.rrght-that-pcter-is-bloncl (rrrJ. An act is a' indir.idual cxcr'plil'yi,g trvo simplc characters; o'e is a species (perceiving, believi'g, rc'ier'be.ing, and so on); the othcr is the thought "in" the act. The l.t irr tluestiorr,for iri.qtance,is the fact u.hich (in my schema) is represcntcd by,b is pcrccir.ing and b is tpr', rvith Lrrepre_ serting thc indi'idual "ir" it.8 s'ch (mcntal) individuals I call arvare8'b i s p e r c e i v i n g ' i s . o t ve r v- id io m :r tic. \ve s' o .ld r a ther sry'b i s o percci vi rg, arrd'b i s a t h o r r g h t - t h at- Pcte r - is- b lo .cl' , i,st a s \1 ,es&y ,peter i s o me', rather t h: rn' P e t e r i s m a n ' e v e n th o r g h *' e ilr so sa y' p e te r is brond,. If percei vi 'g, thet hough t - t h a t - P , m a n , a n d b lo n d a r e fo u r ch a r a cte r s, as i n my rvorl d the;. a,rc, t hen s. c h i d i o m a t i c s t ra in o r a ' r ,k*' a r d n e ss is ir r e le va n t. N or coul d a schema be what it i s s r . r p p o s e dt o b e if it co n fo r m e d in a ll co n te xts to the i di oms of thi s or that language.

oNToLocICAl AL'r'DltNATlvIts

135

ncsses.tpl i5 -simple.Atl thoughtsare simple characters.'l'lrtrsI ltrct'ottttt for tlie unitl'9f thoug[t, meet the first conditiol cvcl'y lidtrtlttllt,tl rcali'qticontcllogl' must meet. Notice, too' that since thougltts art'tttriver-*als,1'ou and I may literally have thc samc thought althorrgh a, thought is a mental crrtity and altliough of the tn'o individuals n'hit,lr cxemplify it rvhen 1'ou and I botli have it one is in my mind and orrc is i rr yours.e tPl MP is a formal In my rvorlcl there is a subsistcrtt',)[, such that tme&ns'&s we sometimesuse it in such fact. ',lI' transcribcs the $'ord scrrtences as'the-thouglit-that-Petcr-is-blond means (intends, is alior.rt)Petcl is blond'. ,\ fact is formal or a fact of (in) the rvorld's fonn if atid only if the serrtenccreprescntiDg it is analytic.l0 Tlirough tlre tlroughts "in" them, acts are by the subsistent I call llf "aonncctcd,, rvith thcir intentions. That immcdiately raisesthrce cluestions. l'lrsl. Is the "conncctioD" sufficictttly "close" to fulfill the thild of the three conditions? Orre only has to consider such complex facts a,s 1r-and-Qand P-or-Q iD order to realizc that some subsistentsestablish connections rvliich are \-ery loose indced. (If thety n'ere closer, fewer oDtologiesrvoulclbe rcistic.) Thus the clucstionis vcry reasonable.All is accourltedfor by tPl MP I can say here is that the recluireclcloseness being a fact in the rvolld's form. Second. The-thought-that-Caesar-rvas-murdercd and the-thoughtthat-Calpurnia's-husband-rvas-murdcredare ttr,o, not one. Yct they seetnlo mean (l'11)the "same" fact. Can trvo thoughts mean a single facrt? The affirmative answer bogs dorvn in difficulties rvhich alc insuperable.This is thc logical problem of intentiorrality;or, rather, it is the logical aspectof the problem of intentionality.ll since it can be kcpt out, I shall keep it out of this essay,merely drop a hirrt that u'ill come e Frege tri ed very hard to ac c ount for thi s pi ec e of "real i s ti c " c ommon s 0ns e. That i s i ndeed a maj or i ntel l ec tual moti v e for hi s i nv enti ng thos e nortrnental enti ti es he cal l s sens es (S i nz ) and rv hi c h therefore, rev c al i ngl l ' c v en thougl t mos t mi sl eadi ngl 5', he al s o c al l s thoughts . r0 Thi s i s merel 1 . a c onv eni ent hi nt. Les t i t be mi s l eadi ng, I ac l d that thr: b:r.s i t: i rl ea, (the $'orl cl 's f orm) i s ontol ogi c al , not l ogi c al (anal 1' ti c i tl ' ). -l rv orl tl ' s form i s rvhat i t i s becaus e i ts s ubs i s tents are w hat thc y are. Our noti ort of tnrr.l y ti c i ty i s gror.rnded i n the form of the onl y l v orl d rv c k now . 11It i s a l neasur e of Frege' s greatnes s thtt he rv as (as far l rs T k rrol v ) not onl y thc fi rst *'ho ci earl y s an the l ogi c al probl en of i rrtenti ontl i tv brrt thet ht: al s o rl rrl i zed i ts ontol ogi c al i mport. That prov i ded another nl rj or i rrtc l l ec trtal moti v e for hi s eventr.ral l y hy pos tati z i ng s ens es (thoughts ) as nontntrtttrl enti ti es - That he mi ssed the soi uti on, i n s pi tc of thi s hy pos tati z ati on arrd ev en though he rec ogni zerl that some s ubs i s tents ex i s t, i s a meas urc of hi s fai l urc . H i s fear rv as that lrc rvoultl ha,ve to give up ali definitions. How typic:r.lly a ma,thematician's fear! sce ai so x[eani rtg antJ E ri s tenc e, p. 217, l i nrl tho t.nrc i al pas s age from "B egri ff

136

oNToLocICAr,ALTERNATTvES

Locrc AND REALrry

in handy soon. The key to the solution is the recognition that the sulrsistentsexist. For, if thcy exist, then trvo facts (complexes)are literalll the same, i.e., one and not tu.o if and only if (l) tlic simples in thcrrr arc the "same" and (2) tliese simples are ,,ticd', t. t'uch other in tlr. same way. Upon tliis strict use of ,same', P a.d e ntay tlierefore lr. two facts and rrot litcrally one even if P-if-and-orl1'-if-e is a fact in tlrr, wor ld' s f or m ( i. c ., i f ' ,{ ' i f a n d o n l y i f Q ' i s a n al l ' tl r,). Third. Wha,t if P does not exist? Xlore preciscly, l,iurr.can tlrl nI p lta a fact if therc is no fact for IPI to intend?1!oncc more, ilie ansrvc. depends orr ,41 being a subsistent. For P-or-Q to be a fact it suffices that cither 1' exists or Q exists. This could not be so if or \\ ere rrot u subsistent but a relational universal, such as, s&y, being to the left of, u'hich is a thing. This book being to the left of somethirig else, fol instance, is a fact only if that something else exists. Not so, rve just sarv, for or. As for or, so for III . (In my rvorld a fact that i-sn,t there yet exists, though only in the mode of possibitity, rvhich is of course the lorvestontological status of all. I call such "nonexistent" facts p-facts.r2 Nor is that an ad hoc construction. But I cannot hcre pursue this matter.) Succinct as it is, almost desperately so, this sketch of an ontology rvill do as a foil for u'hat must be said about Frege before I can intelligently attend to what Dr. Egidi says about him, cxcept that thcr sketch could not even serve this purpose n'ithout sonrc indications as to the ontological assaysof judgment and of truth rvhir,l-rit implies. A judgment is qn act. It rvill keep out issrrcsthat can be kept out of this essayrvithout prcjudging anythi.g that rvill havc to bc said if rve take it for granted that thc species"i'" an act of judgme't is believing. That makcs for u threcfolclrlistinction: (cv)

P

tp1

c(p).

P is the fact intcndc4. r7;l is the thought ,,in,, the judgment. G(p) is the judgment itself. P is ncvcr a constituent of.G(P).If P is nonmental, which is the only cosc \\'(l rrccd consider,that is obvious. tpl is merely one constituent of G(P). 'l'hc other tu.o are the specicsbelicr-ing and the arvarenessrvhich "intlit'iduates" the act, e.g., my judgment non, that P, yours tomorrol, mine yesterday. 'Truth', or, rather, ,true, has und Gegenstand"which is there q.oted (p.46 of the Black-Geachtranslation). structurally the deepestroot of this failure is that in his rvorld there are no facts (eomplexes).See below. 12P obviously existsor doesn't exist dependingon whether the sentence,p, is true or false. It is by now equally obvious, f trust, why at this point I avoid 'true' and 'false'. The root of all matters philosophical, including logic, is ontology.

