EUROPEAN REVI EW OF PHI LOSOPHY
Edi t or i al Boar d Manuel Gar ci a Car pi nt er o Rober t o Casat i Er os Cor azza j ...
12 downloads
542 Views
39MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
EUROPEAN REVI EW OF PHI LOSOPHY
Edi t or i al Boar d Manuel Gar ci a Car pi nt er o Rober t o Casat i Er os Cor azza j ér ôme Doki c Pet r Kot át ko Mar t i n Rechenauer Gi anf r anco Sol dat i TomSt oneham
Cogni t i ve Dynami cs 2
Chr i st i ne Tappol et
U
edi t ed by JEROM E
D0KI C
Csu Publ i cat i ons Cent er f or t he St udy of Language and I nf or mat i on St anf or d, Cal i f or ni a
Copyr i ght ©1997 CSI , I Publ i cat i ons
Cont ent s
Cent er f or t he St udy of Language and I nf or mat i on 1 el and St anl o RI Juni or Uni ver si t y 01 00999897
Par t I : Cogni t i ve Dynamks I nt r oduct i on / 3
321
I SBN: 1- 57590- 073- 2 I SBN: 1- 57586- 072- 4 ( phk. )
JEROME DOKI C
Ri p van Wi nkl e and Ot her Char act er s I 13
:si . 1
Publ i cat i ons r epor t s new devel opment s i n t he st udy of I an gnage, i nf or mat i on, and comput at i on. I n addi t i on t o l ect ur e not es, our publ i cat i ons i ncl ude monogr aphs, wor ki ng paper s, r evi sed di sser t at i ons, and conf er ence pr oceedi ngs. Our ai m i s t o make newr esul t s, i deas, and appr oaches avai l abl e as qui ckl y as possi bl e. Pl ease vi si t
01. 11-
websi t c at ht t p: / / csI i - wvst ant or d. edu/ publ i cat i ons/
l o r comment s on t hi s and ot her aut hor and Pt i hl i 5hei
t i des, as wel l as f or changes and cor r ect i ons by t he
f t c t ext i n t hi s book was set by CSLI Publ i cat i ons i n Mi ni on, a t ypehi ee desi gned by Rober t Sl i mhaeh, The di spl ay t ype i s set i n St one Sans, desi gned by Si mmer St one. The hook was pr i nt ed and hound i n t he Uni t ed St at es of Amer i ca.
JOHN PERRY
The Dynami cs of Si t uat i ons I 41 FRANçOI S RECANATI Dynami c Thought s and Empt y Mi nds / 77 MI CHAEL LUNTLEY
Dynami c and Coher ent Thought s / 105 MAt TE EZCURDI A
Cogni t i ve Dynami cs: An At t empt at Changi ng Your Mi nd / 141 CHRI STOPH HOERL
Bel i ef , Cont ent , and Cause / 159 Tost Es GRI MALTOS ANDCARLOS J. M OYA
Par t I I : Ot her Topi cs Cr i t i cal Not i ce of Fr ançoi s Recanat i ' s Di r ect Ref er ence I 175 ALBERTOVOLTOLI NI
10 t ( Æ 2- S1
Par t I Cogni t i ve Dynami cs
r
cogni t i ve
JEROME DOKI C I nt r oduct i on
It m ay be usef ul t o comment br i ef l y on t he t heme of t hi s vol ume, dynami cs, i n or der t o hi ghl i ght some of t he phi l osophi cal i ssues
r el at ed t o i t , as t hey wi l l be di scussed by our cont r i but or s. 1. 11w phr ase " cogni t i ve dynami cs" was once coi ned by Davi d Kapl an i n hi s cl assi cal wor k " Demonst r at i ves. " Accor di ng t o Kapl an, t her e i s a pr obl emposed by i ndexi cal r ef er ence over t i me: t he pr obl emof cogni t i ve dynami cs can he pot l i ke t hi s: what doe, s i t mean t o say of an i ndi vi dual who at one t i me si ncer el y asser t ed a sent ence cont ai ni ng i l l dexi cai s t hat at some l at er t i me he has ( or has not ) changed hi s mi nd wi t h r espect t o hi s asser t i on? ( 1992: 537, 11. 64) ' 11105
Mor e pr eci sel y, Kapl an says t hat a t heor y of cogni t i ve dynami cs i s concer ned wi t h t he quest i on of what ( t d, Jos/ mcoi s one has t o make i n t he expr essi on of
one' s
st andard
i ndexi cal t hought i f one i s t o expr ess t he same
( i ndexi cal ) opi ni on as t i me passes. A f r t mous exampl e of such an adj ust ment comes f r om Fr ege. I l owcan I expr ess on Tuesday t he t hought I expr essed on Monday when I sai d " ' l oday i s a beaut i f ul cl ay, " Fr ege asks? Hi s answer , i n " The Thought , " whi ch you wi l l meet sever al t i mes i n t hi s vol ume, sounds pr et t y obvi ous: I F someone want s t o say t he same t oday as l i e expr essed yest er day usi ng t he wor d ' ' t oday, l i e t n ust r epl ace t hi s wor d wi t h " yest er day. " Al t hough t he t hought i s t he sa I I I C, i t s ver bal CXI ) messi oi i i n ust he di f l er en t i n or der t hat t he change of sense whi ch woul d ot her wi se be af f ect ed by t he t i mes of ut t er ance may be cancel l ed out ( 195( 1: 296) .
di f f eri ng
4 Eur opean Revi ew of Phi l osophy However , t he seemi ng obvi ousness i n Fr ege' s answer i s mi sl eadi ng. I n t hi s passage, 1r ege has somet hi ng qui t e speci f i c and cont r over si al i n mi nd, as one can gat her f r omt he f amous debat e bet ween John Per r y ( 1993: Essay 1 and i t s Post scr i pt ) and Gar et h Evans ( 1985: Essay 10) : i f I amt o expr ess on Tuesday t he t hought expr essed t he day bef or e wi t h " Today i s f i ne, " I must use t he wor d " yest er day, " and not any ot her expr essi on co- r ef er ent i al wi t h " t oday" ( such as " my bi r t hday' or " Apr i l 26, 1996" ) . I n Evans' s i mage, wor ds l i ke " t oday, " " yest er day" and per haps " t omor r ow" f or m a f ami l y of i ndexi cal adver bs whi ch can be used t o expr ess t he same sense i n di f f er ent cont ext s. W e l i see one di f f i cul t y when we t r y t o gi ve f ur t her , subst ant i al i dent i t y condi t i ons f or i ndexi cal t hought s gr asped at di f f er ent t i mes: t hat t he cr i t er i on of di f f er ence of t hought s of f er ed by Fr ege ( f oes not seem t o he ver y hel pf ul i n t hi s connect i on. I n i t s st andar d f or mul at i on, t he cr i t er i on says t hat i f a r at i onal Subj ect gr aspi ng t he t hought s expr essed by t wo sent ences he accept s one senadopt s di f f er ent epi st emi c at t i t udes t owar d t hem t ence as t r ue, but suspends hi s j udgement about t he ot her ) , t hen t he t hought s cannot be t he same. I l owever , a pr econdi t i on f or appl yi ng t he cr i t er i on seems t o be t hat t he subj ect gr asp t he t hought s a! t he same t i me. The poi nt of i nt r oduci ng a " r at i onal " subj ect i s t hat hi s epi st emi c at t i t udes ar e coher ent ; t he subj ect ' s at t i t udes t ake car e of each ot her . Rat i onal coher ence, t hough, i s f i r st and I br emost a synchr oni c not i on. I f we want t o def i ne a di achr oni c not i on of r at i onal coher ence, we must have an i ndependent account of what i t i s t o gr asp t he same t hought over t i me. An exampl e shoul d hel p t o make t hi s poi nt cl ear . Suppose t hat on Monday mor ni ng, I t hi nk " Today i s l i ne?' i hen t he sky gr adual l y becomes cl oudi er ; i n f act , Monday t ur ns Out t o he a r ai ny day. On Tuesday, I t hi nk " Yest er day was not l i ne. " I hol d di f f er ent epi st emi c at t i t udes t owar d t he sent ences " l oday i s f i ne" ( as used on Monday mor ni ng) and " Yest er day was f i ne" ( as used on Tuesday) . I accept t he f or mer , but I r ej ect t he l at t er . Can t he Pr egean cr i t er i on of di f f er ence be used t o show t hat t he sent ences expr ess di f f er ent t hought s? i nt ui t i vel y, t he answer i s negat i ve. I f t he cr i t er i on coul d be used t o show t hi s, i t woul d al so i mpl y, i ncor r ect l y, t hat t wo t okens of t he " et er nal " sent ence " Apr i l 26, 1996 i s a f i ne day" ( wher e t he " i s" i s at empor al ) expr ess di f f er ent t hought s on Monday and on Tuesday: I accept t he sent ence as t r ue on t he f i r st day; I r ej ect i t on t he next .
I nt r oduct i on
5
2. I n t he exampl e above, i t i s obvi ous t hat t he pr obl em ar i ses because I have changed my mi nd about Monday' s weat her . So we ar e back t o Kapl an' s quest i on: what does i t mean t o say of an i ndi vi dual who at one t i me si ncer el y asser t ed a sent ence cont ai ni ng i ndexi cal s t hat at some l at er t i me he has ( or has not ) changed hi s mi nd wi t h r espect t o hi s asser t i on? chr i suph l - l oer l deal s wi t h Kapl an' s quest i on expl i ci t l y. I nst ead of wor ki ng wi t h an i ndependent not i on of sense t o account f or t he phenomenon of changi ng one' s mi nd, l i e t r i es t o di scover what i s pr esupposed by t hat phenomenon. I n t he cour se of doi ng t hi s, he gi ves some necessar y condi t i ons f or a change of mi nd. The f i r st one i s qui t e i nt ui t i ve: when I change my mi nd, I go f r oma gi ven bel i ef t o a cont r adi ct or y one. For exampl e, I go f r om t he bel i ef t hat Apr i l 26, 1996 i s a l i ne ( l ay t o t he bel i ef t hat Apr i l 26, 1996 i s
not a f i ne day. l - l owever , t her e i s an i mpor t ant const r ai nt her e; as l I ner ! says, i f we want t o descr i be t he st r uct ur e of a change of mi nd i n or der t o under st and what i t means t o r et ai n a bel i ef ( or , i n Fr egean t er ms, t o gr asp t he same t hought over t i me) , we must be car ef ul not t o i nt r oduce an i ndependent not i on of what i t i s f or t wo bel i ef s hel d at di f f er ent t i mes t o cont r adi ct each ot her or not . Hoed t hen dr aws some i nt er est i ng cor ol l ar i es. Fi r st , t he phenomenon of changi ng one' s mi nd depends 01) a capaci t y t o have hi gher - or der t hought s: t hought s about my bel i ef s bef or e and af t er t he change of mi nd, and about t he r el at i ons bet ween t hem. A smal l chi l d wi t h i nsuf f i ci ent under st andi ng of hi s own ment al st at es cannot be sai d t o change hi s mi nd. Second, i f I cannot change my mi nd wi t hout knowi ng t hat I change my mi nd, t hi s knowl edge i s based on memor y. Memor y i s what l i nks my bel i ef s bef br c and af t er my change of mi nd. The t hi r d cor ol l ar y i s par t i cul ar l y cent r al t o 1- l oer l ' s pr oj ect . I f changi ng one' s mi nd i mpl i es knowi ng t hat and howwe have changed our mi nd, t her e must be a way t o he di r ect l y awar e t hat t he same cont ent i s hen I change my mi nd bet ween Monday and Tuesent er t ai ned over t i me. W 1 knowt hat I day, change my mi nd about t he same cont ent , once expr essed i s f i ne, " once expr essed wi t h " Yest er day was l i ne?' ( A r el at i vel y wi t h " Today neut r al not i on of cont ent i s at st ake her e. ) I n ot her wor ds, Hoer l cl ai ms t hat t her e Must be a di achr oni c i dent i t y cr i t er i on f or t hought s hel d at di f f er ent l i mes. ( See al so Mi chael l , unt l ey' s ar t i cl e f or an expr essi on of t hi s cl ai m. ) Somet i mes, t he f act t hat sent ences used at di f f er ent t i mes expr ess t he same
6
Eur opean Revi ewof Phi l osophy
t hought can be an epi st eni i c gi ven. Thi s mi l i t at es agai nst t oo st r ongl y ext er nal i st not i ons of r at i onal i t y and of di achr oni c coher ence. At some poi nt i n hi s ar t i cl e, 1- ben al so t ouches on what I bel i eve i s an i mpor t ant r equi r ement on any t heor y of cogni t i ve dynami cs: t o account f or a change of mi nd, i t i s not enough t o say t hat cont r adi ct or y bel i ef s ar e hel d successi vel y; we have t o expl ai n howt he l at er bel i ef can r epl ace t he ear l i er one i n t he subj ect ' s cogni t i ve economy. Mor e gener al l y, t o expl ai n what i t i s t o ent er t ai n t he same cont ent over t i me, i t i s not enough t o poi nt out t hat t he same cont ent i s ent er t ai ned at di f f er ent t i mes; we have t o say somet hi ng about t he di achr oni c r el ( ui ous bet ween t he r el evant ment al st at es. Change of mi nd i s mor e wi despr ead t hat we may t hi nk, and i t i s easy t o over l ook t he pr esence of a change of mi nd when we come t o i nt er pr et i ng a par t i cul ar exampl e. Thi s mi ght have i mpor t ant consequences f or one' s t heor y of t he i dent i t y condi t i ons of t hought and bel i ef . Thi s i s ' I bbi es Gr i mal t os and Car l os Moya' s cl ai m, who suggest t hat t he st andar d di scussi on of t wo f ami l i ar cases of change i n bel i ef i s f l awed. Let us consi der f or exampl e t he f i r st case t hey di scuss, i . e. , John Per r y' s f amous exampl e of t he shopper . Her e i s Per r y' s own descr i pt i on of t he exampl e: I Once f ol l owed a t r ai l of . sugar on t i t per i n ar kct f l oor , pushi ng owcar t down t he ai sl e on one si de of a t al l cot i t i t er and back t he ai sl e on t he ot her , secki ng t he shopper wi t h t he t or n sack t o t el l hi nt he was ma ki ng a mess. . . But I seemed unabl e t o cat ch up. Fi nal l y i t dawned on me. I was f i l e shopper I was t r yi ng t o cat ch ( 1993: 33) . Ther e ar e at l east t wo r el evant bel i ef s i n t hi s exampl e, The f i r st i s expr essed by " The shopper wi t h a t or n sack i s maki ng a mess; " t he second i s expr essed by " I ammaki ng t mess?' Per r y' s mai n t hesi s about t hi s exampl e i s t hat t he wor d " t , " i n t he expr essi on of t he second bel i ef , i s essent i al : i f we r epl ace i t wi t h ot her desi gnat i ons of t he same per son, we l oose t he f or ce of t he expl anat i on; we cannot expl ai n why I f i nal l y st opped t he car t . Gr i mal t os and Moya have no quar r el wi t h t hi s t hesi s; t hei r cr i t i cal t ar get i s r at her t he way Per r y i s pr one t o descr i be t he change i n bel i ef i n t he exampl c. Per r y f avour s a t wo- t i er ed appr oach i n whi ch t her e i s a di st i nct i on bet ween bel i ef st at e and pr oposi t i on bel i eved. Thi s di st i nct i on i s al so expl ai ned ( and r ef i ned) i n Per r y' s ar t i cl e i n t hi s vol ume, but i n a nut shel l i t coni es t o t hi s: t he pr oposi t i on bel i eved cont ai ns t he obj ect s and t he pr oper t i es t hought about t hemsel ves and not , as i n Fr ge' s t heor y, modes of pr e-
I nt r oduct i on
7
sent at i on of t hem. Now we coul d say t hat t he t wo bel i ef s i n t he exampl e e can expr ess t he same pr oposi t i on: t hat John Per r y i s maki ng a mess. ( W suppose f or t he sake of ar gument t hat t he phr ase " t he shopper wi t h a t or n sack" r ef er s di r ect l y t o John Per r y. ) So t he change i n bel i ef whi ch expl ai ns t he st oppi ng of t he car t i s not a pr oposi t i onal change; f or Per r y, i t i s a change i n t he bel i ef st at e. Gr i mal t os and Moya di sput e t hi s, and t r y t o show t hat what expl ai ns t he change i n behavi our i s a change i n t he pr oposi t i on whi ch i s bel i eved. Roughl y, when I f ol l owed t he t r ai l , I bel i eved t hat one of t he shopper s, hut not r ue, was maki ng a mess. A bi t l at er , 1 r eal i sed t hat i t was me, and not one of t he ot her shopper s, who was maki ng a mess. So I changed my mi nd; I hel d cont r ar y at t i t udes t owar d t he sent ence " 1 am maki ng a mess?' My st oppi ng t he car t i s expl ai ned by my comi ng t o bel i eve a new pr oposi t i on, whi ch I di d not bel i eve bef or e. So one i nf l uent ar gument f or t he di st i nct i on bet ween bel i ef st at e ( mor e gener al l y, causal r ol e) and pr oposi t i on bel i eved ( mor e gener al l y, semant i c r ol e) i s r ebut t ed. 3. I n Gar et h Evans' s ( 1985: Essay 10) i nt er pr et at i on of Fr ege, t he pr obl em of cogni t i ve dynami cs i s t he nar r ow pr obl em of what i t means t o e keep t r ack of t i l e same keep t r ack of an i ndexi cal t hought over t i me. W t hought by keepi ng t r ack of t he r ef er ent of t hat t hought i n t i l e same way. For exampl e, I can keel ) t r ack of t he dei ct i c t hought expr essed by" Thi s i s a bi r d" by keepi ng my eyes on t he bi r d movi ng i n f r ont of me. I n ot her cases, we cannot cont i nue t o keep t r ack of t he same obj ect i n t he same way, because we have l ef t t he cont ext i n whi ch i t was possi bl e t o r ef er t o t he obj ect wi t h t he same sor t of i ndexi cal . Hence i t appear s t o be possi bl e t o keep t r ack of t he same obj ect wi t hout keepi ng t r ack of t he same t hought i n Fr cge' s sense. Thi s shows one way i n whi ch t he nar r ow pr obl em of cogni t i ve dynami cs cal l be
general i sed.