I'

137

(at least) four uscs;one is primary; the secondderivesfrom the first; the third from the second;the fourth from the third. In the pfyygla use, 'true' and 'false' are predicated of thoughts, rvhich are characters. A character is truc if and only if there is an entity such that the character means (M) it and it is a fact. A character is false if and only if there is an entity such that the character means it and its negationl3is a fact. This assay has trvo consequences.First. True aniLfalse, as representedby the primarA use of 'true' and'false', are they are subsistentsof the kind some call subsislenfs.liIore -specifically, defined logical characters.laother entities of this kind are the integers, integer itself, transitir-ity, reflexivity, and so on. sutrsislents are neither tPl bcing true and (2) P itself are mental nlr n()nntenlal.Secorrd.(1) trvo facts, not one. That i-qagain obYious,if only because(1) doesr'vhile (2) cloesnot contain the cortstituent 11'1.Nor is it a source of difficulti es that the sent ence't Pl is t r uc if and only if P'is analyt ic. Just remcmber the hint of tvliich it rvas said a rnoment ago that it would come in handy soon. In their secondaryuse, 'true' aud 'false' are predicatcd of judgments. A judgment is true (false) if and only if thc thought "in" it is true (false). An assertion is a kind of linguistic gesture. Typically, it involves the utterance of a sentence,Typically, it communicatesa judgment of the one nho makes the asser t ion.I n t heir t pr t igt y use'( t r ue'andt f alse'ar e predicated of assertions. An assertion is true (false) if and only if the judgment it communicates is true (false). The sentenceinvolved in an assertion representsthe fact the judgment communicated intends. What a sentenceof a natural language represents depends on the context in $'hich it is uttered. This is not so for the schemata calied ideal languages.The fact representedby a sentence of an ideal language (Begriffsschrzf) is completely determined by the sentenceitself.rsIn their fourth use, 'true' and 'false' are predicated r3 N egati on bei ng a s ubs i s tent, ,P ' and ' not-P ' both repres ent enti ti es , one a fact, one a p-fact. r{ ,D efi ned'suggcs ts c ompl ex i tS ' . If one hol ds , as I do, that among s ubs i s tents the di chotoml ' si mp l e-c ompl ex mak es no s ens e, then the w ord i s mi s l c adi ng' S o, si nce characters are thi ngs , i s ' c harac ter' . r5 To real i ze the dependenc e of natural l angr.tageson eontex t, c ons i der (1) ' It (2)' I am c ol d' . W hat (1) repres c nts depc nds on w hen and i s col d today'and w here i t i s asscrted ; i l -hat (2) repres c nts , on rv hen and by rv hom i t i s as s erted. The i ndependence, i n thi s s ens e, of s c hemata c al l ed i dc al l engrrrges i s one of the rl rtl i ctl di fl erences betw een thc m and natrrral l angttagc s . B ec atts e of thi s di ffcrcnce, those schem ata c oul d not ev en i n pri nc i pl e be tts c d for c ommuni c ati on. A l so, thi s di fference unpac k s part of the metaphor that an i deal l anguage i s , or purports to be, a pi c ture of the w orl d.

138

L OGIC AND NEAL ITY

of the sentenccsof ideal languages.such a scntence is true (false)if and only if .n.hat it represcnts is a fact (p-fact). iI some of thc terrn.sl"rege choseare very a*-kri'ard. l'rolr:rl,ly he sought for rvords tr ser'c him as \I'eaponsin his lifc-long str.'gqr. rtluirst psy-chologism. For irrstance,hc calls "thouglrts" entities l-hicrrilie himself strenuously insists &re nonmental and l'hich one l'oultl thcrefore much ratlrcr call thc (potc'tial) intentions of tliouglits (or of judgments, or of acts in general). under the circumstanccsI sliall continue to use my o\vn *'ords rvith the mea^ings I have given to thcnr, make a spceial point of ayoiding his use of 'thought'. It wili bc converrient, t hot r gh, oc c a s i o n a l l y to rc p l a c e ' n i e n ta l -n onmcntai , l rr.,sul rj ecti t-eobjecti'e', *.liich is the dichotonry he happened to p.cfer, probably becausche fclt that it stressedhi-soppo.ilion to psl,cliologisni. Begri.ffsschrift appearcd sliortly before 1880, Functiort tLrtcrBegri.ff, LIebersinn und Bed,eutung,and ueber Begrif ttncl Gegettstanrlshortl-v aftcr 1890. The threc later es,qayspresent au cxpricit and rathcr detailed ontology of the objectivc (truncated) *'orlcl. Alrrut minds as suc,hthey tell us nothing. They merely specifl' the objct,tive entitics l'hich arc the (potential) intentions of subjectilc a<,ts.Nor cloesthe earlier essaycontain an ontology of mind. Yct it nrlkcs irr $2 arr irrrportant contribution to the or-rtologyof juclgmcnt. ll'o tlris t,ontribution, as far as I ktrorr',none of lrrege'-qlatcr I'r.iti1gs lrlrls:r11.tlrirrg.Some evel blur it . I ' c r ha p s th a t i s rv l i y th a t c l i rl v c .nl r' i l rrrti .' has been some$'hat neglcr,tccl.It is a great rnt.rit of I)r. l,)girlit0 luiYe callcclattentiol t o it . I s hlll r re x t rl c s .ri l rctl ri s l l rrl t,i :rlo rrto l ogyof j udgment, then the et ' c t t t t t lilot t lt t l o g .lo,'[ 1 Irco l ri c c ti r-rrtl o l l rl . T he hc lr r t of $ ? i s l r tl i s { i rrr' l i o trl rn o n g thrcc enti ti es,reprc.scnte{b1' (P)

l',

-P,

lp,

re-rpectivcly.If y.rr ('or)r[)rrrc(B) l'itli (a) abor-c, 1.ou n'ill be able to guc.ssr,hat I tak. t. bc tlr<'natural reading of the paragraph. Fp is t he judgm ent , th c e rrti tv I c a l l G (P ); -p i s the enti tl . I cal l [pl , i .e., t he t hought " i n " th c j rrrl g rrc rt;1 , i s i ts i ' tenti on, or, as Frege herc calls it, its "content." A strirrg of six comme'ts *ill support thi,qreading and prepare the grountl for what follori--q. 1. Frcge here says nothing allout.w'hethcr p is simple or complcx. Nor does anything hc says depcnd on that. This is not to cle'y, tliough, that in his eventual ontology (of thc objective l'orld) all (pote'tial;

O\TOI,OGIC .4,L A LTE It\A TIY E S

139

ideas" (F2)'to intcntions are simple. 2. -P is called a "complcx of hatl not (b) I'rcge That shorvsthree things. (a) -P is a mental entity' intenthe as yet hypostatizecl thoughts into objective entities to serve yerbally completcly not at least, he $,as still iion. of judgments. (c) thoughts complex. 3. - P is not the makes that tradition the from free judgment itself but, rather, that constituent of it which also occurs in of merely errtertaining P n'ithout either believing or disbelieving or-, tPl.4. of the horizontal stroke in -P it is ".t that -P is it. That _"horvs ,,combincs the symbols following into a rvhole" (F2). Liiit -qaiclthat u'ell-formed crally that does not make sense' For one, the symbols of a another, For are a rvhole. Thus the stroke $-ould be redundant' -"errtence out of scnsc this ivhole is a nonmeutal cntity. The only rvay of making of unitv the of thc passagcis to reaclit, rvith Dr. IJgidi, as an assertion fr.m contrary the thoright. ,q,,reacl, it fa| out*'eighs t6c cvidc'ce to after all hol' 2(c).br so at lcrL-.tit scc'rs to rric. O'c nrust not forgct artcurately' things r.crl. clifirr,ultit u-tLsin 1871)t() spcllli lilrt)ttt tltcsc -P is not th:r,t ir. l,lvt'rrtlrorrglr l''l't'gcirlsists Nor is it r.-ct'1'r'lls1'torl:ly. At is thc n-l i ol c,, i F1', l, t 't l. , cs r r ot t t 'll t t s u'lt r Ll;clst '1'lr t 'r t l "it t " ll) ' lllur. a beulme u;ill tt, blanl;. Iitctl'ttl.illy l|rc lian|,: this p|r.Lcelte Leat,es -1, cntitic's,not onc. Nor is thc|c anytliing to indit\-o arc P and 6. ((irl"-P. That shorvs that the contribu,tion 'is real'ist'ic2, cate tliat P is or, at least, tliat it is compatible rvith rcalism:' is of In the eventual ontolofy (of thc objecti'e rvorld) each entity are former The orre of tir-o kinds. It is either a Gegenstandot a Function. ,,independent" iu exactll' the samc senscin $'hich (in my \l'orld) things mathematare. That makes them lhings. Frege's notion of function is pair of (or ordcredan ical. A function projects or maps one thing on things, etc.) on arlothe, thing. That makes a function "dependent" in sense same the the Ihings mapped and mapped upon in exactly on the ii.hich (ii my world) the subsistents called ties are "dependent" need a things iandiacts) they tie into complexes'Nor does a function funcmakes That further tie to tie it to rvhat it maps and maps upon' tions syncalegorematic entities. ate t'he Things are of three kinds. The only members of the lrsl kincJ (F)' The Fal'se two trulh values, the thing True (T) and the thing sensc-lhal-Peter-isthings of the secondkind arc all "senses," e.g., the Judgblo'i, the sense-o.f-Caesar,tlie sense-o/-Calpurnia's-husband' names t4ot9 scnses mcnt is propo-sitional,of course. Hence, only thc judgments. But Frege is contairr thol' o." the (potential) intentions of t6t(F2"refers topage2oftheB l ac k -Geac htrans l ati on' R utl hav etheGerman for 'Ideenuerbindung'' text of Beorrfsschrzlt before me. 'Complex of ideas' stands

140

L OGIC AND REAL IT Y

not (as far as I knorv) committcd to the vie*, (*'hich happens to be mine) that all arvarenc.qsis propositional. Thris tiic senscs rvhose names contain 'of' r'oulclbe ancl probably are in his rvorlclthc (potential) intentions of other acts, e.g., of perception. Ail prssilie senses enist.l7The things of tIrc th,irdkincl are cither ordinarv ,,iliings,' such as Peter or Nfary cir a c.l'recl spot; or they are integers, cresses,arr{ so on. For our pllrposcs it rvill be safc to ignore ali but the,,orclirrr.l'l ,rr"*bers of this kiud. (1'he
ON TOLOGIC A L A LTE R N A TIV E S

141

to infcr it frorn the The onll' l'ay, therefore, of arrivitrg at an answer is propose &11ans\\'cr' structure of rvhat \\,e are told. Presently I shall explain.rvhy the shall I though, First, irrgr', things are alt simple' discuss his clichotorny is so furrdam"rlt"I thut one cannot thoroughly ontology rvithout ausn'cring the cluestion' just a class or colllemember rvhat rvas saiJearlier. The rvorld is not I-say and lcction of clisjoineclentities; it is completely structured' focus on the objecmean the rvorld as a s'hole' But lve may once more is an obvious task or tive ivorid. 'I'o acc,ount for the rvorlci's structure a task and problem evcry aclequateontology must solve' Recognizing The adequacy of a tackling it in a ccrtain waJi or ttyl" it one tliing' rvorld (ont.logy) all solutioi proposecl is anotlier thi.g. Co'sicler a facts. That merely rvhosetliiugs are simples ancl all rvltosecomplexesare to -solvethe task by brushcs asicietletails. Irr such a,$'orltl ol)c may try of fttlts (complcxcs)rvhic'hare in makitig thi[gs (simplcs) r,Onstitrrcrrts f t r ct s. 'l'lr isis o' st ylc' As it turn consti tnc I ] t s( ) t ( ) t ht 'r '( nr , , r 't 'r 'or r r lllt 'x) hupp.,' ,,i ti sl l lest ylc
142

L OGIC AND REAL IT Y

occasionallywe speak and think of his "connections,, ,,facts.,, as (This is a recognitior of, as'-ell as ar attcr'pt to or.erconrc, tlrc difficulty of speaking n-ithout distorti.rr about a style radically clill'e.cut from one's o\\'n.) I'Iiregc's *'orld I'ett't''s bci'g bloncl i'volves at lcast f,ur c'tities; (1) t he t hi, g l) c tc r' ; (2 ) th c th i n g T ;o r, i f Peter i s' ot Lrl .url ,tri e thi ng I', but thc dili'r'r'crr.crnukcs clifiereuce for .*-hat ,r. ,,.,,. about; 'o (3) thc fu'ctiorr lrl.rrcl; (a) the thi.g se'se-that-peter-i.-bro,,J.S; (1) on (2). 1'hrLtis tlic- "f,,"ct" in the case,l.hich itself ^op. is not ar, e.tit;,. (1 ) a. d ( 2) u' c ' c ' c rth c l e s s o b j e c ti r.e l y .I.rrere " c o nnected,,b1,(ri ). i s, h'wcl'cr, .r "co.nection" nhatsoever bet*,ec' (J) on the 'lrjcctive orrc lru,ridarrcl(l), (2), (3), on the.ther. Tliat sho*,s th,t Frege,s wortd is rLotcornpletelgstructured. Onc rnay try to remedy the clefectby bringing in trvo mor.e things, (5) the sense-of-pcterand (6) thc sc'sc-of-(being)-broDd.(6) is 'arnely, the scnseof a fu'ctio'. Are there i' Frege,srvorld such ,scuscs? \yhatever expcdie.ts either he himself or his disciplesmay have rcsortcclto, structurally, I believe, the ansrver is No. But *-e ncecl not insist, may evc'for thc sake of the argumc't assume that there are such senses. If s o, t hc n ( a) *- i l t b e ,,c o ' rre c tc c l ,,rv i th (l ), (2), (3) i f and onl y i f the follorving trvo conditions (a) a.d (b) are fr-rrfilled. ia) (;) a.cl (6) are corstitue.ts of (a). o) (b) and (6) are "con.ected', rvith (I) a'd (3), rcspcctivcly. (a) makes (4) a complex. Ifence, if all thirrgs afc ,sin1pls, thc attcrnpt at rernedying thc clefectfails on this grorurd ulone. lrut lct us n'ai'. thlt argument, look at the first harf of (b), i.c., at trre t*-o things I'r'tt'r :urd sense-of-Petcr.There is no objectir-e ,,conncction,, whatsocr..r lr.t*'con thcm. 'ro appreciate tlie gia'ity of the poi't, introdur,c trvo r',Lc thirrgs, llary and tlie se.se-of-lfary. Therc is no "connccti,r" lr.tiveen arry trvo of thcse four thirrgs. \\-iiat, the', one must ask, is tlrc olrjcctivc "fact" that makes the serrse-oi-Ireter the senseof Petcr nrtlrcr tlrarr that of Xlarl, and couvcrscly? I conclude that the things l,'.cgo .rLlls sensesare totally ctisjoinecl froi alt otlter enti_ ties o;f his objectiucurorld. N'r is that surprising. senses, after ali, are merely the (objccrtive)lrr'prstatizations of (subjectir-c) thougrrts.rs Assume next, for thc s,,k. .f thc argume't, tl'rat there is an otjective "connection" betrvecn (.t) orr thc o'e hand ancl at ieast one of the entities (1), (2), (3) on th. oth.r. Thc only likely candiclate (2), is tlie thing T. Assume, then, corrtuu')r to fact, that there is an objectir.e re T h e t r v o m a j o r d ia lo ctica l m o tive s fo r th is h yp o stati zati on are Frege's anti _ ps y c h o l o g i s m a n d his a w&r e n e ss o f th e lo g ie a l p r o bl em of i ntenti onai i ty. s"u f ns . I a n d 1 1 .