general i sed
i n many ot her John Per r y shows t hat t he pr obl em can be " i nf or m at i on I -I c i . e. , gam es: ' gani es whi ch ways. car ef ul l y anal yses ei ght t he ul t er i or t he of a of i nf or m at i on and pi ece appl i cat i on i mpl y acqui si t i on of t hat pi ece of i nf or mat i on, t ypi cal l y when t he subj ect act s upon t he r el evant obj ect . ' 11w f i r st t wo games i nvol ve onl y i ndexi cal bel i ef s, si nce i n Per r y' s t er mi nol ogy t he subj ect mai nt ai ns hi s cpi st cmi c and pr agmat i c " at t achment " t o t he obj ect . ( Ther e ar e si mpl e t echni ques avai l abl e t o t he
8
I nt r oduct i on
Eur opean Revi ewof Phi l osophy
subj ect f or gai ni ng i nf or mat i on f r omt he obj ect , and f or act i ng upon i t . ) I n t he t hi r d ki nd of game, " det ach and r ecogni se, " t her e i s a gap bet ween acqui si t i on and appl i cat i on i n whi ch t he subj ect i s not at t ached any mor e t o t he obj ect ; i t has become " i ndexi cal l y i naccessi bl e. " I n one sense, he can t hi nk of t he obj ect usi ng an i ndexi cal sent ence such as " That was a bi r d, " wher e hi s t hought i s based on t he memor y of t he bi r d, but t hi s i s not enough t o keep t r ack of t he obj ect i n Per r y' s sense. " Remember i ng" i s one of t he f i ve ot her games di st i ngui shed by Per r y, whi ch ar e t ypi cal l y pl ayed when t he subj ect i s det ached f r omt he obj ect . Usi ng t hese concept ual t ool s, Per r y i s abl e t o gi ve a f r esh anal ysi s of Fr ege' s exampl e wi t h " t oday" and " yest er day?' Consi der t he case wher e I Cont i nue t o hol d t he same bel i ef about Monday' s weat her , a bel i ef I expr ess on Tuesday wi t h " Yest er day was a f i ne day?' I n t he oor ni al case, I j ust pl ay t he game cal l ed " updat i ng" - l have a concept i on of t he cont ext ual change whi ch has t aken pl ace bet ween Monday and Tuesday, and t hi s concept i on al l ows me t o make t he adj nst ment f r om" t oday" t o " yest er day: ' and f r omt he e ar e cl ose t o Kapl an' s not i on of a " st andar d pr esent t ense t o t he past t ense. W adj ust ment . " I l owever , t her e i s anot her game t hat can be pl ayed her e. I can r t ' mcnt hcr Monday' s weat her , so t hat I can t hi nk of Monday by expl oi t i ng t he pi eces of i nf or mat i on I acqui r ed t hen. Thi s game i s si mi l ar t o t he game " det ach and r ecogni se. " ( I t i s onl y si mi l ar , si nce we cannot l i ve t he same day t wi ce. ) t he avai l abi l i t y of t hi s game i s i mpor t ant t o account f or ot her cases wher e t he subj ect does not updat e cor r ect l y- l i ke Ri p van Wi nkl e, f or exampl e, he t hi nks he has sl ept one ni ght but i n f act he wakes up sever al year s l at er . I n Per r y' s f r amewor k, such a subj ect , al t hough gr ossl y mi st aken, can cont i nue never t hel ess t o t hi nk of Monday by pl ayi ng t he backi ng game j ust ment i oned. Ther e i s an i mpor t ant di st i nct i on bet ween t wo ki nds of moves i nvol ved i n an i nf or mat i on game. On t he one hand, t her e ar e compensat or y moves, whi ch ar e j ust i f i ed i nsof ar as t her e has been an obj ect i ve change i n t he con-
compensa- t or y t ext of t hought and bel i ef . For exampl e, " updat i ng" moves ar e i f t he passage of t i me i s consi der ed as an obj ect i ve change. On t he ot her
hand, t her e ar e non- compensat or y moves, i . e. , moves whi ch ar e not j ust i f i ed by obj ect i ve changes. For exampl e, " i nf er r i ng" and " r emember i ng" ar e games whi ch t ypi cal l y i nvol ve l i on- compensat or y moves. Fr ancoi s Reeanat i cl ai ms i n hi s ar t i cl e t hat t her e ar e non- compensat or y moves t hat never t hel ess i nvol ve a cont ext ual change, even t hough, as non-
9
compensat or y, t hey ar e not caused by an ant ecedent obj ect i ve change. Consi der an ut t er ance of " i t i s r ai ni ng" made i n Par i s at some par t i cul ar t i me. Fol l owi ng Per r y ( 1993: eb. 10) , Recanat i says t hat t he pr oposi t i on expr essed by t hi s ut t er ance has an " unar t i cul at ed const i t uent " whi ch cor r esponds t o a pl ace. I t i s i n vi r t ue of t hi s const i t uent t hat t he ut t er ance concer ns Par i s; t he ut t er ance i s t r ue i f and onl y i f i t i s r ai ni ng i n Par i s. I n Recanat i ' s Aust i ni an t er mi nol ogy, Par i s i s t he si ! oaüon whi ch suppor t s t he f act t hat i t i s r ai ni ng. Nowwhen t he subj ect moves f r om" I t i s r ai ni ng" t o " I t i s r ai ni ng i n Par i s, " t her eby r ender i ng expl i ci t t he pl ace whi ch i s i n quest i on, t he si t uat i onal component has shi f t ed. The name " Par i s" i s ext r act ed f r omit cont r ast i ve set cont ai ni ng ot her names f or ci t i es, e. g. , i n Eur ope. So t he si t uat i on whi ch suppor t s t he f act t hat i t ' s r ai ni ng i n Par i s i s Eur ope, not Par i s. The move f r om" i t i s r ai ni ng" t o " i t i s r ai ni ng i n Par i s" i nvol ves a cont ext ual change. Reeanat i gener al i ses i n ef f ect t he not i on of an unar t i cul at ed const i t uent : t her e i s al ways a pur el y cont ext ual component of cont ent , whi ch i s t he si t uat i onal el ement . Recanat i shows how power f ul a t ool i s t he not i on of an Aust i ni an pr oposi t i on, wi t h i t s di st i nct i on bet ween si t uat i on and f act . For i nst ance, t he not i on i s used t o deal wi t h r easoni ng and r easoni ng about
count er f act ual
someone el se' s ment al st at es. I n Recanat i ' s t heor y, new pr obl ems of cogni t i ve dynami cs ar i se. For t he t r ansi t i on f r om" I t i s r ai ni ng" t o " i t i s r ai ni ng i n Par i s" not onl y i nvol ves a cont ext ual change; i t al so i nvol ves a change i n coni c??! ( i f cont ent i s def i ned i n t er ms of Aust i ni an Pr oposi t i ons) . So we have t o expl ai n howi t i s possi bl e t o keep t r ack of t he same phenomenon ( t he r ai n) t hr ough changes i n cont ext and i n cont ent . 4. The pr obl emof cogni t i ve dynami cs i s al so r el evant t o a per enni al t heme i n phi l osophy, i . e. , t he quest i on of t he r el at i ve pr i or i t y of t he t heor i es of t hought and l anguage. Mi chael Dummet t i s f amous f or hol di ng t hat i n t he or der of expl anat i on, t he t heor y of l anguage i s pr i or t o t he t heor y of t hought ; i t i s onl y t hr ough a st udy of l anguage t hat we can have a sat i sf act or y account of t hought . ( Accor di ng t o t ) ummet t , t hi s pr i or i t y t hesi s i s even def i ni t i ve of anal yt i c phi l osophy. ) Now Mi chael Lunt l ey ar gues t hat t he exi st ence of dynami c t hought s poses a pr obl emf or a phi l osopher l i ke t ) ummet t who want s t o def i ne t hought over l i ngui st i c i t ems. W hen I go f r om " I bday i s f i ne" t o " Yest er day was f i ne: ' I gr asp one and t he same t hought ,
10
I nt r oduct i on
Eur opean Revi ewof Phi l osophy
but t he l i ngui st i c expr essi on changes. I n f act , accor di ng t o Lunt l ey, my keepi ng t r ack of a dynami c t hought i s not gr ounded on r ul es of l i ngui st i c meani ng. As an expl anat i on of dynami c t hought s, Lunt l ey woul d pr obabl y r ej ect Kapl an' s suggest i on t hat t her e ar e st andar d adj ust ment s i n t he expr essi on of one' s t hought . To Lunt l ey' s mi nd, a pr oper , Fr ege- i nspi r ed i nt er pr et at i on of dynami c t hought s shoul d showt hat t he " Empt y Mi nds" t hesi s hol ds: we do not need r epr esent at i ons i n or der t o t hi nk, f or t hi nki ng i s not an oper at i on def i ned over r epr esent at i ons. I n Lunt l ey' s sl i ght l y pr ovocat i ve t er ms, " t hi nki ng does not r equi r e t hought s. " The not i on of r epr esent at i on cover s her e not onl y publ i c l anguage sent ences, but al so symbol i c i t ems i n t he l anguage of t hought . I n al l t hese cases, t he not i on of cogni t i ve si gni f i cance i s not def i nabl e i n t er ms of t he " ar r angement " of r epr esent at i ons. Lunt l ey dr aws i n a r adi cal way what he sees as t he nat ur al consequences of Evans' s ext er nal i st not i on of a dynami c t hought . I t i s t r ue t hat i n or der t o expl ai n t hat not i on, we have t o i nvoke t he subj ect ' s egocent r i c poi nt of vi ew. ( By cont r ast , a t heor y of l i ngui st i c meani ng i s " i nsensi t i ve" t o t hat poi nt of vi ew. ) However , t he not i on of an egocent r i c poi nt of vi ew i s concei ved i n t hor oughl y ext er nal i st t er ms: i t i s def i ned by r ef er ence t o t he ongoi ng or gani sat i on of t he subj ect ' s r at i onal behavi our i n hi s envi r onment . Cont r ast Mai t e Ezcur di a' s t hesi s. She cl ai ms, about Fr ege' s exampl e, t hat t he sameness ( or di f f er ence) i n t hought s expr essed by sent ences cont ai ni ng i ndexi cal wor ds l i ke " t oday, " " yest er day" or " t omor r ow" i s f ul l y det er mi ned by t he semant i c r ul es whi ch gover n t hei r uses. Her e i t seems t hat l anguage i s pr i or t o t hought i n t he or der of expl anat i on, f or i t i s t hr ough an i nvest i gat i on of t he l i ngui st i c meani ng of t hese i ndexi cal wor ds t hat we can det er mi ne t he t hought s t hey cont r i but e t o expr ess i n par t i cul ar cases. Ezcur di a opposes Evans' s i nt er pr et at i on of t he Fr egean exampl e. Accor di ng t o Evans, t he ut t er ance on Monday of " Today i s f i ne" and t he ut t er ance on Tuesday of " Yest er day was f i ne" expr ess t he same dynami c t hought onl y i f t hey ar e connect ed by an exer ci sed abi l i t y t o keep t r ack of Monday over t i me. Thi s i s pr eci sel y t he abi l i t y whi ch someone l i ke Ri p van Wi nkl e has l ost . Ri p l ost t r ack of Monday, and wi t h i t t he t hought he expr essed on t hat day wi t h " Today was f i ne. " ( Mor e pr osai c cases i nvol ve a f ai l ur e t o r eal i se t hat mi dni ght has passed. ) Ezcur di a, l i ke Per r y but f or di f f er ent r easons, cl ai ms t hat t hi s aspect of Evans' s t heor y of t hought i s not pl ausi bl e. She suggest s t hat we di st i ngui sh t he psychol ogi cal l evel of t hi nki ng and t he semant i c
11
her eas t hi nki ng may i nvol ve t he ki nd of t r acki ng abi l i t i es l evel of t hought . W descr i bed by Evans, t hought i s i dent i f i ed essent i al l y i n t er ms of l i ngui st i c meani ng. 5. Of cour se, t hi s vol ume has no pr et ence of cover i ng al l t he aspect s of t he pr obl emof cogni t i ve dynami cs. To gi ve but one exampl e, not hi ng has been sai d about t he t heor i es of r at i onal changes of bel i ef - whi ch expl or e t he dynami cs of whol e bel i ef syst ems ( cf . f or exampl e For r est 1986, and Gär denf or s 1988) . I n one sense, t he cont r i but or s t o t hi s vol ume have been concer ned, not wi t h t he quest i on of when i t i s r at i onal t o change one' s bel i ef s, but wi t h t he mor e f undament al quest i on of what i t means t o mai nt ai n t he same bel i ef ( or a bel i ef wi t h t he same cont ent ) over t i me i n one' s cogni t i ve economy. I ncl uded i n t hi s vol ume i s al so a cr i t i cal not i ce, by Al ber t o Vol t ol i ni , of Recanat i ' s book Di r ect Ref er ence: Fr omLanguage t o Thought ( 1997) . CREA, Par i s Uni ver si t y of Geneva
Ref er ences
Evans, G. 1985. Col l ect ed Paper s. Oxf or d: Cl ar endon Pr ess. For r est , P. 1986. The Dynami cs of Bel i ef Oxf or d: Bl ackwel l . Fr cge, G. 1956. The Thought : A Logi cal Enqui r y. Tr ansl at ed by A. and M. Qui nt on, Mi nd 65, pp. 289- 311. Kapl an, D. 1989. Demonst r at i ves. Avai l abl e as a ms. si nce 1977, and t hen publ i shed i n J. Al mog, J. Per r y, and H. Wet t st ei n, eds. , Themes f r om Kapl an, New Yor k: Bl ackwel l , pp. 481- 563. Per r y, J. 1993. The Pr obl em of t he Essent i al I ndexi cal and Ot her Essays Oxf or d: Bl ackwel l . Recanat i , F. 1997. Di r ect Ref er ence: Fr omLanguage t o Thought . Oxf or d: Bl ackwel i .
(
JOHN PERRY
Ri p van Wi nkl e and Ot her Char act er s
I f someone want s t o say t he same t oday as he expr essed yest er day usi ng t he wor d ' t oday; he must r epl ace t hi s wor d wi t h ' yest er day
- Fr ege, " The Thought " 1. I nt r oduct i on I n " Demonst r at i ves' Davi d Kapl an devel ops an account of t he meani ng of i ndexi cal s and sent ences t hat cont ai n t hem based on t he concept s of cont ent , char act er and cont ext . The cont ent of a st at ement i s a pr oposi t i on; whi ch pr oposi t i on a st at ement expr esses depends not onl y on t he
char act er of t he sent ence used, but al so on t he cont ext : who i s speaki ng, t o whom, when, wher e and i n what ci r cumst ances. l i ngui st i c meani ng or
I n hi s essay, Kapl an br i ef l y sket ches an anal ogous concept of bel i ef : we ar e i n a bel i ef st at e wi t h a cer t ai n char act er i n a cer t ai n cont ext and t her eby bel i eve a cer t ai n pr oposi t i on. As Kapl an says, we bel i eve pr oposi t i ons under
char act er s. Kapl an t hen r ai ses t he f ol l owi ng quest i on. Suppose you have f or med a bel i ef of t he sor t t hat you woul d expr ess wi t h a sent ence cont ai ni ng an i ndexi cal , say, " You ar e a comput er sci ent i st " or " Today i s a ni ce day. " W hat do you need t o do t o r et ai n such a bel i ef , af t er you l eave t he cont ext i n whi ch t he sent ence i n quest i on expr esses i t ? W hat do you have t o do t o r et ai n a bel i ef t hat you once expr essed wi t h " You ar e a comput er sci ent i st , " af t er t he hat do you have t o do t o r et ai n t he bel i ef per son you ar e t al ki ng t o has l ef t ? W you once expr essed wi t h " Today i s a ni ce day' af t er t hat day i s gone? Fr ege' s r emar k quot ed above suggest ed a vi ew t o Kapl an: t o r et ai n a bel i ef as one moves i nt o a newcont ext , one must adj ust t he char act er under
14
Ri p van Wi nkl e and Ot her Char act er s
Eur opean Revi ewof Phi l osophy
whi ch one hol ds t he bel i ef , t o a new one t hat conf or ms t o a sent ence t hat woul d expr ess t he same pr oposi t i on i n t he newcont ext . Ther e ar e pat t er ns of char act er change t hat cor r espond t o pat t er ns of change i n cont ext of bel i ef . To cont i nue t o t hi nk t he same t hi ng as t he cont ext of bel i ef changes, i s t o t hi nk under a successi on of char act er s t hat det er mi ne t he same cont ent i n t he successi on of cont ext s, and conf or m t o such a pat t er n. To t hi nk t he t hought I t hought yest er day under t he char act er of " Today i s a ni ce day, " I must nowt hi nk a t hought under t he char act er of " Yest er day i s a ni ce day. " Kapl an r ej ect s t hi s suggest i on because of t he case of Ri p Van Wi nkl e, who f el l asl eep f or t went y year s and woke up t hi nki ng he had sl ept f or onl y a day. Kapl an t hi nks t he Fr ege- i nspi r ed st r at egy woul d l ead us t o deny t hat when Van W i nkl e awoke, he had r et ai ned t he bel i ef he expr essed, t he day he f el l asl eep, wi t h " Today i s a ni ce day. " Howwoul d he expr ess t hi s hel i ef i He mi ght t r y t o expr ess i t wi t h " Yest er day was a ni ce day, " but t hi s woul d f ai l . He woul d not have asser t ed anyt hi ng about t he day he f el l asl eep, but r at her sai d somet hi ng about t he day bef or e he spoke, a day t hat he sl ept t hr ough and of whi ch he has no memor y. I n hi s ar t i cl e " Under st andi ng Demonst r at i ves, " Gar et h Evans f ound t hi s an i nadequat e r eason t o abandon t he Fr ege- i nspi r ed st r at egy. He seems t o t hi nk t hat we shoul d si mpl y say t hat Ri p has f ai l ed t o r et ai n hi s bel i ef : I see no mor e st r angeness i n t he i dea t hat a man who l oses t r ack of t i me cannot r et ai n bel i ef s t han i n t he i dea t hat a man who l oses t r ack of an obj ect cannot r et ai n t he bel i ef s about i t wi t h whi ch he began ( Evans 1981, pp. 87n- 88n. )
speaker i n a l i t er al sense. The cont ext i s a set of f act or s t hat det er mi ne what i ndexi cal s st and f or : t he speaker of t he ut t er ance, t he t i me of ut t er ance, t he pl ace of ut t er ance, t he ci r cumst ance or possi bl e wor l d i n whi ch i t occur s. Char act er s ar e f unct i ons f r omcont ext s t o cont ent s. That i s, a char act er t akes a cont ext as i nput ( as i t s ar gument ) and pr ovi des a cont ent as out put ( as i t s val ue) . These char act er s ar e mat hemat i cal r epr esent at i ons of t he r ul es t hat l anguage assi gns t o expr essi ons. Char act er i s an i nt er pr et at i on of l i ngui st i c meani ng. Suppose t hat at a speci f i c t i me i n 1995- cal l i t t - Kapl an says t o Qui ne, " I l i ve west of you now. " On Kapl an' s t heor y, her e i s what happens: -
-
-
Qui ne. The char act er of ' now" i s a f unct i on f r oma cont ext t o t he t i me of t he cont ext . I n t hi s case, t he cont ent of " now" i s t . The char act er of " l i ve t o t he west of " i s a f unct i on f r om a cont ext t o t he 3- ar y r el at i on x l i ves t o t he west of y at t . Thi s char act er i s not sensi t i ve t o di f f er ences i n cont ext ; i t s cont ent i s t he same at al l cont ext s. 2 The char act er of " I l i ve t o t he west of you now" i s bui l t up out of t he char act er s of t he par t s. I t i s a f unct i on t hat f r oma cont ext t o t he pr oposi t i on t hat a l i ves t o t he west of b at t , wher e a i s t he cont ent of " I " i n t he cont ext , b i s t he cont ent of " you" i n t he cont ext , and t t he cont ent of " now" i n t he cont ext . I n t hi s case, t he cont ent i s t he pr oposi t i on t hat Kapl an l i ves t o t he west of Qui ne at t .