ON TOLOGIC A L A LTE R N A TIV E S

t43

,,conncction,,betrvcen (4) and (2), i.e., betu'ecn the setrsc-that-l)ctcr(2), (3) are "conn('(ltcd," is-blontl and T. If this wcre so, since (1), contrcctedwith (1) and (3) andthis rvorld i+i,.-"ufa over (2) also be 'I'o apprccirvoulcl in its orvn pcculiar rvay be completely structured' is tall' the N'Iary if ate horv peculiar that rvay rvould be, considerthat' sense-thatthe Pcter, blond, and "connection" among the ihree entities as that betrveen the three same the exactly be rvould Pcter-is-bloncl el rti ti esP eter, blond, andt hesense- t hat - x{ar y- is- t all. onecouldar gue ,.conncction,,iS tvorsetharr none. That shorvsthe absurdity that such a ofh5' postati zingt het r vosr r bsist ent st r uear r df alseint ot het r vot hings ' f arrd F. dominated bv his Yirtually all studies of Frege start from and are rvithout any refercnce to sentantics.I cleliberately statecl his ontology This is,'rtt to clcrtythe trrttci.l importance of the linguishis scmar-rtics. though' is tic turn or eYcli of iclclLl llttrgrtltgcs.1'hcir impot'tartce, mcthod othcr by arry mcthodological. oncct otrtr l}lis citlrt'r' by this clr be wcll r vill hc cnucl catecian or r t ology, t 'it lr t 'r 'lt is o\ \ t t oI llt t r t t ] t er 's, mcntioncvcn it rvith
L41

L OGIC AND REAL IT Y

a mental ttconnection" betr.ccr such nonmental thi'gs as, sa]., T a'd the sense-that-P cven thougli in his objective ivorlcl tliesc trvo things remain totally disjoined. Sinn und Bedeuturtgc.'tai.s t*-o crucial passages:o (F65, i'78) to the effcct that a,iud.gmentis tlte "adt'ance',from a sense-that to,,ils,' truth ualue.I do rrot quotc tlrc passagcsonllr becauseI coutinuc to use my orvn n'ords. 'Adva,rrcc'llrrd 'its', though, arc Frcge,s. 'l'hey leave no doubt that tlic advarrt'cis held to pror.ide a mental link Lrctn'censuch nonmental tlrings a,s,s&y, 'l and the sense-that-P. Thus, tlrcl,connection" cstlblisliecl is at best subjective. Linfortunaterl', thorrfrr, the very ideu,of tliis advance is irremediably confused. Fregc liirusolf rvas not u'liolly at case. otheru,ise he r-ould not have rvarned (I,'(i5) against mistaking the italicized formula for a definition. The formula.i-sneverthelcss thc heart of his irremediably confu,"cd e'entual oritology of judgmcnt. The blanh of the early contribut'ionhas becomea bltLr.I shail removc the blur by stating the only clear idea of an "adr-ance" in this context. Supposethat three acts occur successivelyin my mind. The species of the first is entertaining (rvithout eithcr believing or cli-sbellsr.ir*; i the specicsof the secondis pcrceiving;that of the third, belin.ing. The intcntion of all three is the same, P. The thought "iu" alr three is the same, IPl. That is mcrely a schcma of course, b.t it rvill scn.e. One starts by entertaining P; orie ends by judging thrt 1, (bclit,r.ingthat 1'). Or is thcre any other *'ay of "adr.ancing', to',arcl a juclgment? In the scherna,, the "advancc" is from an entertaining through a perceiving to a lrclie'i.g; the thoughts and thc intentions of the three acts happen to bc thc sarnc. f say happen partly becausenot all ,,ad'ances', are that simplc; ptrtly becauseI *.ant to repeat that even if thcre rvere such objcctivc thirrgs as the sense-that-Pand T and even if these tn'o things u'crc tlrc irrtcntions of the first and the third act, respectir.ely, the "connectiorr" *'lrirrli the "adr.ance" establishcsbetrvecn these trvo objective thirigs rvrrrkl still be subjective. Nor, alas, is that alr. T is a th i n g. T hus it . o' lt l rrrt i n F re g c ' s * v o rl d b e th e i ntenti on of a j udgment. That alone shon'sthat the confusionis irremediable..\*otice.too. that unlcss a judgme't h's actually occurred, there is no actual but at mo s t a pos s iblct t c o n n c (rti o rr." Since some of its things arc totally disjoineil from all others,Frege's truncated,worlil is not contpletely struct.ured.If , howet:er,one Jollous him in adding to the (nonmental) "connections" auailable in this world some 20 ch a r a c t e r i s t i c a l l y , th e y a r e a lso cn r cia l fo r th e p r o bl em, from rvl i i ch the s.hol e es s ay i s d c v e l o p e d , o f h o w, if ' a ' a n d ' L r ' a r e b o th ,,n ames,,, 'a:a, and'a:b' c an d i f f e r i n " c o g n i t i v e va lu e ." Se e a lso b e lo r v.

ONTOLOGICAI,

ALTERNATIVES

145

othersw hi charem ent alanclt her ef or enr lt auai'lableat allt l, r l/cotnpletel c. sst hcrycar e workl becomes his truncatecJ minds, then, after a f ashion at least' diagbecauseof the blur')'fliis is thc structu'ecl.(I say after u i"'ftio" nosi satrvhi ch' '"hu'"a, r ived. I t suf f icest oident if yt heidealist ict cndjob' gui it tuitt.le bettcr if rve postpone this ency in I''rege's o"t"f"gy' has rviil support and round out u'hat turn lirst to -qome.om""le"ts that been said so {ar. of judgment,.although sound, The early corttribution to the ontology blank became a blur' Can one so rvas yct fragmentary' Eventually the fi l l thi sbl ankthat l'"gt '"t o'lt lo'"r vholebecom escom plet elyst r ucqueslittle about minds as such' the tured? Since he tclt. ui b"t very some yield does, believe I yet the a'-qrvermay, a'd tio' is rather moot. just one sarv' rve As 1' parts' tu-o dialectical insight. )I1' ansu-cr has possible merely aclmit and rvoulcl have to make truth vlilucs itrtt:trtiotrs .,conttccti ol rs. , , Bot lr cnt ct t t llit iot t slr r t llt it ist r t r ct ur al. ( . I - hisisbut t'tlttfttsiotrrtrrluirirrgthcrn is irremediatiothcr rv:ry of s:rfirrg tlirLt tlrc of frilsc.bclicl'.Sin<,othe intcntions of able.) 2. Renrcnrbcrti""ir,.,,r,r."r ,,r,.nrrct.ti.rr',lr(rt*ct'rro,nqa't and its ir-rtensorrreacts clo.ot .rr.i, tr,,, rvorltl thc rli{lirrrrltyis solvcd lly making ti.. c.'.ot l,c a .clatio'. In my exist' cliflicrulty'All possibltr-scriscs it a subsistent (lll' iircg" ftot no are rvhich "contiri'gs .I.hus one coulcl fill the biank by maki.g minds Or' to say the same'thing in nectecl" rvith tlieir intentions by rciations' and an i" his rvorld things rvhich are minds Frege's style, one .".,1;;l; u'hich Frege-rvorld modined functions' The appropriate class of ;ilt is cmendation the counterstructural is the rcsult of this oJlltion and of features' peculiar three Yet it has indeed completely'ti"ttttt'"a' objcctive things remain as subjecatnong "connections" (a) Certain my that P thcn' irrespectively of t'hether tive as before. if,l fiil"age my mind and the maps bclief is true or rorr",'tr-t"-l"licvi.g-function clear that the judgment as thing T on T. That makes it embarrassingly though one'may have the t"ttt"' such is completely disjoined from """tl lattcr' 3' If acts are the from entertaining "adt'anced" to the fot"t"' called _P has no Frege errtity tlre early relational, the,' t]'e,''.''tal is requircd' \\re must place in the system' Thus a further emendation contribution'2l early abandon a most valuable part of the thoughts'But are but hypostatized that senses evidence 'r All this is further with reiresent:itiverealismin the there is here a striking"ii"r*ii."r connection to me by E' B' Allaire'In those ottt pointed styleof, say,Locke,"it*h;;;;n a percept callcdir'ut r-hollvmvthic^al 3ntitv a mind i. 'l;;;;;dito ontologies binary functions I io one of the b1, a relation ,thith ;;;;o"J* 'ttt"tu'ally is that while ell Frcgean intentions difference ;;ly .r".;;:-if" acldcd to Frege's it is be mental entities' The more striking &re nonnental, perceptsare meant to

146

I l i

L0( ;tC

ANI)

rn lrrege's sernanticsall scrrtcnt,r,sand all definitc clcscriptiors are names.'Peter is blond', 1''r i.sta'.c, is a rrame of r or. 1i, cleperiling on whether Petcr is or is rrrt blcl'd. 'calpurnia,s husbarrcl,is another name of Cacsar.2:'l'his is tlrc sccond structural rca_sollfor hcllrling that in Frcge's o't.lrg' lLll tlri,gs are simplc. Si.ce it is scure'tical and since, for a rcaso, tlriLt,lr.s since been explai.ed, I clicl \\.art to '.t introducc l"r.g.'s s..lurti.s before having stated iris orrtol.g1., I dicl not stltc tliis s.<'.rrl l'eusourvlien stating the first. \ ol- I ari, 'reacl1.. Speakirrglls l.t' orrliriurily do, \\-e use ,name, vcr5, brclaclh,.Trr this ccrtur'.y t,lrtrlrlrilosollhcrsl,ho rvcre most influenceciby lircgc rrsr,tlthe rvot'tivcry tt'r,lrrritrlllly,Their use, rvhich is very lt&rl.ow,car.ricsrrr,,fc or l css c lc lL. l. yt lr r . c . o ' n o ta ti o n s . (1 ) A n a m e re p resentsa si r' 1l l c thi rrg. (2) A r.rnc t.lls about the entity it represcrt-s.r3 (il) hi a rvell'.thing <'orrstlrrctcdBegrifsschrift (ideal ia'guage) a name clocs not o.cur tirrk-'sstlie entity it purports to represcnt cxists. Hacl F'csc liinself alu'.ys arrclclearly used 'namc' l'ith all trrescconnotrtio.s, i u,c,ulcl in vicrv of (2) have made my poi't. I do not make so extlcrne ri claim conccrning Frcge's use of the rvord. I merely claim that all thcse conDotations are n)ore or less clea'ly implicit in trre n'ay he uscsit. ,Iust rcmembe' lrorv it puzzled him that 'a:b', *-hcre ,a' arcl ,b, are (,an con'cy any information.2aNor is it just cha.ce that i' '[mc-\, thc postrfrcgcu' dcbate these connotati,.s lrccarne c\-cl' nrorc r,lcar a'cI e xp lir r it . Irr rny *-o.ld, you rvill rcmember, 1'a'cl r-I,r beirg t^rc arc trvo facts, rrot.rrr'. Irr lfrcge's *'orld tlierc arc rlo facts. Litcr.alll', tlicrcforo, he catirr.l,<'itlre' ugrec or disagree.Rut he comesas closc to di-"agreei'g as lie.urr lr.yrrss.r'tirrg(lt6-1)that the se'-se-that-p a.cl tlie "qc.sc-that-pis-tnrc ur'()olr('tlrirrg, not trvo. we ouglrt to bc able to unilelstancl horv h e car r r
r47

ON TOLOGIC A L A LTE R N A TIV E S

III' ALITY

Pctcr on T is sut'lt & "l'lot'" thing on something else.Rloncl mapping t.l tl l i ta.' /,.tl oesrrot r clllyst ut ea; it ism er elyil'llcnI ( 'of . l'. I , et us crirl'oa' 'I'ho serrtcttcc this b-v sa,viug that 'P' correspotrtls to expre-s-s :' represe.ting a. identity tpt t,.lrrg t,ue is 'P: ?', with.' il;di; projecting trvo things cn T if artd functiotl,rs i.e., a binary function P arid -l on". Tirc iclentity function projecting if the,,t$'o""r" olr11,. tcr plcpared I am .I ,,fact.,, call it B. Are a ancl B trvo or one? is a on shall I you disagree are two' If argue that by the logic of "f&cts" thcy cliffrcult it is to think and very how shou' to take 1'our ciisagreemeut tal kaboutarvorl dr r 'it liout f act s( not : , , f act s', ! ) , par t icular lyr vlr er r . one i scomnri ttedtoasclr cnr au. hosesent ct r cesclor r ot evcr r st at e, , f act s, ' lve nccd not argue' l-or I but are nrcrel5' tlames of 'f or F' Irortunately rvitli you' licld a and B to be one''I'hus' also bcliel'e tliat l"regei-agrcecl t}rey are trvo' \\'c can agree rvhether or rrot ]'or.r {rgrco u-ith rno tlut ttttc. NIttrc<.r'cr'it is ob'io's that t'at he ri.artecl t,.,,,rrJ,'t tlit'ir ltt'irrg Irccoul drrotpossillll'lLsst 'I 't t lr lLt llylt sst . r 't iI r gt lt lLt . l, 'llr r d. P: 1.r vhich , liot h lr t t t lr . t , u'o"lr r . t s" lr y Yir t t r cof thl L t *- . ult l ir r r 1ll.t1. .amc.I.. I,.c.rr, s st cld t hat l f ( '( ) }l( '.St l lt t 'iissct . t ,ilr l ,eter i s l l l ol l d l rtt t l }[ lLr . vis t lt ll lLlst lI 'l'lrut ltllirls to orlt'' llfo trrd tlttr st,rrst:-thlLt-1'-is-tl'tl(] the scnsc-tlurt--I', trvo further obsct'r'ltiorls' (P-1r1' ttames'I' of 1| u'hilc First.If. '1'' is fitl-sc,theu it is a name Il ence.1,,attdtp:||, dor iot havet lr cSalnesensc. That slr or r . st hat t lr e s1' rrrmctri cal treat m er r t of Talr clFism er esliar n'Nor ist hat sr t r pr is- by job rvhich in my u'orld is done siuce T (but tloi n; cloesthc ing, ^,,r?rrrptitic"tion *'hich arel not a'd ex.mplificati'u prod*ccs e'titics merely in the rnode of possibility' Considcr thc itrfinite scries Seconcl.

( P:T) ,

( P:7.) :7,

l ( P:D :r l :T ','

',

compare it rvith the familiar all of lvhose members are rveil-formed, classicalregress I knorv that I knorv P' I knon' P, that I knorv P'' ' ' know I knorv that I ' true catrrlot be subsistent tlre for nrar,k the sclrema my ttrrtic.c'thatirr itclrrtctl ri'ithor.rt adding corncrs ; rl , i sr truc, r tp l iq t r uel is t r ue, t t t Pl ie t r uel is t r uel is t r uc, ' evcn identity 25L ,rJcg.rfsscl.rzlr ide.tity hokrsbet*.eennernes,rvhichmakes identity function'Grossobjective an i"ttoatt"es .,tr,:..iit". fvcntu;lli:'F-"tg" ti me of c hange ot"ut' "d maru) argLtcs cortv tl tc rng)y that thi s P SS i tJ ' s l l re tl trr',' grel t