I.
t er mi nol ogy.
2. Li ngui st i c Char act er s and Rol es The cont ent of a st at ement i s a pr oposi t i on, " what i s sai d" by t he
The char act er of " I " i s a f unct i on f r oma cont ext t o t he speaker of t he cont ext . I n t hi s case, Kapl an i s t he speaker , and so he i s t he cont ent of " I " i n t hi s cont ext . ' The char act er of " you" i s a f unct i on f r oma cont ext t o t he per son t hat i s addr essed by t he speaker of t he cont ext , at t he t i me of t he cont ext , i n t he ci r cumst ances of t he cont ext . So t he cont ent of " you" i n t hi s case i s
I n t hi s essay I f i r st r evi ew Kapl an' s t heor y of l i ngui st i c char act er , and t hen expl ai n and mot i vat e a concept of doxast i c char act er . I t hen devel op some concept s f or deal i ng wi t h t he t opi c of bel i ef r et ent i on and t hen, f i nal l y, di scuss Ri p Van Wi nkl e. I come down on Kapl an' s si de wi t h r espect t o t he Fr cgc- i nspi r cd st r at egy, nar r owl y const r ued. But I advocat e somet hi ng l i ke t he Fr ege- i nspi r ed st r at egy, i f i t i s const r ued mor e br oadl y. On my vi ewi t i s r emar kabl y easy t o r et ai n a bel i ef , and I t hi nk Evans i s qui t e wr ong about Ri p and Kapl an. The cent r al concept I devel op, however , t hat of an i nf or mat i on game, i s i n t he spi r i t of much of Evans' wor k. I al so bor r ow some of hi s
15
2.
Kapl an devel ops hi s account wi t hi n possi bl e wor l ds semant i cs, and i n t hat set t i ng t akes t he val ue of an expr essi on l i ke " I " at a cont ext t o be a r i gi d i nt ensi on r at her t han an i ndi vi dual . I n t hi s case, i t woul d be an i ndi vi dual concept t hat pi cked out Kapl an i n each wor l d. Thi s compl i cat i on i s basi cal l y an ar t i f act of t he semant i cal f r amewor k, not par t of t he i nt ui t i ve set of i deas Kapl an t r i es t o convey. I i gnor e t hi s and ot her i mpl i cat i ons. 1 ami gnor i ng t ense.
6
Eur opean Revi ewof Phi l osophy
Kapl an doesn' t assi gn char act er s excl usi vel y t o i ndexi cal s ar i d sent ences t hat cont ai n t hem. Ever y expr essi on get s a char act er . The char act er s of expr essi ons t hat ar e not sensi t i ve t o changes of cont ext ar e const ant f unct i ons. The name " Davi d Kapl an, " f or exampl e, i s a f unct i on t hat r et ur ns Davi d Kapl an at ever y cont ext . I ' l l cal l char act er s l i ke t hi s " l oyal , " and char act er s, l i ke t hat of " I " and " you" t hat ar e sensi t i ve t o changes i n cont ext " f l i ght y. " I l i ke t o i nt er pr et Kapl an' s char act er s i n t er ms of what I cal l ut t er ance- r el at i ve or l i ngui st i c r ol es. Thi s i nt er pr et at i on l i nks Kapl an' s i deas wi t h an ol der t r adi t i on t hat emphasi zes t hat i ndexi cal s ar e " t oken- r ef l exi ve. " 3 Li ngui st i c convent i ons assi gn a name l i ke " Davi d Kapl an, " t o a par t i cul ar i ndi vi dual . But t hey assi gn an i ndexi cal l i ke " I " or " you" t o an ut t er ance- r el at i ve ( or l i ngui st i c) r ol e: bei ng t he speaker of u, ei ng- t h bei ng t he addr essee of u, et c. An ut t er ance of an i ndexi cal does not st and f or or r ef er t o t he r ol e assi gned t o i t ; i t st ands f or or r ef er s t o t he obj ect t hat pl ays t hat r ol e, r el at i ve t o t he ut t er ance i t sel f .
W hy ar e i ndexi cal s i mpor t ant ? The answer has t wo par t s. Fi r st , l i ngui st i c r ol es ar e cl osel y associ at ed wi t h ot her r ol es t hat obj ect s pl ay i n our l i ves. The speaker of t he ut t er ance I hear i s of t en t he per son I aml ooki ng at ; t he pl ace an ut t er ance i s made i s usual l y t he pl ace t he speaker occupi es, and usual l y near t he pl ace t he l i st ener occupi es. So, when I l ear n t hat an obj ect pl ays a l i ngui st i c r ol e r el at i ve t o an ut t er ance I hear , I l ear n about ot her r ol es t hat i t pl ays. Gi ven t hi s, i ndexi cal s ar e usef ul i n t wo si t uat i ons. Somet i mes one want s t o know mor e about an obj ect t hat pl ays some l i ngui st i c r ol e or associ at ed r ol e. I want t o knowmor e about t he per son I see bef or e me. I ask: " W ho ar e C om m on sense and at assur e m e t hat t he you?" my f aci l i t y l anguage per son I see bef or e me wi l l be t he addr essee of my ut t er ance. So i f I f i nd out whot he addr essee i s, I f i nd out who t he per son bef or e me i s. I want t o know t he name of t he ci t y I f i nd i t sel f i n. I ask: " W hat i s t he name of t hi s ci t y?" Common sense t el l s me t hat t he ci t y I ami n i s t he ci t y my ut t er ance wi l l t ake pl ace i n. So i f I f i nd out t he name of t he ci t y i n whi ch my ut t er ance occur s, I wi l l f i nd out t he name of t he ci t y i n whi ch I f i nd i t sel f . Somet i mes I need t o know what r ol e some obj ect i s pl ayi ng i n my l i f e, what i t s cur r ent r el at i on t o me i s. I ask, " W ho i s Davi d Kapl an?" You answer , 3.
See, i n par t i cul ar , Rei chenbach ( 1947) , Bur ks ( 1949) , and Per r y ( 1979) .
Ri p van Wi nkl e and Ot her Char act er s
17
poi nt i ng: " That man i s Davi d Kapl an. " I want ed t o know whi ch of t he peopl e I coul d see was Davi d Kapl an. Common sense t ol d you t hat I woul d be abl e t o i dent i f y t he per son you demonst r at ed as one of t he ones I saw. I ask, "W hen i s Jul y 4t h?" You answer , " Tomor r ow. " Common sense t el l s you t hat I wi l l r eal i ze t hat your ut t er ance occur s at t he pr esent t i me, and so t hat t he day I ami nt er est ed i n i s j ust 24 hour s away. W hen we hear an i ndexi cal , t he f i r st way of t hi nki ng of t he r ef er ent t hat i s af f or ded t o us, i s t hi nki ng of i t as t he t hi ng t hat pl ays t he ut t er ance- r el at i ve r ol e. So when Qui ne hear s Kapl an' s ut t er ance uof "I l i ve t o t he west of you now' t he f i r st way of t hi nki ng of Kapl an pr ovi ded by t hi s ut t er ance i s as t he speaker of t he ut t er ance of "I . " Qui ne' s f i r st gr asp of t he t r ut h condi t i ons of u i s somet hi ng l i ke, " The speaker of u l i ves t o t he west of t he addr essee of u at t he t i me of u. " Thi s phase of under st andi ng usual l y sl i ps t hr ough one' s mi nd wi t hout st oppi ng, per haps wi t hout r i si ng t o consci ousness, as t he ut t er ance- r el at i ve r ol es gi ve way t o mor e i nt er est i ng associ at ed r ol es. I n nor mal ci r cumst ances Qui ne wi l l r eal i ze t hat t he per son he sees i s t he speaker of t he ut t er ance he hear s, t hat t hat per son i s Davi d Kapl an, t hat Kapl an i s addr essi ng hi m, and t hat t he t i me of ut t er ance i s, f or al l pr act i cal pur poses, t he t i me t he ut t er ance i s hear d. So Qui ne wi l l t hi nk somet hi ng l i ke, " Kapl an l i ves west of me now. " I n ot her cases, however , t he ut t er ance- r el at i ve r ol e may be our onl y way of t hi nki ng of t he obj ect s an ut t er ance i s about , at l east unt i l we have done some det ect i ve wor k. I f i nd a f aded not e i n my ol d copy of Wi t t genst ei n' s I nvest i gat i ons: " You ar e bei ng an ass. " I knowt hat t he not e i s t r ue i f and onl y i f t he per son t o whomi t was addr essed was bei ng an ass at t he t i me i t was wr i t t en. That may be t he onl y gr asp of t he t r ut h- condi t i ons I can get , unt i l I l ook t hr ough some paper s and r ecor ds and wr ack my br ai ns. Then I r emember : someone passed t he not e t o me, appar ent l y t hi nki ng my r emar ks i n a semi nar on t he pr i vat e l anguage ar gument wer e l ess pr of ound t haj l di d. NowI knowt hat t he not e- or mor e accur at el y, t he ut t er ance f or whi ch t he not e was t he t oken, was t r ue i f and onl y i f I was bei ng an ass. Kapl an' s syst em al l ows t he possi bi l i t y t hat we can say t he same t hi ng ( ut t er st at ement s wi t h t he same cont ent ) i n qui t e di f f er ent ways ( usi ng sent ences wi t h qui t e di f f er ent char act er s) . Thi s happens i f we use t he sent ences i n cont ext s wher e, due t o par t i cul ar and per haps pecul i ar ci r cumst ances, t hey t ur n out t o have t he same cont ent . Thi s happens when t he same obj ect
18
Eur opean Revi ew of Phi l osophy
pl ays t wo qui t e di f f er ent ut t er ance- r el at i ve r ol es. And i t can happen wi t hout our knowi ng i t . Kapl an asks us t o consi der t he sent ences " My pant s ar e on f i r e" and " Hi s pant s ar e on f i r e. " The char act er of t he f i r st i s a f unct i on t hat f or a cont ext wi t h a as t he speaker , r et ur ns t he pr oposi t i on t hat a' s pant s ar e on f i r e. The char act er of t he second i s a f unct i on t hat f or a cont ext wi t h a as speaker , i n a wor l d i n whi ch a i s poi nt i ng at b or ot her wi se cal l i ng at t ent i on t o hi m, ( " demonst r at i ng b" ) r et ur ns t he pr oposi t i on t hat Us pant s ar e on f i r e. Suppose, t o cont i nue wi t h Kapl an' s exampl e, t hat he says, " My pant s ar e on f i r e. " The cont ent of hi s r emar k i s t he pr oposi t i on t hat Davi d Kapl an' s pant s ar e on f i r e. Suppose nowt hat he sees hi msel f i n a mi r r or , doesn' t r eal i ze t hat he i s seei ng hi msel f , and, poi nt i ng at t he man i n t he mi r r or , says " Hi s pant s ar e on f i r e. " The sent ence i s qui t e di f f er ent , wi t h a qui t e di f f er ent char act er , but i n t hi s par t i cul ar cont ext t he cont ent i s t he same, t hat Davi d Kapl an' s pant s ar e on f i r e. Kapl an has sai d t he same t hi ng i n t wo di f f er ent ways wi t hout knowi ng i t , al t hough pr esumabl y he wi l l r ecogni ze what i s goi ng on bef or e l ong. Nowconsi der t he di f f er ence bet ween ( 1) ( 2) ( 3)
My pant s ar e on f i r e. Hi s pant s ar e on f i r e. Davi d Kapl an' s pant s ar e on f i r e on Jul y 4, 1984 at 5 p. m.
One can t hi nk of ( 1) and ( 2) as t ool s. ( 1) i s a t ool f or sayi ng t hat onesel f has bur ni ng pant s. ( 2) i s a t ool f or sayi ng of someone t hat one can see and demonst r at e, t hat t hey have bur ni ng pant s. Si nce l ot s of peopl e coul d, i n pr i nci pl e at l east , f i nd t hemsel ves wi t h bur ni ng pant s, or f i nd t hemsel ves i n a posi t i on t o demonst r at e someone t hat has bur ni ng pant s, t hese t ool s mi ght be used, agai n and agai n, i n di f f er ent si t uat i ons, t o say di f f er ent t hi ngs about di f f er ent peopl e. ( 3) doesn' t cont ai n any i ndexi cal s, j ust names. Even t he ver b can be t aken as t ensel ess, gi ven t he way t he dat e i s f i l l ed i n. As we sai d, Kapl an assi gns al l expr essi ons a char act er , not j ust i ndexi cal s, but f or non- i ndexi cal s t he char act er doesn' t do much. I t ' s j ust a way of maki ng t he t heor y wor k smoot hl y. So " Kapl an, " f or exampl e, i s assi gned a char act er t hat has Davi d Kapl an as cont ent i n al l cont ext s. And ( 3) as a whol e has a char act er t hat has t he same pr oposi t i on as cont ent i n al l cont ext s, t he pr oposi t i on t hat Kapl an has bur ni ng pant s on Jul y 4, 1984. Thus t he char act er of ( 3) i s ver y l oyal . I t
Ri p van Wi nkl e and Ot her Char act er s
19
st i cks wi t h t he same cont ent i n cont ext af t er cont ext . To put t he poi nt anot her way, wi t h ( 3) we have a t ool t hat al l ows us t o expr ess t he same pr oposi t i on, no mat t er when i t i s or wher e we ar e or t o whomwe ar e t al ki ng. As a t ool , ( 3) can seema bi t odd. W hy woul d we need or want a t ool f or of one sayi ng, j ust per son, t hat he has bur ni ng pant s? I t seems l i ke a ver y speci al pur pose t ool . We' l l r et ur n t o t hi s quest i on l at er .
3. Doxast i c Char act er s Howcan we concei ve of bel i ef s, so t hat char act er s may be i nt el l i gi bl y assi gned t o t hem? One concept i on i s t hat of bel i ef as an at t i t ude t owar ds a sent ence; t he bel i ef i nher i t s t he char act er of t he sent ence at whi ch i t i s di r ect ed. I t hi nk t hi s i s a r at her unsat i sf act or y concept i on. For one t hi ng, i t seems t hat f ai r l y compl ex t hought s about t hi ngs of al l sor t s woul d 4 pr ecede t hought s about sent ences, bot h l ogi cal l y and psychol ogi cal l y. A mor e sat i sf act or y way of concei vi ng of bel i ef s st ar t s wi t h t he i dea t hat t hey ar e concr et e cogni t i ve st r uct ur es t hat ar i se i n one' s mi nd i n cer t ai n si t uat i ons; t r aces, as i t wer e, of exper i ences of per cept i on, l ear ni ng, and i nf er ence. These st r uct ur es have cont ent ; when one has a bel i ef , t her e i s ( at l east when t hi ngs go r i ght ) , somet hi ng t hat one bel i eves, a pr oposi t i on P such t hat one bel i eves P. Bel i ef s so concei ved wi l l have t wo aspect s t hat must mesh, causal r ol e and The bel i ef s ar e caused by cer t ai n ki nds of per cept i ons, and cause cer t ai n ki nds of act i ons. But t hey al so have a cer t ai n cont ent ; t hey ar e bel i ef s
cont ent .
t hat such and such i s t he case, bel i ef s t hat P f or some pr oposi t i on P. These t wo aspect s shoul d mesh. Abel i ef t hat P shoul d cause act i on appr opr i at e t o i t s bei ng t he case t hat P. gi ven one' s desi r es and ot her bel i ef s. 4.
I ment i on t hi s concept i on mai nl y because many r easonabl y at t ent i ve r eader s seemt o f i nd i t i n " The Pr obl emof t he Essent i al I ndexi cal " ( Per r y 1979). But i t was not what I had i n mi nd. See " Bel i ef and Accept ance" ( Per r y 1980). 1 t hi nk t hi s mi si nt er pr et at i on has t hr ee r oot s. 1 di dn' t pr oper l y di st i ngui sh bet ween l i ngui st i c and cogni t i ve r ol es. I under est i mat ed t he ease wi t h whi ch peopl e woul d suppose t hat my vi ew, t hat sent ences wi t h i ndexi cal s ( or t hei r char act er s) , wer e mor e adequat e ways of cl assi f yi ng bel i ef st at es t han wer e pr oposi t i ons, woul d have t o r est on t he vi ew t hat sent ences ( or t hei r char act er s) wer e what bel i ef s wer e di r ect ed at . Fi nal l y t he pr obl em r ef er r ed t o i n t he t i t l e had t o do wi t h t he f act t hat i ndexi cal s scent ed essent i al t o expr essi ng cer t ai n t hought s; f r om t hi s some r eader s seemt o have assumed t hat I t hought t hat i ndexi cal s wer e necessar y f or havi ng t hose t hought s.
20 Eur opean Revi ew of Phi l osophy By t he causal r ol e of a st at e I mean t he var i ous combi nat i ons of f act or s t hat br i ng t he st at e about , and t he var i ous combi nat i ons of f act or s i t br i ngs about i n t ur n. Consi der t he st at e of bei ng nauseat ed. Thi s st at e i s uni ver sal and par t i al . To say i t i s uni ver sal i s t o say t hat di f f er ent peopl e at di f f er ent t i mes can be i n t he same st at e. You wer e nauseat ed l ast week, af t er a r i de on t he Fer r i s Wheel . I was nauseat ed yest er day, af t er bi ngi ng on sushi andhambur ger s. To say i t i s par t i al i s t o say t hat bei ng nauseat ed const i t ut es onl y par t of one' s t ot al st at e at a gi ven t i me. I was nauseat ed and embar r assed and gui l t - r i dden and i n a number of ot her st at es at t he same t i me. Each case of nausea has i t s own speci f i c causes. But t her e ar e pat t er ns. Cer t ai n combi nat i ons of f act or s br i ng about nausea, and nausea, t oget her wi t h ot her f act or s, has cer t ai n r esul t s. Thi s pat t er n i s t he causal r ol e of nausea. On a gi ven occasi on, t he causes of a speci f i c case of nausea may be pr et t y si mi l ar t o t he causes of some ot her st at e. Per haps t he mai n cause of my nausea was pr et t y much t he same t hi ng as my f eel i ngs of gui l t : massi ve over eat i ng. But t he ef f ect s of t he t wo st at es ar e di f f er ent . My nausea l eads t o me t o t ake Al ka- Sel t zer ; my gui l t l eads me t o t ur n on t he t el evi si on. And i n gener al , t he pat t er ns ar e not t he same. Lot s of t hi ngs cause nausea t hat don' t cause gui l t and vi ce ver sa, and nausea, i n combi nat i on wi t h var i ous f act or s, causes l ot s of t hi ngs t hat gui l t , i n combi nat i on wi t h t hose same f act or s, woul dn' t cause. One can t hi nk of causal r ol es i n var i ous st r i ct and phi l osophi cal ways or i n a sor t of l oose and casual way. The second wi l l suf f i ce f or my pur poses. Thi nk of t he causal r ol e of a st at e as i t s t ypi cal causes, t he t hi ngs you woul d expect mi ght cause an i nst ance of t hat st at e, i n mor e or l ess or di nar y ci r cumst ances, and si mi l ar l y wi t h i t s t ypi cal ef f ect s. W e sai d t hat t he causal r ol e of a bel i ef shoul d mesh wi t h i t s cont ent . But Kapl an' s case shows t hat t hi s meshi ng i s not si mpl e and di r ect . Fi r st , not e t hat t he causal r ol es of t he st at es we i magi ne t o occur ar e qui t e di f f er ent . I f we hear d t hat someone bel i eved t hat t hei r ver y own pant s wer e on f i r e, we woul d expect t hat somet hi ng l i ke t hi s happened. Fi r st , t hey dr opped an ash f r om a pi pe i nt o t hei r own l ap, or st ood t oo cl ose t o a f i r epl ace, or somet hi ng l i ke t hat . Then t hey f el t some unaccust omed war mt h i n t hei r net her r egi ons. Then per haps t hey smel l ed somet hi ng l i ke wool or cot t on or r ayon bur ni ng. Then t hey l ooked down' wher e t hi ngs wer e heat i ng up, and sawt he smoke. Those ar e t ypi cal causes of t he st at e one i s i n when
Ri p van Wi nkl e and Ot her Char act er s 21 one says, " My pant s ar e on f i r e. " The t ypi cal ef f ect s woul d i ncl ude st r ong emot i ons l i ke f ear , and at t empt s t o douse onesel f wi t h wat er or put out t he f i r e i n some ot her way, and sayi ng, " My pant s ar e on f i r e. " Nowconsi der t he st at e Kapl an was i n when he sawt he man i n t he mi r r or , who j ust happened t o be hi m, wi t h hi s pant s on f i r e. Thi s st at e i s t ypi cal l y caused by seei ng smoke and f l ames er upt f r omt he t r ouser s of someone st andi ng i n f r ont of one. I t t ypi cal l y l eads t o concer n and t he at t empt t o hel p and shout i ng, " Hi s pant s ar e on f i r e. " So t he t ypi cal causes of t hi s st at e ar e qui t e di f f er ent f r omt hat of t he f i r st st at e. Next , not e t hat t hese di f f er ent bel i ef st at es do not l i ne up di r ect l y wi t h pr oposi t i ons bel i eved, so t her e i s not a si mpl e and di r ect meshi ng bet ween causal r ol e and cont ent . W e can i magi ne a l ot of peopl e bei ng i n t he " My pant s ar e on f i r e" bel i ef st at e; t hey woul dn' t t her eby bel i eve t he same pr oposi t i ons. Kapl an woul d bel i eve t hat Kapl an' s pant s wer e on f i r e, Sear l e woul d bel i eve t hat Sear l e' s pant s wer e on f i r e, and so f or t h. And we can i magi ne a l ot of peopl e bei ng i n t he " Hi s pant s ar e on f i r e" st at e. They al l have somet hi ng i n common, but i t ) not what t hey bel i eve- i t ' s not t he cont ent . They woul d bel i eve qui t e di f f er ent pr oposi t i ons, dependi ng on whomt hey wer e l ooki ng at . At t hi s poi nt , i t i s ver y hel pf ul t o post ul at e a l evel of meani ng f or bel i ef s t hat i s anal ogous t o but not der i ved f r om t hat of char act er f or sent enceschar act er s t hat i n f act ar e qui t e i ndependent of l anguage. Abel i ef wi l l have a cer t ai n cont ent because ( i ) i t has a cer t ai n char act er , and ( i i ) i t occur s i n a cer t ai n bel i ef cont ext - i t i s a bel i ef hel d by a cer t ai n per son, at a cer t ai n t i me, i n a cer t ai n pl ace, at t endi ng t o cer t ai n obj ect s, et c. The vi r t ue of t hi nki ng of a bel i ef i n t hi s way, i s t hat i t makes i nt el l i gi bl e howt hi s meshi ng bet ween causal r ol e andcont ent coul d t ake pl ace, i n a syst emat i c psychol ogy of cont ent , f or t hese char act er s wi l l cor r espond mor e cl osel y t o bel i ef st at es t han pr oposi t i ons do. Awi der cl ass of r ol es i s needed t o char act er i ze bel i ef s t han i s needed t o char act er i ze ut t er ances. W e not ed t hat t he char act er s t hat ar e associ at ed wi t h i ndexi cal s ar e based on ut t er ance- r el at i ve r ol es. These r ol es ar e based on r el at i ons obj ect s mi ght have t o a gi ven ut t er ance. To char act er i ze bel i ef s we need what we mi ght cal l t hought - r el at i ve or cogni t i ve r ol es. These r ol es ar e based on r el at i ons t hat an obj ect can have t o a gi ven epi sode of t hought or a par t i cul ar bel i ef , such as bei ng t he owner of t he t hought , ( t he sel f r ol e) ,
22
Eur opean Revi ewof Phi l osophy
bei ng at t ended t o by t he owner of i t , bei ng r emember ed by t he owner of i t , bei ng hel d i n t he r i ght hand of t he owner of i t , bei ng above t he owner of i t , bei ng sat on by t he owner of i t , and so f or t h. I ' l l i ndi cat e doxast i c char act er s by quot i ng t he sent ence a speaker mi ght use t o communi cat e t he bel i ef s one has under t he char act er . Of cour se t he char act er s of t hese sent ences ar e l i ngui st i c r at her t han doxast i c; I suppl ement t he i ndexi cal expr essi ons wi t h br acket ed mat er i al t o i dent i f y t he under l yi ng cogni t i ve r ol e i nvol ved. For exampl e, " t hat man { t he one I see] " and " t hat man [ t he one I r emember ] , " " I [ sel f ] " and " now [ t he moment of t hought ] . Ther e i s a di f f er ence bet ween r ef er r i ng t o someone as " you" and desi gnat i ng t hemwi t h t he phr ase " t he per son I am addr essi ng, " even t hough t he r ol e of bei ng t he addr essee i s i nvol ved i n bot h cases. W hen one r ef er s t o a per son as " you, " one does not say t hat one i s addr essi ng t hem; one expl oi t s t hat f act t o r ef er t o t hem. One can l ear n how t o use " you, " wi t hout bei ng abl e t o expl i ci t l y f or mul at e t he condi t i ons under whi ch a use of t hat wor d r ef er s t o a per son ( per haps because one has not t hought about i t , or has not yet at t ai ned t he concept s, such as t he concept of a wor d and t he concept of r ef er ence, t hat woul d be necessar y t o t hi nk about i t ) . Even so, one has some gr asp of t hese condi t i ons; one has some sor t of posi t i ve doxast i c at t i t ude t owar ds t he f act t hat a use of " you" r ef er s t o t he addr essee. But one may not be pr oper l y sai d t o bel i eve t hat t he r ef er ence of " you" i s t he addr essee. One knows howt o use " you" t o r ef er t o t he addr essee, even t hough one may not know t hat a use of " you" r ef er s t o t he addr essee. I n t hese cases, I say t hat one i s at t uned t o t he way " you" wor ks, Si mi l ar l y t her e i s a di f f er ence bet ween bei ng abl e t o t hi nk of a t hi ng or per son i n vi r t ue of some r ol e i t pl ays i n one' s l i f e, and bei ng abl e t o ar t i cul at e t hat r ol e i n t hought or speech and t hi nk of i t as t he t hi ng or per son pl ayi ng t hat r ol e i n one' s l i f e. Consi der a chi l d, who i s t hi nki ng about a dog t hat she sawan hour or so bef or e. She has a cer t ai n memor y of t he dog, and i t i s i n vi r t ue of t hi s memor y t hat she i s abl e t o t hi nk of t he dog. I woul d r epr esent t he char act er of her t hought wi t h " That one I r emember ] [ t he dog was ver y cut e. " Thi s i s not t he same as " The I r em ember was t hi nki ng dog ver y cut e. " The chi l d mi ght l ack t he concept s needed t o t hi nk t hi s t hought ; even i f she has t hem, i t may t ake a bi t of t i me and wi t t o f i gur e out t hat t he dog can be char act er i zed as t he one she r emember s. St i l l , t he chi l d woul d be at t uned t o t he f act t hat t he dog she i s t hi nki ng of i s one t hat she r emember s, i n t hat she woul d know howt o consul t her memor y f or mor e i nf or mat i on
Ri p van Wi nkl e and Ot her Char act er s
23
about t he dog: t o f i nd t he dog she woul d go back t o wher e she had l ast seen i t , f or exampl e. e i magNowl et ' s r et ur n once agai n t o Kapl an and hi s bur ni ng pant s. W med Kapl an maki ng t wo set s of obser vat i ons of hi s own bur ni ng t r ouser s. W e i magi ned hi musi ng t he sent ences, " My pant s ar e on f i r e" and " Hi s pant s ar e on f i r e, " t o expr ess what he bel i eved. But he mi ght make t he obser vat i ons, and acqui r e t he bel i ef s, wi t hout sayi ng anyt hi ng. Even i f t hey don' t l ead t o di st i nct ut t er ances, t he bel i ef st at es ar e qui t e di f f er ent . I f Kapl an not i ces t hat hi s pant s ar e on f i r e i n t he usual way, he wi l l be i n qui t e a di f f er ent st at e t han i f he not i ces a man wi t h bur ni ng pant s i n a mi r r or . The di f f er ence i n t he act i ons he woul d t ake i n t hese ci r cumst ances, i ncl udi ng t he di f f er ence i n sent ences he woul d ut t er i f he wer e t o put what he not i ced i nt o wor ds, r ef l ect s a di f f er ence i n bel i ef s. The concept of bel i evi ng under a char act er i s i nt ended t o capt ur e t hi s di f f er ence. The di f f er ence i n bel i ef woul d be t her e, even i f Kapl an di dn' t say anyt hi ng. And t he di f f er ence bet ween t he t wo cases i s not t he pr oposi t i on t hat i s bel i eved, but t he char act er under whi ch i t i s bel i eved. Ther e ar e t wo char act er s her e, one cor r espondi ng t o t hat of " My[ sel f ] pant s ar e on f i r e" and one cor r espondi ng t o t hat of " Hi s [ t he man I aml ooki ng at ] pant s ar e on f i r e. " These char act er s come much cl oser t o l i ni ng up wi t h causal r ol es t han do t he pr oposi t i ons bel i eved. Thus t wo t hi ngs st and out at t he l evel of char act er t hat get obl i t er at ed at t he l evel of cont ent . Fi r st , t he di f f er ence i n causal r ol e of t he t wo bel i ef s. Sec- / ( ond, t he common nat ur e t hat di f f er ent bel i ef s wi t h di f f er ent cont ent s j) bel ongi ng t o di f f er ent peopl e at di f f er ent t i mes mi ght have, and i n vi r t ue of whi ch t hese di f f er ent si t uat i ons mi ght i nst ant i at e t he same psychol ogy of cont ent . Recal l t he di st i nct i on bet ween f l i ght y and l oyal char act er s f r omour di scussi on of l i ngui st i c char act er s. Loyal char act er s yi el d t he same cont ent f r omany cont ext ; t hese char act er s ar e t he sor t t hat bel ong t o sent ences t hat don' t cont ai n i ndexi cal s. Fl i ght y char act er s yi el d di f f er ent cont ent s i n di f f er ent cont ext s; t hese char act er s ar e t he sor t t hat bel ong t o sent ences t hat cont am i ndexi cal s. The same di st i nct i on appl i es t o doxast i c char act er s. " Davi d Kapl an has bur ni ng pant s at 4 p. m. on Jul y 4, 1996" i s a ver y l oyal char act er ; " I [ sel f ] have bur ni ng pant s now [ t he pr esent moment ] " i s a ver y f l i ght y char act er , expr essi ng a di f f er ent pr oposi t i on wi t h each var i at i on i n t i me or t hi nker . I f we know t hat an ut t er ance ent ence wi t h a f l i ght y char act er
I
a/
24
Eur opean Revi ewof Phi l osophy
expr esses a cer t ai n pr oposi t i on, we may knowqui t e a bi t about t he cont ext i n whi ch t he ut t er ance occur r ed. I f an ut t er ance of " I ama l ogi ci an" has t he cont ent t hat Davi d Kapl an i s a l ogi ci an, t hen t he speaker must be Davi d Kapl an. I f an ut t er ance of " That man wor ks f or Mi cr osof t now" has t he cont ent t hat El wood Fr i t chey wor ks f or Mi cr osof t on Oct ober 27, 1995, t hen t he ut t er ance occur r ed on Oct ober 27 and t he speaker was someone i n a posi t i on t o demonst r at e El wood Fr i t chey. The obj ect s t hat get r ef er r ed t o by t he i ndexi cal s must be pl ayi ng cer t ai n l i ngui st i c r ol es, and hence cer t ai n r ol es i n t he l i f e of t he speaker . The same i s t r ue wi t h bel i ef s. 1f f bel i eve t hat Davi d Kapl an has bur ni ng pant s under t he char act er " t hat man [ t he one I see] has bur ni ng pant s, " t hen I must be l ooki ng at Davi d Kapl an. I f anyone bel i eves t hat pr oposi t i on under t he char act er " I ( sel f ] have bur ni ng pant s, " t hat per son must be Davi d Kapl an. I f my bel i ef s ar e under char act er s, i t seems t hat many of t he char act er s t hey ar e under ar e ver y l oyal and not ver y f l i ght y. Ri ght nowI bel i eve t hat Davi d Kapl an i s a l ogi ci an. I amnot speaki ng t o hi m, seei ng hi m, or hear i ng hi mr i ght now. He i s pl ayi ng no r ol e i n m y l i f e t hat woul d enabl e me t o r ef er hi m by ut t er i ng " you" or by demonst r at i ng hi mand ut t er i ng " he. " W hat char act er do I bel i eve t hese t hi ngs under , and what i s t he poi nt of such bel i ef s? W hat causal r ol e, i f any, do t hey have? W hy do peopl e have such bel i ef s, and why do ot her s car e whet her or not t hey do? 4. I nf or mat i on Games W e ended each of t he l ast t wo sect i ons by wonder i ng about l oyal char act er s. W hy do we want sent ences t hat have l oyal char act er s- what use do we have f or such speci al pur pose t ool s? And what sor t s of bel i ef s have l oyal char act er s, and why ar e t hey i mpor t ant ? To consi der t hese quest i ons, I want t o i nt r oduce t he concept of an i nf or mat i on game. An i nf or mat i on game i nvol ves t he acqui si t i on and l at er appl i cat i on of a bel i ef about an i ndi vi dual . That i s, at some t i me one comes t o bel i eve somet hi ng about some per son or obj ect . Then, l at er , t hat bel i ef gui des one' s behavi or t owar ds t hat obj ect or at l east an obj ect t hat one t akes t o be t he same as i t . I cal l t he obj ect about whi ch one acqui r es t he bel i ef t he sour ce. 5 I cal l t he obj ect t o whi ch one appl i es t he bel i ef - t he obj ect one t akes t o be t he 5.
See Evans ( 1973) f or t he t er m" sour ce" used i n mor e or l ess t hi s way.