at thc approx i mate

148

Locrc aND REALrry

and you will seethat 'r also does the job of that mental character which r call the species"in" an act of judgment. or are we to infer that only a true judgment is o'tologically a judgment, only a truc belief a belicf? ' These observatio's shorv the absurdity of hypostatizi.g true a'd false into trvo objccti'c things. The next paragraph is stricily necessaryi^ thc ..'tcxt of this essay. yet it is short aritl'ot it at least states the ans\l'crs to some questions that must har.c ariseri in the mind of the readcr. So I shall indulge in the digre_ssion. rr a schcm* that reflects a rvorld like mine no compound expression is a name. A rrarnois a primitive symbol that represe'ts a simple thing. No simple thi'g has more than one name. sameness a'cl diversity thctnselvesare not representedbut mereiy "sho' thcmscl'es,, by the s&mcness and diversity of (types of) expressio's. (Thnt shorvs,once more, that ideal languagesare really ra.guagcs. Also, it u'packs 'ot part of the metaphor that thoy are pictures of their Two ''orlcrs.) things are identical and not one, or, as one says, the same, if a'd onry if wlratever can be said about thc one also can saluat^eritate be said about the other (Leibnizia. identity of indiscer.ible.s). Iclentity is representedin the scherna(Leibniz-Russelldefinition of idcntity). It is, hon'ever, a categorial feature of.thc truncaled *-orld that twosimples are identical. T*'o awarenesses, it seems,can be trvo a'd'o }.et identical (disccrnibility of identicals). Alr this u'packs part of thc aphorism that sumcllessand identity are primary. Notirre, fiDally, that irilrrege,s case I sp.ke of an iderrtity function, not of a sameressfu'ctio.. A sameness fun.ti'rr rrould be a monadic functio'mapping er-cry thi'g on itserf. sulijc.ti'e idealists hold that only minds exist. Upon the commonscnsi.ril usc of 'mental' and 'nonmental', I shall not abandon, '.,hich somc tliings ilrc mertal, some are nonmental. That makes subjective idcalisnr :ibsrrr<1.objcctive idealists hold that all nonmental things are, in a vcr.y spcrrialse.se of the .rvord,,,mental.,, The only rvay to find out rvhat this vcry special se'se is, is to state commonsensically those featurcs .r llllegerl features of the nonmental are held to '.hich ma k e it " m ent al." ' l ' h rrs ' rre c a n c l i s c o v e rtl i e ci i al ecti calcore of obj ective idealism. Its p.op.rr.rts may or may not add that the nonmc'tal (the truncated rvorld) is litcrally a xlind of which our o\\,, minds are "moments." That is mcrt'ly spcc,ulation.so I merely ignore it. But I reject any ontology, rvith or rvithout spccurativeaccretions,rvhose dialectical core is absurd. r'\rc dialecticar core of objecttte ide:alismis the proposition that minds conlribute (create) the structure of the nonmental. That is absurd. Frege'struncated u.orld is not "mental.,, Nor cloeshe claim that it is.

oNToLocrcAL Ar'TDRNATTvES

149

fcrv realistsz That nrakcs liirn a realistr. He is indeed one of the very makes his That partly structured' least at is world who-qetrturcatecl reism creates a realisni stulclier than most others. For the inadequacy of rve discussed the very strotrg pull tolvarcl idealism. That rve salv rvhen on the intcracr,,orr betrveen the three centers of the basic dialectic. some time other hancl, according to the diagnosisat u'hich'we arrived by ago, part of the structure of his truncated rvorld is contributed .I.n"t makes his ontology idealistic in tendency ,ni,ta. nonmental lirege also hypo-statizeclmental things (thoughts) into tendency? in materialistic his ontology make that no"t on". ("."rt."r). which is things, mental are there that denies either A materialist absurcl, or, tvhich is eclually absurd, he "identifies" them'ivith "ordialas, most nary', nonmental things. Frcge's T ancl F and his Sensesare, rvhy they do not make him "extraorclinary" nonmlntal tiiings' That is rvhich he clung a materialist. Ratht'r, thcy tcstify to the tenacity with before him to his realism2uridcr clialcr,tir,t'lprcssurcsto rvhich no one had been as scnsitivc lrs lltl l'as. III the three Dr. Egicli is excluisitely sensitive to thc pressures from commitment centers of th. basic dialectic. IIer strongest and clearcst inadequacy the on insists \\.&y she her o\Yn In thought. of is to the unity is hcr proThis adequation' problem of the of is of reism. She ",r"r" That fundity. She seesno realistic: ivay of resolving the basic dialectic. for opts she is ivhy, under the pressureof her strongest commitment' a makes also objecilve idealism. But she not only makes a choice, she Not idealist' clai^. Str.tcturally, she claims, Frege is an objective her claim supports she which by reasons or argument the surprisingly, u'hy before dcpend oi those by rvhich she supports her choice' That is at examining the former I shall examine the latter' The conclusion choice rvhich I shall arrir.e is that none of her arguments, either for the or for the claim, is a good argument'26 choose Dr. Egidi believeq l'ith one qualification, that one must qualification conbetrveen Aristotelianism ancl objective idealism' The yet Aristotelianism, of inadequacy the through see rvho cerns those persist in rejecting objectil"e idealism' These, she believes' are-forced she also io rvithdraw into thc desert of dead-end nominalism' Since ?6N oti ce once m ore the heav y emphas i s on the di al ec ti c . I need not and do not of Frege i s "mi s tak en' " cl ai m that her ontol ogy i s "fai s e" or that her readi ng gi v es for them are not good s he arguments or reas ons I merel y cl ai m that thl from G' E ' Moore' reasons. Thi s i s j us t one of the many l es s ons I hav e l earned

150

Locrc AND REALrry

seemsto believc that I am among the dn-ellersin this desert, I shall try to conviuce her that I am not by sho'w-ingfirst of all rvhat is or ought to be obvious, namcly, ti,at dead-end nominalism is not a \ray o u t of t he inr ur s seo f rc i s rn . Rcalistsr holcl thrt "in" l)eter's being blond there are trvo things, Pcter and blond. llcistic rcalisml fails becauseit cannot "connect" the trvo. Dead-cnd non'rirrlrlistsfind "in" this fact only oric thing, namely, Petcr. Ilencc, if tlicy could otherrvisesolve the problem of universals, i.e., if tlicy coukl assigtran or.rtologicalground to Peter's and trIary's both bcirrg lrlond, thcv rvould not at this point have to face the problem rro rcist can solt'e. Yct they rvould be up again."t it at the yery next stcp. What ttcol)nccts" the several facts of thcir rvorld? .Sincethey arc all reists, any trvo facts (or should I say things?) remain disjoined. 'l'hat shorvs thrit the only rvay out of the irnpa-sscof reism is not deadcnd nominalism but, rather, thc recognition of thc ontological status of a class of subsisterrtssufficient not only to make realisml r'iable but also to account for a r,orld that is completely stmctured. Aristotle's individuals are substancesl his character-q,attributes.2T The former "create" or "produce" the latter, the latter "inhere" in the former. The traditional lvords ('create', 'produce', 'inhere') suggest a characteristicfeature. Attributes "depend" on substancesin a sensein rvhich the lattcr do not "depend" on the folmer. 'l'lrut spots a nomin a lis t ic t ender ic y . On th e o th e r l ta n d ,,\r' i s totl e' s srrbst:rncesand attlillutcs are both things. Arry onc u-ho rloesnot rrsc the n'ord as narro l ly as f s om et i rr-rcdso n -i l l tl rc l t' fo rt' r' rrl ll ri nr u ' ' rcal i str." N or i s i t fti r t o c lr , llhinr li l c i s t. t' l rrl rc rc rrrc "o r, (' o n v c r s(,Iy,t(crcati on" i s a sort o f tie. ' l' lr e t r r r r r lrl ci s tl rrr,lr,l , o s o ' ru rrrl t.s ire s r.cal si t to bc i rrcmedi abl y a n thr opor r r or ' plr i <' l,.' l ri s i s l l rc I' rrt:rl{ l a rv n o t j ust of A ri stotel i ani sm b u t of lill s r t lr s t r r rrti rrl ios ltrto l o g i t' s .:s Dr. Iigicli irlcrrtifit's ;\ristotclirrnism (substantiali.sm)not only ri'ith realismr but rilso rvitlr lt'isnr. nlore fatally still, since Aristotle is of coursc a rea,list::rrntI sirrccslrc sccsin him the only alternative to either objcctive idealisrn or tlrrr rrorninalistdesert, she identifies reali-smrand realismz.Thus she fllils to rlistinguish betrveenan)r t\yo of four things as different from clrch othcr as Aristotelianism (substantialism), realisml, realism2,and lcisrn. 'l'liis is her.frsl major mistal,e.It vitiates all her arguments. 27For our purposcs it is safcto ignorethc distinctionbetweenattributesand ac c id e n t s . 2s T h i s s h o * ' s h o w th e issu e o f b a r e Da r ticttla r s ve r s us sub.stancesti es i nto the bas ic d i a l e c t i c .