Ri p van Wi nkl e and Ot her Char act er s
25
sour ce- t he appl i candum. I n any i nf or mat i on game, one f aces t he pr obl em of maki ng r easonabl y sur e t hat t he sour ce i s t he appl i candum. Suppose I meet El wood Fr i t chey at a par t y; he t el l s me he i s a pr ogr ammer f or Mi cr osof t and I bel i eve hi m. So I acqui r e a bel i ef . Lat er I ask hi mi f he knows why t he Maci nt osh ver si on of W or d 6. 0 wor ks so sl owl y. I di r ect t hi s quest i on t o hi mbecause I bel i eve hi mt o be a Mi cr osof t pr ogr ammer . I f I di dn' t bel i eve t hat , I woul dn' t ask hi m. I t doesn' t make much sense t o ask some r andom per son t hi s quest i on. M y bel i ef t hat he i s a Mi cr osof t pr ogr ammer i s par t of t he r eason I ask hi m; my aski ng hi mi s an appl i cat i on of t he bel i ef . I n an i nf or mat i on game I acqui r e a bel i ef , and t hen l at er I appl y t he bel i ef I acqui r ed t o a cer t ai n obj ect : because I have t he bel i ef , and t ake t he obj ect t o be t he one t he bel i ef i s about , I deal wi t h i t i n a cer t ai n way. I n t hi s case, I ask a per son a cer t ai n quest i on, because I t hi nk he i s i n a posi t i on t o knowt he answer . I wi l l descr i be ei ght i nf or mat i on games, whi ch I cal l " st r ai ght - t hr ough, " " t r acki ng, " " det ach- and- r ecogni ze, " " updat i ng, " " r ecol l ect i on, " " i nf er ence, " " pl anni ng, " and " communi cat i on. " I i nt r oduce t he f i r st t hr ee by consi der i ng t hr ee scenar i os f or what hap_pens i n bet ween my acqui r i ng t he bel i ef and my appl yi ng i t . Fi r st , vi r t ual l y not hi ng comes bet ween. I am t al ki ng t o Fr i t chey, f ace t o f ace. He says t o me, " 1 ama pr ogr ammer at Mi cr osof t . " I acqui r e a bel i ef I woul d expr ess wi t h " You ar e a pr ogr ammer at Mi cr osof t . " I say, on t he basi s of t hat bel i ef , " W or d 6. 0 r un so sl owl y?" hy does t he Maci nt osh ver si on of W Thi s i s t he st r ai ght - t hr ough i nf or mat i on game. I st and i n a cer t ai n r el at i onshi p or f ami l y of r el at i onshi ps t o Fr i t chey. He i s i n f r ont of me; I am l ooki ng at hi m; I amt al ki ng t o hi m; he i s t al ki ng t o hen someone has t hese r el at i ons t o me, or as I shal l put i t , occupi es me. W t hese agent - r el at i ve r ol es i n my l i f e, I have ways of f i ndi ng out i nf or mat i on about hi m. ( I ' l l say t hat t her e ar c cpi st cmi c t echni ques associ at ed wi t h t he r ol e, or f or shor t , t hat i t i s an epi st emi c r ol e. ) I can l ook, and I can ask and l i st en t o t he answer , t o ment i on t he most obvi ous ways of gat her i ng i nf or mat i on about t he per son one i s t al ki ng t o f ace- t o- f ace. Ther e ar e al so ways of act i ng, t he success of whi ch wi l l t ur n on t he char act er i st i cs of t he per son t hat pl ays t hi s r ol e. ( I ' l l say t hat t her e ar e pr agmat i c t echni ques associ at ed wi t h t he r ol e, or f or shor t t hat i t i s a pr agmat i c r ol e. ) The act i on of aski ng, or d 6. 0 wor k so sl owl y, " wi t h t he goal of f i ndi ng out t he "W hy does W
26
Eur opean Revi ewof Phi l osophy
answer , has a chance of bei ng successf ul i f t he per son I amt al ki ng t o i s a pr ogr ammer at Mi cr osof t . I n t he st r ai ght - t hr ough i nf or mat i on game, maki ng sur e t hat t he sour ce i s t he appl i candumi s not a bi g pr obl em. The epi st emi c r ol e and t he pr agmat i c r ol e ar e cl osel y associ at ed and t he t i me di f f er ence bet ween acqui si t i on and appl i cat i on i s ver y shor t . Ther e i s not t i me f or one obj ect t o cease t o pl ay i t and anot her obj ect t o t ake i t s pl ace. I n par t i cul ar , I don' t need t o know a l ot about Fr i t chey t o successf ul l y pl ay t he st r ai ght - t hr ough i nf or mat i on game. I don' t need t o knowhi s name, or have any way of r ecogni zi ng hi m. I j ust need t o be abl e t o t el l t hat onl y one obj ect pl ays t he r ol e i n quest i on i n my l i f e f or t he f ewmoment s i t t akes f or t he game t o be pl ayed. Second case. I br eak of f my conver sat i on wi t h Fr i t chey, af t er l ear ni ng t hat he i s a Mi cr osof t pr ogr ammer but bef or e aski ng hi mabout t he Maci nt osh ver si on of W or d 6. 0. But I keep my eye on hi mas he and I move our var i ous ways ar ound t he par t y. Ot her peopl e assume t he r ol e i n my l i f e t hat Fr i t chey had: t hat i s, I conver se wi t h t hem. Fr i t chey pl ays a successi on of r ol es: obj ect t o my l ef t , obj ect t o my r i ght , obj ect I am gl anci ng at . Lat er we end up f ace t o f ace and I ask hi m, " W or d 6, 0 r un so sl owl y. " hy does W Her e t he t ask of maki ng sur e t hat t he appl i candumi s t he sour ce i s mor e compl i cat ed. I have t o t r ack Fr i t chy, t o use anot her apt t er m of Evans' s ( Evans 1981) . Thi s r equi r es t hat I be abl e t o ascer t ai n t hat a si ngl e obj ect has pl ayed a successi on f r ol es i n my l i f e. Thi s r equi r es mor e t han i s r equi r ed f or t he st r ai ght - t hr ough l anguage game, but i t doesn' t r equi r e t hat I have t he abi l i t y t o r ecogni ze or r e- i dent i f y Fr i t chy i f I l ose t r ack of hi m. I n t he st r ai ght - t hr ough and t r acki ng games, t he pl ayer st ays at t ached t o t he sour ce, i n t he sense t hat he r emai ns i n an epi st emi c r el at i on t o t he sour ce f r omacqui si t i on t o appl i cat i on. I n t he st r ai ght - t hr ough game i t i s t he same r el at i on and vi r t ual l y t he same t i me. I n t he t r acki ng game i t i s a successi on of r el at i ons t hr ough an i nt er val of t i me. The t hi r d game I cal l t he " det ach- and- r ecogni ze" game. I l ear n t hat Fr i t chy i s a pr ogr ammer f or Mi cr osof t . Then I go home. I don' t seemhi m f or days or even year s. At some l at er poi nt I see hi magai n. I r ecogni ze hi m, r emember hi s j ob, and ask hi m, " W or d hy does t he Maci nt osh ver si on of W 6. 0 r un so sl owl y?" Af t er I l eave t he par t y, Fr i t chey i s not pl ayi ng an epi st emi c or pr agmat i c r ol e i n my l i f e. He i s of cour se st i l l r el at ed t o me. He i s a cer t ai n di st ance and
Ri p van Wi nkl e and Ot her Char act er s 27 di r ect i on f r omme, but I amnot awar e of i t ; i t i s not par t of my concept i on of hi m. Ther e ar e var i ous act i ons I coul d t ake t o f i nd out t hi ngs about hi m, or t o have an ef f ect on hi m. But t her e i s no si mpl e t echni que, l i ke l ooki ng or aski ng a quest i on t hat I can use t o f i nd out mor e about hi m, and no si mpl e t echni que l i ke t al ki ng t o hi mor shovi ng hi mor gest ur i ng t o hi mt hat I can use t o have an ef f ect on hi m. Never t hel ess, Fr i t chey st i l l does pl ay a r ol e i n my l i f e. He i s t he man of whoma cer t ai n t hought of mi ne i s a memor y. Thi s r ol e, t hi s r el at i on t o my t hought and t o me, gi ves me a way of t hi nki ng of hi m. I t i s a ver y l oyal char act er. I can car r y t he memor y wi t h me as I t r avel away f r om Fr i t chy and he f r om me; i t s bei ng a t hought of Fr i t chey depends on t he way I acqui r ed i t , but not on my pr esent ci r cumst ances. As an anal ogy, consi der t he not e I l ef t i n my copy of t he I nvest i gat i ons. I don' t r emember who wr ot e i t . The si gnat ur e i s har d t o make out . The wr i t i ng, never ver y l egi bl e, i s f aded. I n spi t e of al l t hi s, t he per son who pl ayed t he r ol e, " wr i t er of t hi s l et t er ' i s t he same as i t was t hi r t y year s ago. I t i s a ver y l oyal r ol e. Andt he sour ce of memor i es i s si mi l ar l y l oyal . I n such a case, what i s t he poi nt of cont i nui ng t o bel i eve anyt hi ng about Fr i t chey? I t doesn' t have much of a poi nt , unl ess t her e i s a good chance I wi l l encount er hi magai n and be abl e t o appl y t he bel i ef usef ul l y t hen. But of cour se t hi s i s somet hi ng we do a l ot . W e acqui r e a bel i ef about an obj ect at one t i me, when t he condi t i ons ar e f avor abl e f or doi ng so. W e appl y t he bel i ef at anot her t i me, under qui t e di f f er ent condi t i ons. These l at er condi t i ons mi ght make i t i mpossi bl e t o acqui r e t he i nf or mat i on i f we di d not al r eady have i t . Doi ng t hi s r equi r es t hat we be abl e t o r ecogni ze t he obj ect i n quest i on. Unl ess I r emember a l i t t l e bi t about Fr i t chey, so I can r ecogni ze hi mt he next t i me I see hi m, I won' t be i n a posi t i on t o use t he i nf or mat i on I have r et ai ned. The i nf or mat i on t hat we need, t o make det ach- and- r ecogni ze a vi abl e i nf or mat i on game, f al l s i nt o t wo cat egor i es. Ther e i s t he i nf or mat i on we wi l l use when we encount er t he obj ect agai n, t o deci de what t o do. The i nf or mat i on t hat Fr i t chey i s a pr ogr ammer f or Mi cr osof t i s what l eads me t o or d 6. 0. Cal l t hi s sor t of i nf or mat i on, " i nf or ask about t he sor r y st at e of W mat i on f or act i on. " But t her e i s al so " i nf or mat i on f or r ecogni t i on. " My t i dbi t of i nf or mat i on about Fr i t chey wi l l si t usel ess i n my br ai n, unl ess I r emember al so t hat he i s a bi g man wi t h a r ed bear d, et c. , i nf or mat i on t hat wi l l al l ow me t o r ecogni ze hi mnext t i me I meet hi m. Of cour se, t hese ar e di f f er ent
28
Ri p van Wi nkl e and Ot her Char act er s
Eur opean Revi ewof Phi l osophy
f unct i ons or uses f or i nf or mat i on, not di f f er ent cat egor i es of i nf or mat i on i t sel f . I n many cases, a bi t of i nf or mat i on wi l l ser ve bot h pur poses. To be usef ul , t hen, a det ached bel i ef , l i ke t he bel i ef t hat Fr i t chey i s a pr ogr ammer f or Mi cr osof t , wi l l have t o be par t of a l ar ger f i l e on Fr i t chey. The addi t i onal el ement s i n t he f i l e pr ovi de t he f act s about hi mt hat mi ght enabl e me t o r ecogni ze hi m. Never t hel ess, t her e i s no guar ant ee t hat bel i ef s wi l l be usef ul . Consi der my bel i ef t hat Fr i t chey was a Mi cr osof t pr ogr ammer . Under what char act er do I cont i nue t o bel i eve t hi s, once I det ach f r om Fr i t chey and l eave t he par t y? Si mpl y t he char act er : That man [ t he sour ce of t he bel i ef ] was a pr ogr ammer f or Mi cr osof t t hen [ t he t i me t he bel i ef was r equi r ed] . The bel i ef i s si mi l ar t o t he not e I di scover ed i n my ol d copy of Wi t t genst ei n' s I nvest i gat i ons. By i t sel f , wi t hout connect i on t o and augment at i on f r om ot her sour ces, i t i s usel ess. Thi s bel i ef , by i t sel f , i ncl udes no i nf or mat i on about Fr i t chey' s name, f or exampl e, so I coul dn' t even use i t t o cont act hi mi f I had a di r ect or y of Mi cr osof t empl oyees. A bel i ef t hat i s det ached and vi r t ual l y usel ess can never t hel ess be qui t e consci ous, even vi vi d. The f r ust r at i on t hat such bel i ef s can engender i s wel l known, and we have l i ngui st i c devi ces f or expr essi ng t hem. The f ol l owi ng conver sat i on not onl y makes per f ect l y good sense, i t i s al l t oo f ami l i ar f or some of us: " Do you know any pr ogr ammer s f or Mi cr osof t ?" " Yes . . . uh, uh, what ' s- hi s- name was a pr ogr ammer f or Mi cr osof t . I don' t r emember hi s name. " "W hen di d you meet hi m?" " I ' mnot sur e. " "W her e di d you meet hi m. " " I don' t r emember t hat ei t her . I can' t r emember much of anyt hi ng about hi m, except t hat he was a pr ogr ammer f or Mi cr osof t . " " Thanks. " A bel i ef l i ke t hi s one, t ot al l y usel ess at one t i me, may become usef ul l at er . Recover ed memor i es, or out si de sour ces, may di scl ose mor e i nf or mat i on about t he per son I r emember . Then t he appar ent l y usel ess bi t of i nf or mat i on may pr ove qui t e usef ul .
29
The det ach- and- r ecogni ze game pr ovi des par t of t he answer t o our quest i on about t he poi nt of bel i ef s wi t h l oyal char act er s. Such bel i ef s ser ve as component s of l ar ger , usef ul , bel i ef s, when t he obj ect s t he bel i ef s ar e about ar e r ecogni zed. They ar e compl et i ons f or pr agmat i cal l y at t ached bel i ef s, br oken of f f r omear l i er epi st emi cal l y at t ached bel i ef s. 6 Det ached bel i ef s do not si mpl y si t i n our mi nds and gat her dust whi l e we wai t t o encount er t he obj ect s t hey ar e about . They ar e par t s of var i ous act i vi t i es t hat do not di r ect l y af f ect t he obj ect s t hey ar e about , but may have pr of ound i ndi r ect ef f ect s on t hem. W e use such bel i ef s i n f i ve addi t i onal i nf or mat i on games. I n t he r ecol l ect i on game, I t r y t o squeeze mor e i nf or mat i on out of my memor i es. I may have f or mul at ed t he expl i ci t bel i ef t hat Fr i t chey was a pr ogr ammer f or Mi cr osof t as I t al ked t o hi m; he l eaves and I det ach t he bel i ef , t hi nki ng of hi mas " t hat man [ t hat I r emember ] . " But associ at ed wi t h t he bel i ef , at l east f or a whi l e, ar e memor y i mages f r omwhi ch I can cul l some mor e bel i ef s, t hat may ai d i n r ecogni zi ng hi m- e. g. , t hat he had a r ed bear d and wor e gr een suspender s over a pl ai d shi r t and had an unusual f asci nat i on wi t h Tupper war e. I do not need t o be at t ached t o Fr i t chey t o r ecol l ect mor e about hi m, and I al so do not need t o be at t ached t o hi mt o pl ay t he i nf er ence game- t o dr aw i nf er ences on t he basi s of my bel i ef s about hi m. I combi ne t hese bel i ef s wi t h ot her bel i ef s I have, t o f l esh out my concept i on of Fr i t chey. Gi ven hi s j ob, I i nf er t hat he i s br i ght and l i kes comput er s. Gi ven hi s dr ess, I assume t hat he i s happy i n Seat t l e and enj oys bei ng out door s. Gi ven t hat he wor ked f or Mi cr osof t when I t al ked t o hi m, I i nf er t hat he wi l l pr obabl y cont i nue t o do so f or some l engt h of t i me. I cal l a r at her speci al ki nd of i nf er ence updat i ng. Thi s i s an i nf er ence made not on t he basi s of obser ved or i nf er r ed movement s or changes i n t he t hi ng my bel i ef i s about , but on t he basi s of changes i n my own si t uat i on, or hen I meet Fr i t chey, I t hi nk gener al changes, l i ke t he passage of t i me. W 6.
What we have descr i bed so f ar i s per haps t he mai n var i at i on on t he det ach- andr ecogni ze i nf or mat i on game. Ther e i s a sour ce of my bel i ef , or of my f i l e of bel i ef s, about a cer t ai n per son. But we can al so have sour cel ess f i l es. Somet i mes we can f i gur e out qui t e a bi t about an obj ect t hat meet s a cer t ai n descr i pt i on, and wor k up a " f i l e" about i t , even t hough we have never encount er ed i t or communi cat ed wi t h anyone t hat has. I n t hi s case, once we have done t he r easoni ng, we ar e l ef t i n a si t uat i on anal ogous t o havi ng exami ned a sour ce.
30 ç
95/
:
;j
Eur opean
Revi ewof Phi l osophy
" Thi s man [ t he man I see] i s a pr ogr ammer at Mi cr osof t now [ t he t i me of t he t hought ] . " Lat er I t hi nk " That man [ t he man I r emember ] was a pr ogr ammer at Mi cr osof t t hen [ t he t i me I r emember [ . " The change f r om" now t i me of t he t hought ] " t o " t hen [ t he t i me of t he t hought [ " does not r ep[ [ t he r esent an i nf er ence about howFr i t chey has changed si nce I l ast saw hi m. I amj ust updat i ng; changi ng t he char act er i n a way t hat pr eser ves t he cont ent , gi ven t he new ci r cumst ance of bel i ef . Thi s updat e i s qui t e di f f er ent f r om anot her addi t i onal t hought I may have. Gi ven t he r el at i ve per manence of j obs at Mi cr osof t , I may f i gur e at the l at er t i me " That man [ t he man I r emember ] i s a pr ogr ammer at Mi cr osof t now [ t he moment of t hought ] . " Thi s i s not an updat e t hat pr eser ves t he previ ousl y bel i eved cont ent , but an i nf er ence- al t hough per haps one t hat i s not ver y r i sky, i f t he t i me i nt er val i s short. I t i s possi bl e t hat I have met Fr i t chey bef or e. Or per haps I have r ead an ar t i cl e about hi m- l et ' s suppose he has been i nvol ved i n t he Seat t l e Commons pr oj ect and was pi ct ur ed and br i ef l y descr i bed i n an ar t i cl e about i t . I al r eady have some concept of hi m. Upon r ef l ect i on, I may f i nd t hat t hi s concept f i t s t he man I t hought I had j ust met f or t he f i r st t i me t oo ni cel y t o be coi nci dent al , and concl ude t hat Fr i t chey i s t he man I r ead about. W hen I r ead about Fr i t chey i n t he paper , I was pl ayi ng t he communi cat i on i nf or mat i on game. Someone el se had met hi mand t al ked t o hi mand t aken hi s pi ct ur e. They had been i n an epi st emi c r el at i on t o hi m, and acqui r ed i nf or mat i on about hi m. The i nf or mat i on I get i s det ached. W hen I r ead about Fr i t chey, he pl ays a speci al sor t of epi st emi c r ol e in l i my f e, one medi at ed by symbol s. He was t he man I was r eadi ng about ; the man r ef er r ed t o by t he wor ds I sawi n t he paper . W hen I qui t r eadi ng the paper , I cont i nued t o have bel i ef s about hi m- det ached bel i ef s. For a whi l e, my bel i ef s about Fr i t chey may have been t i ed t o memor i es of t he ar t i cl e or t he paper . But of t en we r et ai n bel i ef s about a per son or t hi ngs we have encount er ed or r ead about , l ong af t er t he memor i es of t he per cept i ons, conver sat i ons or t ext f r omwhi ch we l ear ned about t hemhave f aded t o i nsi gni f i cance. The r ol e t hese obj ect s pl ay i n our l i ves i s j ust t hat of bei ng t he source of t he bel i ef s we have about them. Thebel i ef s I gai ned about Fr i t chey f r om r eadi ng t he ar t i cl e may be qui te det ached and f r agment ar y. Per haps I don' t r emember hi s name; I j ust have a concept of a manwhowas a pr ogr ammer at Mi cr osof t , wor ked on t he Commons, wor e gr een suspender s, and had a bi t of an odd name t hat r emi nded
Ri p van Wi nkl e and Ot her Char act er s
31
me of a Br i t i sh Rock St ar of an ear l i er er a. Such bel i ef s woul d seemqui t e usel ess. But t her e coul d he enough t her e t o make a pr et t y pl ausi bl e conj ect ur e, goi ng over t hi ngs af t er meet i ng Fr i t chey at t he par t y, t hat i t was t he same man. And t hen I woul d have a bel i ef t hat , next t i me I met Fr i t chey, coul d l ead t o a quest i on about hi s wor k on t he Commons, as wel l as t he one about W or d 6. 0. The f r agment ar y bel i ef s t ur n out t o be usef ul af t er al l . So now I am si t t i ng at home. I acqui r ed and det ached a bel i ef about Fr i t chey. I have r ecol l ect ed and i nf er r ed. And nowI i nt end t o ask hi m, next t i me I see hi m, about W or d 6. 0 and Seat t l e Commons. I amnot nowi n a t o do t posi t i on anyt hi ng o hi mdi r ect l y; he pl ays no pr agmat i c r ol e i n my l i f e at t he moment . But he wi l l , and I amf or mi ng pl ans about what t o do t hen- pl ayi ng t he pl anni ng i nf or mat i on game. Al l of t hese l ast f i ve games, r ecol l ect i on, i nf er ence, updat i ng, communi cat i on and pl anni ng, coul d he pl ayed whi l e I was st i l l at t ached, epi st emi cal l y or pr agmat i cal l y, t o Fr i t chey. But our mai n i nt er est i n t hemi s t hat t hey can be pl ayed wi t h det ached bel i ef s, bel i ef s wi t h l oyal char act er s, and hel p us see t he poi nt of havi ng such bel i ef s. Bel i ef s wi t h l oyal char act er s t hen have a causal r ol e t hat has t o do wi t h t he ki nds of i nf er ences and pl ans t hey l ead t o, and t hei r ef f ect s on act i on when combi ned wi t h ot her at t ached bel i ef s. 7 5. Mi si dent i f i cat i on
Suppose t hat a f ew days l at er I mi st akenl y t ake someone el se t o he El wood Fr i t chey- El wood' s br ot her Al phonse, say. I see Al phonse, as t al l and r ed- bear ded as El wood i s, one day at t he f eed l ot . " El wood, " I say, t al kor d 6. 0 r un i ng di r ect l y t o Al phonse, " W hy does t he Maci nt osh ver si on of W so sl owl y?" her e I have not onl y l ost t r ack of El wood, I have appl i ed my El wood bel i ef t o someone el se. Thi s i s no r eason t o say t hat I have l ost t he bel i ef . My bel i ef , t he one I acqui r ed at t he par t y, t hat El wood i s a Mi cr osof t pr ogr ammer , i s par t of t he expl anat i on f or my quest i on t o Al phonse- i t i s a mi sappl i cat i on of t hat ver y bel i ef . Si nce t he bel i ef expl ai ns t he mi st ake, t he mi st ake can har dl y he r eason f or supposi ng t he bel i ef t o have been l ost . I asked Al phonse t he quest i on, 7.
1 amher e concer ned wi t h what seems t o be a r at her basi c par t of our concept ual scheme, bel i ef s about i ndi vi dual t hi ngs and peopl e. Per haps as an expl i ci t di scl ai mer I shoul d not e t hat t her e ar e ot her i mpor t ant cl asses of bel i ef s t o whi ch such det ached bel i ef s ar e r el evant . For exampl e, a det ached bel i ef can ser ve as a conf i r mi ng i nst ance or a count er exampl e t o a gener al i zat i on.
32
Eur opean Revi ewof Phi l osophy
because I bel i eved t hat El wood was a pr ogr ammer at Mi cr osof t , and I t ook Al phonse t o be El wood. Suppose I say, by way of expl ai ni ng my quest i on t o t he puzzl ed l ooki ng Al phonse: " You ar e El wood Fr i t chey, t he Mi cr osof t Pr ogr ammer . " Thi s woul d be an i ndi r ect speech act , aski ng hi mi f i ndeed he was El wood. But consi der t he l i t er al st at ement I make. Cl ear l y i t i s f al se. The f act t hat I i nt end t o be speaki ng t o El wood and expr essi ng a bel i ef about El wood does not change t he f act t hat t he per son who i s pl ayi ng t he r el evant l i ngui st i c r ol e i n my l i f e i s Al phonse. Do I have, at t hi s poi nt , a bel i ef i n t he si ngul ar pr oposi t i on, t hat Al phonse Fr i t chey i s a Mi cr osof t Pr ogr ammer or t he pr oposi t i on t hat El wood i s Al phonse? I have sai d t hese t hi ngs, but do I bel i eve t hem? I t i s not necessar y t o say t hat I do. These bel i ef s ar e not r equi r ed t o expl ai n my quest i on; i t ' (
i s adequat el y expl ai ned by my t r ue bel i ef about El wood Fr i t chey and my f al se bel i ef t hat I amt al ki ng t o El wood Fr i t ch/ . Those bel i ef s expl ai n why I t hi nk I can say somet hi ng t o and about El wood Fr i t chj by usi ng a sent ence / t hat cont ai ns " you. " Ther e ar e r easons f or sayi ng t hat I don' t bel i eve t hese pr oposi t i ons. W hen t he mi st ake i s di scover ed, I woul d qui t e nat ur al l y say, " I t hought you wer e El wood" or " I t ook you t o be Hl wood. " I t seems t hat not ever y posi t i ve doxast i c at t i t ude i s comf or t abl y descr i bed as a bel i ef . My mi st ake, i n t aki ng Al phonse t o be El wood, may l ead t o f al se bel i ef s, i f i t r emai ns undi scover ed. But i f i t i s si mpl y a f r ansi t or y t hought , my t r ue bel i ef about El wood and my f al se bel i ef t hat t hi s man was El wood suf f i ce t o expl ai n my f al se st at ement . 6. Thi nki ng about Days Det ach- and- r ecogni ze i s a r easonabl e st r at egy f or deal i ng wi t h t he f act t hat i ndi vi dual s come and go. They become i ndexi cal l y i naccessi bl e, cease t o pl ay any epi st emi c or pr agmat i c r ol e i n our l i f e f or a whi l e, and t hen r e- ent er i t agai n. I n t he meant i me, by det achment , we can r et ai n our bel i ef s, updat e t hem, use t hem as t he basi s of i nf er ences, make pl ans about f ut ur e i nt er act i ons and communi cat e wi t h ot her s. Keepi ng t r ack of t he r el at i ve l ocat i ons t hat t i mes and i mmobi l e t hi ngs have t o us i s easi er t han keepi ng t r ack of peopl e and ot her mobi l e t hi ngs, at l east i n pr i nci pl e. I f I keep t r ack of my own movement s i n space, t hen I can cor r ect f or t hose movement s and st i l l t hi nk of al l t he pl aces I have been and bui l di ngs and l andmar ks I have seen i n t er ms of t hei r r el at i ve di st ance and
Ri p van Wi nkl e and Ot her Char act er s
33
di r ect i on f r omme. But I don' t usual l y do t hi s. Ri ght now, f or exampl e, I am wor ki ng at my home. I r emember wal ki ng at Pal o Al t o' s Bayl ands a f ew hour s ago. I don' t t hi nk of t he Bayl ands i n t er ms of t hei r r el at i ve di st ance and di r ect i on f r om me. I coul dn' t say r i ght of f t he t op of my head whet her t hey wer e t o my l ef t or r i ght , i n f r ont or behi nd me. I f I want t o t hi nk of t hem as " over t her e" I woul d st op and consi der wher e I l i ve i n r el at i on t o t hem, and what my or i ent at i on i n my house i s, and f i gur e i t out f r omt her e. I mi ght do t hi s f or exampl e i f I saw some f i r ewor ks, and wonder ed i f t hey coul d be comi ng f r omt he Bayl ands- i f t he Bayl ands wer e over t her e [ wher e I aml ooki ng] . W e do a much bet t er j ob keepi ng t r ack of whet her each of our exper i ences l i es i n t he past , pr esent or f ut ur e. Thi s i s f ai r l y easy because t her e i s an exact cor r espondence bet ween t he mode of t hi nki ng about t he exper i ence ( r emember i ng i t or pl anni ng t o have i t or havi ng i t ) and i t s posi t i on i n t i me r el at i ve t o t he pr esent moment . But we don' t usual l y keep ver y pr eci se t r ack of event s i n t er ms of t hese cogni t i ve r ol es. I f I have an i mpor t ant appoi nt ment comi ng up, I wi l l def i ni t el y be awar e t hat i t i s i n t he f ut ur e and not t he past ; I wi l l pr obabl y r emember t he t i me. As t he day goes on I may occasi onal l y f i gur e out how l ong i t i s bef or e t he event . But I don' t r et ai n my bel i ef by const ant l y updat i ng i n t er ms of " mi nut es f r omnow?' I r emember t hat t he appoi nt ment i s at 5; [ l ook at t he cl ock and see t hat i t i s 3: 30, and t hen I t hi nk, " An hour ai / a hal f f r omnowI see t he dean" or somet hi ng l i ke t hat . " W e can t hi nk of " now [ t he pr esent t i me] and " t oday [ t he pr esent day] " as ways of t hi nki ng about days t hat ar e bot h pr agmat i cal l y and epi st emi cal l y at t ached. On Sept ember 1st I can f i nd out what Sept ember 1st i s l i ke i n Pal o Al t o by l ooki ng. On Sept ember 1st I can make Sept ember 1st a day on whi ch I t ake a wal k by t aki ng a wal k. 8 On Sept ember 2nd, I can no l onger make any changes i n what happens on Sept ember 1st . 9 I amnot pr agmat i cal l y at t ached. I amno l onger i n a posi t i on t o act i vel y expl or e Sept ember 1st . I am not epi st emi cal l y at t ached. I can, however , st i l l expl or e my own memor i es, and may f or mnewbel i ef s as a r esul t of t hat expl or at i on.
8. 9.
See Evans ( 1981) . Set t i ng asi de " Cambr i dge changes. " I can t al k about Sept ember 1st on Sept ember 2nd, t hus maki ng i t , what i t hadn' t been bef or e, a day t al ked about by me on Sept ember 2nd. But t hi s doesn' t change what Sept ember 1st was l i ke; i t doesn' t change what happened on t hat day.
Ri p van Wi nkl e and Ot her Char act er s
34 Eur opean Revi ewof Phi l osophy Howabout t omor r ow? I amnot pr agmat i cal l y at t ached t o t omor r ow. I can' t do anyt hi ng t oday t hat wi l l di r ect l y change what happens t omor r ow. I can i ndi r ect l y af f ect t omor r ow by maki ng changes t oday i n mysel f and ot her s t hat wi l l have an ef f ect t omor r ow. I can pl an a t r i p t o a concer t , or cal l a meet i ng. And I am not now per cei vi ng t omor r ow' s event s; I am not epi st emi cal l y at t ached. I f I amnot now, and never have been, epi st emi cal l y or pr agmat i cal l y at t ached t o t omor r ow, howcan I t hi nk about t omor r ow, r eason about i t , make pl ans f or i t , and t he l i ke? Our syst em of dat es pr ovi des det ached ways of t hi nki ng about days t hat we have not yet encount er ed. Thi s syst em expl oi t s i mpor t ant met aphysi cal di f f er ences bet ween days on t he one hand and t hi ngs and peopl e on t he ot her . W e can syst emat i cal l y t al k about days we have not encount er ed and keep t r ack of t hemi n our t hought s and pl ans, si mpl y i n vi r t ue of t hei r posi t i on i n t he sequence of days, as r ef l ect ed by t he cal endar . Thi s i s i n par t because days ar e connect ed t o one anot her i n a l i near , pr edi ct abl e, f ashi on, dependent onl y on t he most gener al f act s t hat ci r cumscr i be exi st ence on ear t h. I t i s i n par t because of t he somewhat puzzl i ng f act t hat t he dat e of a day, i t s posi t i on i n t he cal endar , al t hough seemi ngl y a r el at i onal f act about i t , seems al so t o be i t s most essent i al f eat ur e. W e can hol d a day i n our modal e can i magi nat i on and change ever yt hi ng about what happens on t hat day. W suppose t hat al l sor t s of t hi ngs wi l l happen on Jul y 4, 1997; we can coher e can even ent l y i magi ne Jul y 4, 1767 havi ng been qui t e di f f er ent t han i t i s. W suppose t hat ever yt hi ng t hat wi l l happen on Jul y 4, 1997 coul d have happened i nst ead on Jul y 4, 1767, due t o t he absence of some sequence of event s a coupl e of mi l l i on year s ago t hat del ayed ever yt hi ng 230 year s. ( I n t hi s case, of cour se, Jul y 4, 1767 woul d be cal l ed " Jul y 4, 1997. " ) But we cannot consi der a gi ven day as bei ng i n a di f f er ent pl ace i n t he or der of days t han i t act ual l y i s. At l east i t i s not easy. Can next Jul y 4t h he a sour ce of my t hought s? No, f or i t l i es i n t he f ut ur e, and cannot be par t of t he cause of my bel i ef s. But I can have a sour ccl ess det ached bel i ef about next Jul y 4t h. The combi nat i on of our syst em of dat i ng, our hol i days and our t r adi t i ons al l ows me t o f i gur e out qui t e a hi t about next Jul y 4t h, even t hough I have not causal l y i nt er act ed wi t h i t . 10
10. 1 do not deny t hat we can and do have sour cel ess bel i ef s about per sons and
t hi ngs as wel l as t i mes and pl aces.
35
e Consi der what i s i nvol ved i n a si mpl e mat t er of ar r angi ng a meet i ng. W deci de t o meet on a cer t ai n day- say t wo weeks hence. The deci si on i s r ecor ded on a cal endar . At t hi s poi nt I amnot i n a pr agmat i c r el at i on t o t he day. I have made a pl an f or what wi l l happen on t he day, but I cannot make i t happen. I can' t at t end or not at t end t he meet i ng unt i l t he day i t sel f ar r i ves. The t ask of r ecogni zi ng a day i s di sanal ogous i n many ways t o r ecogni zhen we can i dent i f y t oday, yest er day, or t omor r owby i t s i ng an i ndi vi dual . W dat e, i t i s because we have been keepi ng t r ack, or ar e expl oi t i ng someone el se ( l i ke t he newspaper publ i sher ) , who does. Not i ce her e t hat " keepi ng t r ack" does not mean f ol l owi ng a par t i cul ar day. To keep t r ack of t i me i s usual l y not t o pi ck out a day and t r ack i t , but t o be awar e of whi ch day i t i s. That i s, t o be awar e of t he i mpor t ant pr oper t i es of t he day t hat pl ays t he " t oday" r ol e- what t he dat e i s, what event s ar e pl anned, et c. To l ose t r ack of what day i t i s, means not knowi ng t hat t he day t hat pl ays t hat r ol e has some ot her i mpor t ant at t r i but es, l i ke bei ng one' s anni ver sar y or t he day a phi l osophy paper was pr omi sed t o an edi t or . For t he most par t , apar t f r omhuman act i vi t i es and human i nst i t ut i ons, l i ke cal endar s and t he dat es on newspaper s, days don' t have par t i cul ar f eat ur es, easi l y ascer t ai nabl e by t he or di nar y ci t i zen, t hat set t hem of f f r om hen Ri p awakes i n t he hi l l s, t her e ot her days, i n r oughl y t he same season. W i s not hi ng about t he l ook of t hi ngs t hat suggest s t hat i t i s a 1786 day r at her t han a 1766 day. I t ' s onl y when he get s t o t own, and obser ves humans and human ar t i f act s, t hat t hi ngs begi n t o not f i t . So nowl et us r evi ew t he anal ogi es and di sanal ogi es we have f ound i n t er ms of our i nf or mat i on games. Ther e ar e cl ear anal ogues, wi t h days, t o t he st r ai ght - t hr ough and t r acki ng games. Ther e i s not a di r ect anal ogue t o t he det ach- and- r ecogni ze game, si nce we cannot r e- encount er a day once t i me has l ef t i t behi nd. However , t he pecul i ar met aphysi cal st at us of days, t hat i s r ef l ect ed i n our syst em of dat es, and per mi t s us t o ar r i ve at sour cel ess bel i ef s about t hem, pr ovi des an al t er nat i ve r out e t o / def at h¬d- Xhi nki ng. det ached t hi nki ng, pl anni ng, i nf or mat i on exchange, and t he l i ke. Wi t h t hi s al l i n mi nd, l et us, f i nal l y, r et ur n t o Ri p Van Wi nkl e. 7. Ri p van W i nkl e
What shoul d we say about Ri p VanWi nkl e? Van Wi nkl e acqui r ed a bel i ef t he day he f el l asl eep- Jul y 3, 1766, Ri p bel i ef t hat t hat day was a f i ne day. He hel d t hi s bel i ef under t he char say- a
36
Eur opean Revi ewof Phi l osophy
act er " Today [ t he day of t hi s t hought ] i s ni ce. " Then he sl ept f or t went y year s and t wo days, unt i l Jul y 5, 1786, and wal ked back t o t own. 1' W hat happened next ? The possi bi l i t y t hat st r uck Kapl an and Evans i s t hat Ri p mer el y updat ed hi s bel i ef . On Jul y 3r d he never f or ms any expl i ci t bel i ef ot her t han " Today hen he awakes on Jul y 5t h, t he [ t he day of t hi s t hought ] i s a ni ce day. " W bel i ef i s updat ed, due t o hi s awar eness of havi ng sl ept t hr ough a ni ght , and hi s l ack of awar eness of havi ng sl ept t went y addi t i onal year s, t o " Yest er day [ t he day bef or e t he day of t hi s t hought ] was a ni ce day. " He f al l s out of epi st emi c cont act wi t h t he cur r ent day when he f al l s asl eep, but has a r eady- _ , , made char act er i n mi nd f or when he wakes up. But t hen what i s t her e l ef t of t he or i gi nal bel i ef except t he f al se one about Jul y 4t h? But t he f al se bel i ef can not be t he t r ue bel i ef , so hasn' t Ri p l ost t he bel i ef i n quest i on? Thi s seems t o be t he ar gument t hat t hr eat ened Kapl an and appeal ed t o Evans. But even i n t he case of such t hi n updat i ng, t her e ar e backup char act er s f or Ri p t o hol d hi s bel i ef under . W hen Ri p bel i eves, t owar ds eveni ng, as i t dar k, of t hi s [ t he gr ows " Today day t hought ] was a ni ce day, " he has memor i es of seei ng t he f l ower s and f eel i ng t he sun, and so f or t h. So t he char act er , " That day [ t he day I r emember ] i s or was a ni ce day" i s avai l abl e t o sust ai n hi s bel i ef , when t he at t empt at updat i ng goes awr y. Even i f t hese memor i es f ade, t her e i s t he char act er , " That day [ t he day t hi s bel i ef was acqui r ed] i s or was a ni ce day. " So my vi ew i s t hi s. W hen he awakes on Jul y 5t h, Ri p updat es hi s bel i ef accor di ng t o hi s vi ew of howt he cont ext has changed. I l i s vi ew about t he change of cont ext i s mi st aken, and t he new char act er , " Yest er day [ t he day bef or e t he day of t hi s t hought ] was ni ce" i s not a way of bel i evi ng t he or i gi nal cont ent . But t hat i s no r eason t o say t hat Ri p has l ost hi s or i gi nal bel i ef . He r et ai ns i t under var i ous ot her backup char act er s. That ' s what I have t o say about Ri p; what about t he ot her concr et e char act er s t hat f i gur e i n our st or y, Fr ege, Kapl an and Evans?