O \ TO I , O G I C A L

A L TE I TN A TI V I i S

1l-rl

an ontologist but also a'logicilrn' Aristotle, Iikc liregc, rvas not only next some distinctiorrs I)r' That mrtkcs it convenient to intrtcluce though this second failure affccts oi logit' Egidi misscs i,, tftt ""t" than those for her choice' First' "t"' the argrtmcnts for ntt tf"iti more of this are&on which rve completelv thougli, I must dclineate " tttl'o'"o A gr(' (' . & calculus' A calculus acquires Logic rvithout ontology is merely ideal language tt its author claims that it is an philosophical import ;; ontoladequate an reflects (Begt'i.fsschriJt), i'e', tftlt it' perspicuously two the using so by consistently tlogic' for a og)'. I shall mark this distinction calculust a' $'hat is merely $'orcls that 'calculus' stands for or been attached' Ordinary grammar has claim that calculus to rvhich guide reliable a not is it' start from language, although *" t""""t but tol ogi c,otrr(Indo ger m ar r ic) I anguagcsar eallof t hesubject - pr edicat e l"''s p"ttpit"rously reflects his ontology' form. Aristotle'. logit ntore'or it hugs the o't'i'i'-pridicate calculus'i'e'' ,"nggt His calculus l, it reflects that closcly ",t forln of'n"a so orclinlrry grrmniltr subject-preclicatc ontology' his of inltlt:tlrtacics (rather perspicrttousl;j t;t't"itt tq".t:i1',t1 is, lovcd by of '{[ r y The subj ect of ' I'et eiiuut t Nt "'y'is'l'ct er '; t hr Lt P eter' i sl l ary' ft' to'- '"'t "gget at ct lsubjct r t - pr cclicat ccalculust hat isan in rvhich an imporiant senscof 'same' important difference' Y"t ilt"t" is loves fact.2e'Loves NIary' in 'Peter botrr senten.", ,"p."."r-rt the same inadean r ef lect s That r ed'' is N l ary' i s coustrued it 'it r ed'in'This "t transcription of 'All dogs The otu-l'uv quate ontologi.uf '"tutlo"t' preserves the subject-predicate are mammals' into irt" calculus ontological assay of gencralitv' form. 'I'hat rcflects an inadequate not an exaggerated subject-predicate And so on. (Ordinary language is the one a calculus' That is rvhy on calculus simply rl"ti'o* it i''"ot other, the on rvhile, e'erything hand it is flexible enough to represent represent ner5.ni,clou:ll l'^^" therc is much rvhich it does not more hu""-ttt" all this very clcarly' to Egicii O'' *".iiof great It is a are more insights these irin'hose-ambience clearly indeed tno" in insisting --tv As far as I knon" she is also right nt^' i" tf-t"tt rvidely sprcad for admiration her share I all this; and that Frege was the fi"t t"ho sarv though' our agreement.,ends' his momentot, *.rti",ltment' There' calcnltts'i'c'' it may be A ctlculus may be a mYderatenftject-preii'cate hugging this form so'closely that of the subje.t-p,"altuil form l'ithout t"fftt from those specific any ontology it perspicuously reflects ^lu*tt :t.P eterl oves\Iar y i fandonl y i fN ' Iary i s l ov edby P eter' i s anal y ti c .B utn' e as to whether the-thought-that-Peterneecl not at this poini commit ourselves are two things or one' trt"-iftot'gtll-ttt"t-lf loves-)Iary ,rta -y-isJoved-by-Peter

152

Locrc AND REALr.fy

inadecluaciesrvith respect to rclatirlns, gencralitl-, anrl so on, u'hir.lr mar Aristotic's. f am not surc that Dr. ligidi disagt'ee,s. But I am r.ery sure, alas, that she belicvcs any such calculus to be inadequate in an even more ladical scrr-sc. Acrrordingto her, any ontolclgy perspicuously reflectcd by a m.tl<'rlrte subjcct-predicate caiculus mu.st be reistic. This is hcr secortd. nrojrr mistal;;. l'our comments u-ill srrorvhorv bad a mistakc it is aurl rLt tlrc same timc prepare the ground for the third m ajor t ' r il ir is r n . 1. I'rincipia lIu[hcmotica (I,n{) is a modcrate subject-pr.edicatecalculus. Yct it clin lrc uscd and in fact has been useclfor thc correction of those specilic irradccluaciesof the '\ristotelian ontologl-. 2. Notice the conrlition of perspicuous reflection. To unpack tliis labcl is & \-ery major job. '\lso, the'iatter is very teclinical. If onc u'dcrstancls it completell., tlic' he sccsthat a calculus u'hich pcrspicuou-rlyrcflects a (non-reistic!) r'orld l'hose fundamental tie is exemplificatio' must be a moder:rtc subjcct-prcdicate calculus and conversely. 'I'liis is not to say that all philosophers u'ho propoundcd a pll-t1,pe logic realizcd that thc syncategormatic cntities exist. But, thcn, neither clid they fully understand those very technical matters. (8. presently 'lveshall sce that in all calculational respects Fregc's logic is a rnoderate subject-predicatc calculus. Its perspicuity is a dirlcrcnt rnarrer. That 'rvc havc seen already.) 4. I belic'o, {irst, tluit thc logic of thc truncatcd rvorld is of the PII-type. I liclicyc, scc,orrd,tliat the logic of our tvorld as a rvholc, that is, ineludirrg is rrot of this type. Or, to say thc sarnc thing in a n'ay that 'ri'tls, suits our purpose, while su t ' lr : l c alr , r r lusr,a n b c rrs < '
C N TOLOGIC A L A LTE R N A TIV E S

153

the otlicr, namciy, G(P). wiih this I agrce. This is indccd tlrr.'vcry p0irrt of rvhat I just said. A I',N,f-typccalculus can be used to rt'prtrscnt evcry nonmeDtal inteDtion but contains no nlcal)s lor so pcr.spir,uously rt'prcscrrtingG(P) . There, though, our agrecnent ends' il.. t';giai holclsthat uo calculus cau perspicuouslyrcprcsent th. atrt of judging. ,\ny logic for u-hich this claim is madc (as I make it for ,,,ii,,,) ,.he calis,,formalistic." The idea is that no formalistic logic ca' This is her thi'rd major lrcrspicuou,slyreprescnt an adequate ontology' judgirrg it be a mistake may bc to rnistat;e.Some of ny reasons for thesc reasons,thougli, \Iost of Ttr.o. mc|ely implicit in SectionsODcancl is tedious orr Repetition so far. said arc quite explicit in\vhat has been occasion. tliis practical on rrot is any occasiorr.complete explicitness is mistake the bchind motivc So I shall ncxt shorvthat thc intellectual structural idealism. Dr. Egidi's orrtologicalassa,yof judgnrcrit is idcalistic. According to such an o.r"y, thc niinci fi|st "p
154

L OGIC AND REAL IT Y

conterrtion is what is said about judgment in the Begrif sschrift, particularly in $2. Let me once niorc drarv a clcar line betrveenagreementand disagreement. She givcs Frege credit for distinguishing betrveen FP and -P, i.e., betrvccrru juclgrnent arrd l'hat I call thc thought "in" it. If you rccalllr'hat hls ircrcusaid about that early contlibution in Sectakc,qrvhat lircge says tion'f5.o, thel yol f ill sce that I agreed. -(he in $2 to be an a,sscrtionof the urrity of thought. I agrccd that he may plausibly be croditcd rvith this fundamerrtal insight. 'fltere, though, our agreemcnt t-'ntis. tr'<-rrI also believe to have shorvtr three things lghich are inr:ornpatiblervith her contention. I.'irst. Tiic riatural reading of $2 is rcalistic2.Second. This early contribution to thc ontology of judgmcnt is fragmentary; it lear.'esa blank. Third. Evcntually the blank bccomcsa blur. Frege's idea of judgmcnt as &n "advance" ftom a senscto a truth value is irrcmediably blurred. Dr. Egidi takes advantage of the blur by filling the blank u-ith her orvn idealistic ontology of judgmcnt. In this she shorvsonce more her keen senseof structure. For the blur is indecd the seat of the idealistic tendency of the system. If you s.ant to verify this diagnosis,turu to NTLFF 19, lvhere she says perhaps most bluntly rvhat n'rore subtly slie intimates again and again, namely, that a Fregean "Advallcc" is realll' a Fichtean "po.it."" It does not follorv, alas, that her claim is sound. I,'rcgc'sT and F and his scnscsarc most "cxtlaordinary" thing-sindecd. So arc his functions, even though tc-ra mathcmatician bent abor.eall on refuting psychoiogismtliat may not have been as obvious as it rcallf is. Yet all thcsc entities are clearly nonmental. For one, l,\'cge him.self,as far as I kn
AL'|IIRI-ATIVI'IS ON'|OL0(iICAL