Ri p van Wi nkl e and Ot her Char act er s
37
W e can' t pi n anyt hi ng much on Fr ege, f or we can' t hol d hi mr esponsi bl e f or t he st r at egy about bel i ef t hat hi s r emar k on sayi ng i nspi r ed. W e coul d consi der whet her he was r i ght about sayi ng, but we won' t do t hat i n t hi s essay. The st r at egy t hat Fr ege' s r emar k suggest ed t o Kapl an and Evans i s t hat r et ai ni ng bel i ef consi st s i n movi ng f r omf l i ght y char act er t o f l i ght y char act er i n ways t hat r ef l ect change i n cont ext . I t hi nk I have r ef ut ed t hat st r at egy. But my own st r at egy i s t he br oad i nt er pr et at i on of t he Fr ege- i nspi r ed one, gener al i zed and f r eed f r omi t s associ at i on wi t h hi s par t i cul ar exampl e. The det ach- and- r ecogni ze st r at egy f or handl i ng i nf or mat i on i t sel f embodi es a r egul ar t r ansi t i on, f r om st r ong char act er s t o l oyal char act er s and, when r ecogni t i on occur s, back t o st r ong char act er s. Thi s Fr ege- i nspi r ed doct r i ne, l i ke most , per haps, i s i nadequat e when const r ued nar r owl y, pl ausi bl e when gi ven a br oader i nt er pr et at i on. W e can' t f aul t Kapl an f or t hi nki ng t hat t her e was mor e t o be sai d about Van W i nkl e f or i n t hi s he was cor r ect . Evans was wr ong, I t hi nk, about Ri p and about Ri p Kapl an. But much of what I have sai d i s si mi l ar i n spi r i t t o i deas one f i nds i n t he body of hi s wor k. I wi l l end wi t h a coupl e of r emar ks on i nt er pr et i ng Evans. I n appr oachi ng Evans, i t seems t o me one must t r y t o separ at e hi s own i nf or mat i on- or i ent ed appr oach t o t hi ngs f r omt he devot i on t o a ver si on of Fr ege f i l t er ed t hr ough Davi dson t hat cr ops up nowand t hen, most especi al l y i n " Under st andi ng Demonst r at i ves' t he essay f r omwhi ch t he quot e above was t aken. Davi dson' s r el i ance on Tar ski and T- sent ences i n expl ai ni ng hi s vi ews on meani ng has i nspi r ed a t r adi t i on i n semant i cs t hat one mi ght cal l " homogeneous meani ng expl anat i on. " W e expl ai n t he meani ng of a sent ence by usi ng t he same sent ence, or one wi t h t he same meani ng. The wor k of Kapl an bel ongs squar el y i n t he t r adi t i on of " het er ogeneous meani ng expl anat i on. " One expl ai ns t he meani ng of a sent ence by showi ng, usi ng what ever l an-
11. Act ual l y, Kni cker bocker does not t el l us t he exact dat es, j ust t hat t he Decl ar at i on
guage one mi ght want , what sor t of t ool i t i s, howi t conveys var i ous t hi ngs i n var i ous ci r cumst ances. I magi ne t hat you ar e expl ai ni ng a Mer cat or pr oj ect i on map t o a chi l d.
asi de f r om t he possi bi l i t y of connect i ng humbl e anal yt i cal phi l osophy wi t h i mpor t ant l i t er at ur e, t he ext r a t went y year s doesn' t mat t er much.
One way of expl ai ni ng, per haps, i s t o use anot her Mer cat or pr oj ect i on map. A bet t er way i s t o use a gl obe. I t hi nk t he benef i t s of het er ogeneous over homogeneous expl anat i on ar e si mi l ar . As Davi dson hi msel f not es, t he homogeneous st r at egy wor ks wi t h i ndex-
of I ndependence and t he Amer i can Revol ut i on occur r ed whi l e Ri p sl ept . I ' ve added a day t o hi s t went y year sl eep f or exposi t or y conveni ence; t he i mpor t ant t hi ng i s t hat Ri p f el l asl eep on Jul y 3 and awoke on Jul y 5 t hi nki ng i t was Jul y 4;
38
Eur opean Revi ew of Phi l osophy
i nkl e and Ot her Char act er s Ri p van W
39
i cal s onl y when one supposes t hat t he semant i ci st ' s expl anat i on and t he st at ement expl ai ned ar e made by t he same per son at t he same t i me. , , One
1979. On t he l ogi c of demonst r at i ves. The Jour nal of Phi l osophi cal Logi c, 8: 81- 98. Repr i nt ed i n Mi dwest 1: 401- 412.
i nt er pr et at i on of " Under st andi ng Demonst r at i ves' i s t hat Evans t r i es, usi ng a var i et y of i deas t hat ar e i nt er est i ng i n t hei r own r i ght but ar e not necessar -
1989a. Demonst r at i ves. I n J. Al mog, et al . , eds. Themes Fr om Kapl an. New Yor k: Oxf or d Uni ver si t y Pr ess, pp. 481- 563.
of i ndexi cal s beyond t he speci al case i n whi ch i t wor ks. I t hi nk t hi s exper i ment f ai l s. And I al so t hi nk t hat t he i nf or mat i on- or i ent ed appr oach t hat one f i nds i n much of Evans wor k, and i n par t s of " Under st andi ng Demonst r at i ves' i s i l y wel l - sui t ed t o t he pur pose, t o ext end t he homogeneous t r eat ment
basi cal l y het er ogeneous i n i t s i mpl i cat i ons f or semant i cs.
Phi l osophy Depar t ment and CSLI St anf or d Uni ver si t y
Ref er ences Al mog, J. , J. Per r y, and f -I . Wet t st ei n, eds. 1989. Themes Fr om Kapl an. New Yor k: Oxf or d Uni ver si t y Pr ess. Bar wi se, J. and J. Per r y. 1983. Si t uat i ons and At t i t udes. Cambr i dge, MA: MI T- Br adf or d. Bur ks, A.
1949. I con, I ndex and Symbol . Phi l osophi cal and Phenomenol ogi cal Resear ch, vol . I X, no. 4, June 1949: 673- 689.
Cr i mmi ns, M . W 92. Tal k About Bel i ef s. Cambr i dge, MA: MI T- Br adf or d Books. Evans, G. 1973. The Causal Theor y of Names. Ar i st ot el i an Soci et y, Suppl ement ar y Vol ume 47, pp. 187- 208. 1981. Under st andi ng Demonst r at i ves. I n I -I . Par r et and J. Bouver esse, eds. Meani ng and Under st andi ng Ber l i n and NewYor k: Wal t er de Gr uyt er , pp. 280303. Repr i nt ed i n P. Your gr au, ed. Demonst r at i ves ( 1990) , pp. 71- 96. Fr ench, P. A. , T. E. Uehl i ng, Jr . , and H. K. Wet t st ci n, eds. 1979. Cont empor ar y I ' er specl i ves i n t he Phi l osophy of Language. . Mi nneapol i s: Uni ver si t y of Mi nnesot a Pr ess. I sr ael , D. , J. Per r y, and S. Tut i ya. 1993. Execut i ons, Mot i vat i ons and Accompl i shment s. The Phi l osophi cal Revi ew, Oct ober , 1993: 515- 540. Kapl an, U. 1978. t ) t hat . I n Mi dwest 1: 383- 400. Repr i nt ed i n P. Your gr au, ed. Demonst r at i ves ( 1990) , pp. 11- 33.
12.
See Per r y 1994, l ast sect i on.
19891) . Af t er t hought s. I n J. Al mog, et al , eds. Themes Fr om Kapl an. New Yor k: Oxf or d Uni ver si t y Pr ess, pp. 365- 614. Fer r y, J. I 979. The Pr obl emof t he Essent i al I ndexi cal , Nods 13, no. 1: 3- 21. Repr i nt ed i n J. Fer r y, The Pr obl emof t he Essent i al I ndexi cal and Ot her Essays ( 1993) , pp. 349. 1980. A Pr obl emabout Cont i nued Bel i ef . Paci f i c Phi l osophi cal Quar t er l y 61, no. 4: 317- 322. Repr i nt ed i n J. Per r y, The Pr obl emof t he Essent i al I ndexi cal and Ot her Essays ( 1993) , pp. 69- 90. 1993. The Pr obl em of t he Essent i al I ndexi cal and Ot her Essays NewYor k: Oxf or d Uni ver si t y Pr ess. 1994. Davi dson' s Sent ences and Wi t t genst ei n' s Bui l der s. APA Pr oceedi ngs. f or t hcomi ng. I ndexi cal s and Demonst r at i ves. I n B. Hal e and C. Wr i ght , eds. Compani on t o t he Phi l osophy of Language Oxf or d: Bl ackwel l . Rei chenhach, H. 1947.
50. Token- r ef l exi ve wor ds. I n El ement s of Symbol i c Logi c
NewYor k: The Fr ee Pr ess, pp. 284f f . Your gr au, P. , ed. 1990. Demonst r at i ves. Oxf or d: Oxf or d Uni ver si t y Pr ess.
-
FRANc0I S RECANATI
The Dynami cs of Si t uat i ons
1. Ret ai ni ng cont ent t hr ough cont ext change 1. 1 Cogni t i ve dynami cs i f someone want s t o say t he same t oday as he expr essed yest er day usi ng t he wor d ' t oday: he must r epl ace t hi s wor d wi t h ' yest er day Al t hough t he t hought i s t he same i t s ver bal expr essi on must be di f f er ent so t hat t he sense, whi ch woul d ot her wi se be af f ect ed by t he di f f er i ng t i mes of ut t er ance, i s r eadj ust ed.
- Fr ege, " The Thought " I n t hi s wel l - known passage Fr ege dr aws a consequence f r om t he cont ext - sensi t i vi t y of nat ur al l anguage sent ences. Si nce t he same sent ence, wi t h t he same l i ngui st i c meani ng, can expr ess di f f er ent pr oposi t i ons i n di f f er ent cont ext s, we may have t o use a di f f er ent sent ence wi t h a di f f er ent l i ngui st i c meani ng i f we want t o expr ess t he same pr oposi t i on i n a changed cont ext . As Per r y and Kapl an emphasi zed, t he l anguage of t hought behaves si mi l ar l y: i n or der t o t hi nk t he same t hi ng t oday whi ch I t hought yest er day usi ng t he ment al wor d ' t oday, ' I must r epl ace t hi s wor d wi t h ' yest er day. ' Thi s r ai ses t he pr obl em of cogni t i ve dynami cs. I n or der t o r et ai n an at t i t ude ( e. g. , bel i ef ) t owar d a cer t ai n cont ent one may have t o change t he sent ence t hr ough whi ch t hat cont ent i s appr ehended. I t f ol l ows t hat r et ai ni ng a bel i ef cannot be equat ed wi t h keepi ng t he same sent ence i n one' s bel i ef box. Yet r et ai ni ng t he bel i ef i s not mer el y a mat t er of accept i ng, at t ' , a sent ence s' expr essi ng t he same cont ent whi ch was expr essed by t he sent ence s accept ed at t ; f or one may f ai l t o r eal i ze t hat s and s' expr ess t he same cont ent . Kapl an gi ves t he f ol l owi ng exampl e:
41
42
The Dynami cs of Si t uat i ons
Eur opean Revi ew of Phi l osophy
has a t ense and a pr oper name. But i t i s l ess sensi t i ve t han t he sent ence I r ead on t he car d; i t has a const ant or near - const ant i nt er pr et at i on over a wi der r ange of change i n t he cont ext . I nt er pr et i ng down. Thi s i s t o f i nd a mor e sensi t i ve sent ence wi t h t he same i nt er pr et at i on. On a t r i p t o San Fr anci sco I r ead i n my Mobi l - Gui de, " San Fr anci sco has di l api dat ed cabl e car s. " I wr i t e on my not e- pad, " Thi s ci t y has di l api dat ed cabl e car s. " Or I t hi nk i t . But I do not get on t he cabl e car s I see. Lat er al i nt er pr et i ng. Thi s i s t o f i nd a sensi t i ve sent ence t o i nt er pr et a sensi t i ve sent ence. My f r i end shout s, " You ar e about t o be hi t by a r ock. " I t hi nk, " I am about t o be hi t by a r ock. " ( Per r y 1986a: 352)
I f i r st t hi nk " Hi s pant s ar e on f i r e. " I l at er r eal i ze " 1 amhe" and t hus come t o t hi nk " My pant s ar e on f i r e. " St i l l l at er , I deci de t hat I was wr ong i n t hi nki ng " I amhe" and concl ude " Hi s pant s wer e on f i r e. " I f , i n f act , I amhe, have I r et ai ned my bel i ef t hat m y pant s ar e on f i r e si mpl y because I bel i eve t he same cont ent , t hough under a di f f er ent char act er ? ( Kapl an 1989: 537n. )
The obvi ous answer t o t hi s r het or i cal quest i on i s ' No. ' To r et ai n a bel i ef i t i s not suf f i ci ent t o accept a sent ence s' expr essi ng t he same cont ent as t he pr evi ousl y accept ed sent ence. Ther e must be some speci al r el at i on bet ween t he newsent ence and t he ol d one- a speci al r el at i on whi ch i t i s t he busi ness of t he t heor y of cogni t i ve dynami cs t o i nvest i gat e. W hat I want t o do i n t hi s paper i s l ess t o f ur t her t he st udy of cogni t i ve dynami cs t han t o br oaden i t s scope. I want t o consi der a newdi mensi on of cont ext ual change, and t he speci f i c f or mof r et ent i on of cont ent t hat goes al ong wi t h i t .