155

,scnsetliat tltey tnlLlitt"cxpcrictlce" possible.Shc also tlalls tlrcln "objectiyt'." Ilgt tlrun it tr.arrsltilcs(OLLF 207) that all shc l)lcillls by that ('possiblc" is indepcndcrrt tlf its is tllrt thc t'xpclit,rrlc tlrgy make Iinguist,icr.cplt'scrrtrltionin cxactly the same sensein rvhich a ccrt:ritr fact tluLt ri'lLsttit'tttiotrcdearlier is independent of its being rcprcsclltcd b1' ci tl i cr' l ' t' t cL loves \ I ar y'or 'X'f ar y is loved by Pet er '' This I f ind ,'{ ci i srrpl uri rrting. rvritcs l'rcq<,,likc his successorRussell and like Russell's successors, ltor f(-r)', 'r(.u y)', and so on. The only differenceis in the semantics' possible Iius-s1ll,tlie substitution instancesof f(r)' represent facts or frcts; It-rr Frege, they are names of either T or F. And so on' That makcs it obvious tha| i.n all catculational respects Frege's logic is a moclclate subject-predicate calculus. But it rvili help if bcfore turning to thc second argumcnt I state tlie trvo qualifications rvhich al'e covered by tlie italicized phrase. (a) On the one hand, the assertion sign (iF,) antl the horizontal stroke ('-') do rrot fit into a modcrate subject-predicate calculus. on th.e otlicr hand, l,'rcge hirnself makcs no real calculational use of thcm. That is $,hy the latcr logicians wete puzzled by them and everrtually dropped thcm, rvhich in turn fits rvell ii-ith rvhat Dr. Egicli antl I agree upon, nanely, that Frcgc in $2 introducecl thc trvo signs in order to make by means of thcm a point in the ,T, and 'F' do not fit into a moderate subontology of judgmerrt. (b) ject-predicate calculus. xfore precisely, if they a,retaken to reprcscnt in such a ca, lculusill- f or m ed' But thi rrgs,then'P: T'and 'P: F'ar e then, as far as I kno$., neither l'rege nor, n'ith the possible exception of Alonzo Church, any of his follorvers har.c paid any attcntion to these trro e\pressiolls. Rathcr, it l'as I, rvho, for a purpose of my ou'n, insistcd that they are, or, at least, that they ouglit to be consideredas well-formccl. lly purpose, or, rather' thc use to I'hich I put them in section T$.o, was to exposeby means of them some of the perplcxities of Frcge,s ontology as well as the lack of perspicuity with rvhich it is reflected by his calculus. Dr. Egidi makes much of the fact that in dealing $'ith the sentcntial calculus the Frege of the BegrifsschriJt uses sentential variables, writir.rg,p'instead of, say f(r)'. This she takes to be hfs asscrtion of the unity of thought, upon\vhich she foists her ontology of judgment' The truth of the matter is that no mathematician of evetr tlte most moderate .qkillrvho had either conceived or been told the idea of the sentential calculus rvould in constructing it use any but sentential variables. senteltial logic is the most fundamental part of logic. To have recognized that is rvithout doubt one of Frege's major achievements. But it is disappointing to see Dr. Egidi build so much of her argument on the

156

Locrc AND REALrry

trivial fact that he rvrites the sentential calculus in sentential notation. Besidcs,if it lucre the job of the single variable to exprcssthe unity of thought, may I ask hon' she rvould explain rvhat shc also asserts, namely, that thc sarnc job is also done by the horizontal stroke? The indictment f have dran n up against Dr. Egidi is scvere. Yet I do not rvant to end on a note of disappointment. So I shall changesides, from the prosccution to the defense,as it rvere,and concludc rvith some remarks lr,hich rvill add pcrspective in a rvay that amounts to a plea of attenuating cilcumstances. The irasic dialcctic of the realism-idealismissue lurks in the things thcmsclvcs. 'I'he trend torvard idealism began orrly about thrce hundrcd years ago, that is, roughly, at the time of the Caltcsian revolution. \Yhy did this trend start at just that timc? The gist of what I belicvc to be the right ansu'er can be stated very briefly.33The Aristotelian-Thomistic account of perception, n hich l.as dorninant until then, is realistic2 in structure. A single substantial form irrforms the mind of the perceiver and the thing perceived. This account \yas supported by and is compatible u'ith thc idea that a mind can only knorv what is "in" it. Thc revolution overthrerv the old account of perception. The idea continucd to be taken for granted. That suffices to account for the rise of the trend. The lure of idcalism continues undiminished. The structure of Derveyan instrumcntalism is idcalistic.3{'Ihe same is true of the preserrt miscry, at Oxford and elservliere,rvhich goes by thc name of ordinary-Lrnguagcphilosophy. If the temptation of a philosophy so absurd cclntinucs so strong for so long, it stands to reason that, for all its absurdity tnd cvcn though at a proliibitir.e price, it accounts more adcquatt:lv thorr its competitors for at least one striking feature of the rvorld. Thc thing to rlo, therefore, is to idcntify this feature and try to do justice to it in u,rcalisticr ontology. Then and only then will the temptation ccasc. The role of minrl,s in the world is unique. This is the feature. The idealists' r.vay of safcgrurrding it is to insist, with or ri-ithout some attenuation, that mirrds "create" their intentions, rvhich is absurd. X[y way of safeguarding it is different. Consider (1) 'P and Q', (2) 'not-(not-P or not-Q)', (3) 'P and Q if and only if not-(not-I'or not-Q)'. Call the facts representedby (1) and (2) F1 and F2, respectivcly.Are F1 and F2 trvo or one (the,ss.mg)7 There is an important meaning of 'samc' upon rvhich they are the same. 'l'his 33l ' o r a m o r e d e t aile d sta tcm e n t se e e ssa y XII o f thi s book. 3{ See N{ay Brodbeck, "La filosofia di John Dewey," Riuista d.i Filosofi,a, 50, 1959 , 3 9 1 - 4 2 2 .

ON TOLOGIC A L A LTE R N A TIV E S

I57

meaning I explicate by the analyticity of (3). sincc the subsistents exist, there is also a strictcr meaning of 'same', such that trvo farlts arc tlie same if and only if the same simplesare in the same way conttct,tcd by the same subsistents.With this meaning of 'same', Fr and l|2 rlre trvo aud not one. The subsistentsand thc truths u.hich dcpertd only on them are "the l.orld's form." The ontological ground of F1 and Il2 being trvo and not one thus lies not in the rvorld's things but'w'holly in its form. Yet there are in my \vorld two things, namcly, the-thoughtthat-P-and-Q and the-thought-that-not-(not-P-or-not-Q) u'hich are trvo and not one for the sole reason that if the rvorld's form is taken into account the facts they represent are tx'o and not one. Therc is thus a kind of things, namely, thouglits, tvhich are unique in that thcy and they alone among the lvorld's things reflect its form.35This, I submit, is an adequatc rcalisticz accorttttof the featurc. I shall of coursc ttot cottvitt<'cI)r. I')gidi.'I'o expect that rvould lle mcreiy presumptuous. IJrrt I do trotrrish tt ntol'o nrotlest hope. I may q'lro irr tlrcir tt'tt"ytry to rcsolve the convince her tluit tltcrtr lrrc relrli-sts:r dialectical tcnsiotrs to rvhirrh shc so kccnly rcspontls. If this hope is justified, thcn I am confidcnt she tvill cvcntually find her way of rcjecting idcalism. 35Thi s i s al s o the deepes t s tntc tural reas on for s o ex pl i c ati ng ' anal y ti c ' that rtpl 11pt beco mes anal l ' ti c or, s y nony mous l y , that i t bec omes a truth i n the rvorl d's form. There arc al s o qui tc a few l ogi c al (c al c ul ati onal ) reas ons for thi s step. P hi l osophi c al l y , though, s uc h reas ons tl o not c arry c onv i c ti on unl es s thc y .qupport and are s r.rpported by a s truc tural reas on that l i es rather deep.