1. 2 Hor i zont al and ver t i cal i nt er pr et i ng To ' r et ai n' t he cont ent of a speech or t hought epi sode even t hough t he cont ext changes, we must f i nd a sent ence expr essi ng t he same cont ent i n t he new cont ext . Thi s can he done i n t wo ways. I n Fr ege' s exampl e, ' t oday' i s r epl aced by ' yest er day; ' but i t coul d al so be r epl aced by an expl i ci t dat e ( e. g. , ' Wednesday' or ' t he t ent h of Mar ch' ) . Usi ng t he second met hod we make t he sent ence l ess cont ext - dependent i n or der t o f r ee i t f r om t he cont i ngenci es of t he or i gi nal cont ext . Usi ng t he f i r st met hod we keep t he degr ee of cont ext - sensi t i vi t y const ant but we compensat e f or t he change i n cont ext by a compl ement ar y l i ngui st i c change: we ' adj ust ' mdcxi cal s. The second met hod ( subst i t ut i ng names or dat es f or i ndexi cal s) cor r esponds t o Per r y' s not i on of ' i nt er pr et i ng up' whi l e t he f i r st met hod ( shi f t i ng i ndexi cal s) cor r esponds t o hi s ' l at er al i nt er pr et i ng: ' To i nt er pr et a sent ence hear d or r ead or ot her wi se appr ehended i s t o f i nd a sent ence wi t h t he same i nt er pr et at i on i n one' s own si t uat i on, as t he appr ehended sent ence had i n t he ut t er ance of or i gi n. Ne can di st i ngui sh sever al ki nds of i nt er pr et i ng. I nt er pr et i ng up. Thi s i s t o f i nd an i nt er pr et i ng sent ence wi t h a l ess sensi t i ve meani ng. My f r i end i n San Fr anci sco sends me a car d on whi ch he has wr i t t en, " Thi s ci t y has di l api dat ed cabl e car s. " I wr i t e i n t he dr af t of my t r avel - gui de: " San Fr anci sco has di l api dat ed cabl e car s. " Not e t hat t he sent ence I f i nd i s not i nsensi t i ve, I t
43
Let us si mpl i f y t he t er mi nol ogy and t al k of ' ver t i cal i nt er pr et i ng' when what i s af f ect ed i s t he degr ee of cont ext - sensi t i vi t y of t he accept ed sent ence. I nt er pr et i ng up and down ar e t wo i nst ances of ver t i cal i nt er pr et i ng. I n cont r ast , l at er al ( or hor i zont al ) i nt er pr et i ng i s t he sor t of t hi ng Fr ege t al ks about : we go f r om" You ar e about t o be hi t " t o " I amabout t o be hi t " or f r om" I t ' s a l ovel y day t oday" t o " I t was a l ovel y day yest er day. "
1. 3 Ver t i cal i nt er pr et i ng and cont ext ual change Hor i zont al i nt er pr et i ng essent i al l y i nvol ves a change of cont ext . I t i s because t he cont ext changes t hat we have t o adj ust t he expr essi on i n or der t o mai nt ai n t he cont ent f i xed. The l i ngui st i c change compensat es f or t hat cont ext ual change, and i s j ust i f i ed onl y i nsof ar as t he l at t er occur s. I wi l l say t hat a l i ngui st i c change i s of t he ' compensat or y' var i et y whenever t he f ol l owi ng condi t i ons ar e sat i sf i ed: i)
Had t he accept ed sent ence been l ef t unchanged, t he change of cont ext woul d have caused a change i n t he expr essed cont ent .
ii)
Had t he cont ext been l ef t unchanged, t he change i n t he accept ed sent ence woul d have caused a change i n t he expr essed cont ent .
I t i s easy t o check t hat t he t wo condi t i ons ar e sat i sf i ed i n Fr ege' s exampl e. The or i gi nal sent ence " Today i s F" was appr opr i at e t o t he cont ext i n whi ch i t was t okened ( vi z. Wednesday) ; but i n a di f f er ent cont ext ( vi z. t he next day) t he same sent ence woul d expr ess a di f f er ent cont ent . To mai nt ai n t he cont ent f i xed, t he sent ence has t o be changed. Condi t i on ( i ) , t her ef or e, i s sat i sf i ed. The second condi t i on al so i s sat i sf i ed: t he changed sent ence " Yest er day was F" i s appr opr i at e t o t he newcont ext ( one day l at er ) ; but i f i t had been t okened i n t he or i gi nal cont ext ( Wednesday) t he expr essed cont ent woul d have been di f f er ent .
44
Eur opean Revi ewof Phi l osophy
W i t h ver t i cal i nt er pr et i ng, t he si t uat i on i s r at her di f f er ent , or so i t seems. Tr adi t i onal l y, ver t i cal i nt er pr et i ng i s not consi der ed as i nvol vi ng a change of cont ext at al l , but mer el y a change of wor di ng. W hat f ol l ows i s a pr i ma f aci e ar gument i n behal f of t he t r adi t i onal pi ct ur e. I n a gi ven cont ext , a gi ven cont ent can be expr essed i n t wo manner s: ei t her i n a cont ext - sensi t i ve manner or i n a cont ext - i nsensi t i ve manner . Ver t i cal i nt er pr et i ng i s mer el y t he t r ansi t i on f r omone manner of expr essi on t o t he ot her . Thus I can t hi nk " Today i s F, " r eal i ze t hat t oday i s Wednesday, and i nf er " Wednesday i s F. " The cont ext r emai ns what i t was: t he day of ut t er ance i s unchanged, i n cont r ast t o what happens i n Fr ege' s exampl e ( t he shi f t f r om ' t oday' t o ' yest er day' ) . Thi s shows t hat ver t i cal i nt er pr et i ng can t ake pl ace i n t he absence of cont ext ual change- i n such a way t hat t he f i r st condi t i on above does not appl y. Ver t i cal i nt er pr et i ng can al so t ake pl ace as a r esul t of cont ext change, as I poi nt ed out ear l i er : t o expr ess on Thur sday t he same cont ent t hat was expr essed by " Today i s F" on Wednesday, I may opt f or t he ver t i cal met hod and r ef er t o t he pr evi ous day as ' Wednesday Even i n t hat case, however , t he l i ngui st i c change i s not compensat or y; f or t he second condi t i on st at ed above i s not sat i sf i ed. Had t he cont ext been l ef t unchanged, t he change i n t he accept ed sent ence woul d not have caused a change i n t he expr essed cont ent . For ' i nt er pr et i ng up' i s a l i ngui st i c change whi ch mai nt ai ns cont ent f i xed whet her or not a change of cont ext occur s. Thi s ar gument suppor t s t he cl ai m t hat , i n cont r ast t o hor i zont al i nt er pr et i ng, ver t i cal i nt er pr et i ng i s de- coupl ed f r om cont ext ual change. Yet I bel i eve t hat i t i s not : ver t i cal i nt er pr et i ng, i n t he f r amewor k I wi l l set up, does i nvol ve a change of cont ext . ' Thi s concl usi on i s wel come, f or i t f aci l i t at es t heor et i cal uni f i cat i on. Ver t i cal i nt er pr et i ng has many char act er i st i cs i n common wi t h hor i zont al i nt er pr et i ng, and i t poses much t he same pr obl ems. I n par t i cul ar i t r ai ses t he pr obl em of cogni t i ve dynami cs. Thus we have no t r oubl e const r uct i ng a count er par t t o Kapl an' s exampl e i nvol vi ng ver t i cal r at her t han hor i zont al i nt er pr et i ng: I f i r st t hi nk " Today i s F. " I l at er r eal i ze " Today i s Wednesday" and t hus coni c t o t hi nk " Wednesday i s F. " St i l l l at er , I deci de t hat I was wr ong i n t hi nki ng " Today i s Wednesday" and concl ude " Today i s F. " I f , i n f act , t oday i s Wednesday, have I r et ai ned my bel i ef t hat Wednesday i s F si mpl y because I bel i eve t he same cont ent , t hough under a di f f er ent char act er ? 1. 4 Basi c ver t i cal i nt er pr et i ng
Ther e ar e t wo f or ms of ver t i cal i nt er pr et i ng, i l l ust r at ed by t he f ol l owi ng pi eces of r easoni ng: 1.
Evi dence f or t hat cl ai m wi l l be of f er ed onl y wi t h r espect t o basi c ver t i cal i nt er pr et i ng- t he speci f i c t opi c of t hi s paper .
The Dynami cs of Si t uat i ons
1)
Thi s ci t y has di l api dat ed cabl e car s Thi s ci t y = San Fr anci sco Er go: San Fr anci sco has di l api dat ed cabl e car s
2)
I t ' s r ai ni ng Thi s i s Par i s
45
Er go: i t ' s r ai ni ng i n Par i s ( 1) i s Per r y' s exampl e of ' i nt er pr et i ng up. ' ( 2) i s a di f f er ent t ype of exampl e, whi ch I wi l l f ocus on i n t hi s paper . I t i l l ust r at es what I cal l ' basi c ver t i cal i nt er pr et i ng. ' I n bot h cases, we st ar t wi t h a cont ext - sensi t i ve sent ence: " I t ' s r ai ni ng, " or " Thi s ci t y has di l api dat ed cabl e car s. " The sent ence i s cont ext - sensi t i ve because a const i t uent of t he pr oposi t i on whi ch i t expr esses i s det er mi ned by t he cont ext . " Thi s ci t y has di l api dat ed cabl e car s: ' ut t er ed i n San Fr anci sco ( i n t he ci r cumst ances i magi ned by Per r y) , i s t r ue i f and onl y i f San Fr anci sco has di l api dat ed cabl e car s. San Fr anci sco i s a const i t uent of cont ent det er mi ned by t he cont ext . I n t he same way, when, at pl ace p, I l ook out t he wi ndowand say " I t ' s r ai ni ng, " what I say i s t r ue i f and onl y i f i t ' s r ai ni ng at p. I n bot h ( 1) and ( 2) , t he cont ext ual component of cont ent comes t o be obj ect i vel y desi gnat ed i n t he concl usi on. The di f f er ence bet ween ( 1) and ( 2) l i es i n t he manner i n whi ch t he cont ent of t he or i gi nal sent ence depends on t he cont ext . I n ( 1) , bef or e get t i ng t o be obj ect i vel y desi gnat ed ( i n t he concl usi on) , t he cont ext ual component of cont ent i s al r eady ' ar t i cul at ed: by means of t he i ndexi cal phr ase ' t hi s ci t y. ' That phr ase st ands f or San Fr anci sco. I t does so by vi r t ue of t wo f act or s: ( i ) a l i ngui st i c f act or - t he meani ng of t he phr ase ' t hi s ci t y; ' and ( i i ) a cont ext ual f act or - t he ci t y whi ch t he speaker mani f est s t he i nt ent i on t o demonst r at e. But i n ( 2) , t he cont ext - dependent const i t uent of cont ent i s not ar t i cul at ed at al l . I t i s det er mi ned by t he cont ext ual f act or al one. The not i on of an unar t i cul at ed const i t uent i s pr esent ed i n t hi s passage f r omPer r y: I t i s a r ai ny Sat ur day mor ni ng i n Pal o Al t o. I have pl ans f or t enni s. But my younger son l ooks out t he wi ndowand says, " I t i s r ai ni ng. " I go back t o sl eep. W hat my son sai d was t r ue, because i t was r ai ni ng i n Pal o Al t o. Ther e wer e al l sor t s of pl aces wher e i t wasn' t r ai ni ng: i t doesn' t j ust r ai n or not , i t r ai ns i n some pl aces whi l e not r ai ni ng i n ot her s. I n or der t o assi gn a t r ut h- val ue t o my son' s st at ement , as I j ust di d, I needed a pl ace. But no component of hi s st at ement st ood f or a pl ace. The ver b ' r ai ni ng' suppl i ed t he r el at i on r ai ns (t , p) - a dyadi c r el at i on bet ween t i mes and pl aces, as we have j ust not ed. The t ensed auxi l i ar y' i s'
46
suppl i es a t i me, t he t i me at whi ch t he st at ement was made. ' I t ' doesn' t suppl y anyt hi ng, but i s j ust synt act i c f i l l er . So Pal o Al t o i s a const i t uent of t he cont ent of m y son' s r emar k, whi ch no component of hi s st at ement desi gnat es; i t i s an unar t i cul at ed const i t uent . ( Per r y 1986b: 2116)
Per r y cl ai ms, cor r ect l y, t hat t her e i s a di f f er ence bet ween " I t ' s r ai ni ng" and " I t ' s r ai ni ng her e?' The cont ent i s t he same: bot h ut t er ances ar e t r ue i f and onl y i f i t ' s r ai ni ng i n Pal o Al t o. I n bot h cases, Pal o Al t o i s a cont ext ual const i t uent of cont ent . But t hat const i t uent i s det er mi ned pur el y cont ext ual l y i n
one case, par t l y cont ext ual l y i n t he ot her case. 2 I n t he same paper ( p. 2 18) , Per r y makes a f ur t her cl ai m: t hat " I t ' s r ai ni ng her e" i s i nt er medi at e bet ween " I t ' s r ai ni ng" and " I t ' s r ai ni ng i n Par i s?' To go f r om" I t ' s r ai ni ng" t o " I t ' s r ai ni ng i n Par i s: ' we must f i r st conver t " I t ' s r ai ni ng" i nt o t he i ndexi cal t hought " I t ' s r ai ni ng her e?' Ent er t ai ni ng t he i ndexi eal t hought " I t ' s r ai ni ng her e" ( i n cont r ast t o t he pur e cont ext ual t hought " I t ' s r ai ni ng" ) al r eady i nvol ves r ef l ect i ng on t he r el evant aspect of t he cont ext ( t he pl ace wher e we ar e) and maki ng i t expl i ci t i n t he r epr esent at i on i t sel f . I f Per r y i s r i ght t he t r ansi t i on r epr esent ed i n ( 2) compr i ses f our st eps r at her t han t hr ee: a) b) c)
I t ' s r ai ni ng I t ' s r ai ni ng her e her e = Par i s
d)
Er go: I t ' s r ai ni ng i n Par i s
I n ( a) t he pl ace i s a pur el y cont ext ual const i t uent of cont ent . I n ( b) t hat const i t uent i s r ef l ect ed on and t hought about i n an egocent r i c manner . Then t he ' br i dgi ng t hought ' ( c) occur s, and t he egocent r i c f i l e i s l i nked t o an encycl opedi c one. The concl usi on ( d) i s an encycl opedi c t hought . 3 The st ep f r om ( h) t o ( d) i s an i nst ance of t he t r ansi t i on r epr esent ed i n ( 1) . The t r ansi t i on r epr esent ed i n ( 2) t her ef or e i s mor e compl ex t han and
2.
3.
The Dynami cs of Si t uat i ons
Eur opean Revi ew of Phi l osophy
Ar guabl y, t her e i s a cont i nuum bet ween pur e i ndexi eal i t y and unar t i cul at ed const i t uency. Demonst r at i ves f al l i n bet ween. The meani ng of a demonst r at i ve can be ver y poor ( as i n ' he' or ' t hat man' ) ; at l east i t does not det er mi ne t he r ef er ent , whi ch depends upon t he i nt ent i ons mani f est ed by t he speaker . That i s qui t e di f f er ent f r om' I' , wher e t her e i s a r ul e whose appl i cat i on does not have t o be l i censed by t he speaker ' s i nt ent i ons. That i s al so di f f er ent f r om unar t i cul at ed const i t uency cases, wher e communi cat i ve i nt ent i ons do al l t he j ob. 1 amusi ng t he t er mi nol ogy of Recanat i ( 1993) .
47
cont ai ns t he t r ansi t i on i l l ust r at ed by ( 1) . But i t cont ai ns somet hi ng el se as wel l : a mor e basi c t r ansi t i on f r omunar t i cul at edness t o ar t i cul at edness ( i . e. , t he t r ansi t i on f r om( a) t o ( h) ) . The cl ai mt hat i ndexi cal i t y medi at es bet ween unar t i cul at ed const i t uency and obj ect i ve desi gnat i on i s not speci f i cal l y Per r y' s; i t has been made by var het her or not i ous aut hor s such as e. g. , Sechehaye ( 1926) or Lyons ( 1975) . W t hey ar e r i ght , I t hi nk we cannot but acknowl edge t hat t he t r ansi t i on i n ( 2) i nvol ves a basi c st ep whi ch ( 1) does not i nvol ve: t he st ep f r om unar t i cul at edness t o ar t i cul at edness. Tr ansi t i ons i nvol vi ng t hat basi c st ep ( or i t s conver se: t he st ep f r om ar t i cul at edness t o unar t i eul at edness) ar e i nst ances of what I cal l ' basi c ver t i cal i nt er pr et i ng' - t he pr ocess whi ch I amgoi ng t o st udy i n t hi s paper . 2. Aust i ni an semant i cs 2. 1 Aust i ni an pr oposi t i ons
The t heor et i cal f r amewor k I wi l l assume i s t hat of ' Aust i ni an semant i cs' ( Bar wi se and Et chemendy 1987; Recanat i 1996) . I n Aust i ni an semant i cs, a sent ent i al r epr esent at i on r epr esent s a st at e of af f ai r s, but i t s cont ent cannot be r educed t o t he st at e of af f ai r s i t r epr esent s. To use a r epr esent at i on, i n speech or i n t hought , i s t o ' appl y' i t t o some si t uat i on. The compl et e cont ent of a speech epi sode or a t hought epi sode i n whi ch a r epr esent at i on r i s used t her ef or e i nvol ves not onl y t he st at e of af f ai r s a whi ch r r epr esent s, but al so t he si t uat i on s whi ch t he r epr esent at i on i s supposed t o f i t . The compl et e cont ent of t he r epr esent at i onal act i s an ' Aust i ni an pr oposi t i on' , consi st i ng of a si t uat i on and a f act pr esent ed as hol di ng i n t hat si t uat i on. An Aust i ni an pr oposi t i on i s a pr oposi t i on of t he f or m: 5 Ht a
That t he compl et e cont ent of an ut t er ance or t hought i s an Aust i ni an pr oposi t i on i s i l l ust r at ed by t he f ol l owi ng exampl e, due t o Bar wi se and Et chemendy ( Bar wi se and Et chemendy 1987: 121- 2) . Looki ng at a poker game, I say " Cl ai r e has a good hand?' I descr i be t he si t uat i on I amwi t nessi ng as a si t uat i on i n whi ch Cl ai r e has a good hand. I f I ammi st aken and Cl ai r e i s not a const i t uent of t he si t uat i on ( i f she i s not among t he pl ayer s of t he game I amwat chi ng, cont r ar y t o what I bel i eve) , my ut t er ance i s not t r ueeven i f Cl ai r e i s pl ayi ng poker i n some ot her par t of t he ci t y and has a good
48
Eur opean Revi ewof Phi l osophy
The Dynami cs of Si t uat i ons
hand t her e; t he ut t er ance i s not t r ue because t he r el evant si t uat i on i s not as descr i bed. I n ot her wor ds, r at her t han bei ng t r ue i f f Cl ai r e has a good hand, t he ut t er ance i s t r ue i f f t he si t uat i on concer ned by t he ut t er ance suppor t s . The compl et e cont ent of t he ut t er ance t her ef or e i s: t hat poker game F= 2. 2 Si t uat i ons and f act s I n si t uat i on t heor y ( e. g. , Bar wi se 1989) , si t uat i ons ar e model l ed as set s of at omi c f act s. An at omi c f act ( or st at e of af f ai r s- I wi l l use t he t wo phr ases i ndi f f er ent l y) i s a t r i pl e consi st i ng of an i i r el at i on, a of n pl ace sequence appr opr i at e ar gument s, and a pol ar i t y i whi ch can he 0 or 1. Thus t he f act t hat Cl ai r e has a good hand can be r epr esent ed as: At omi c f act s ( whet her posi t i ve or negat i ve) ar e t he onl y f act s whi ch si t uat i ons cont ai n. W hat about non- at omi c f act s, e. g. , di sj unct i ve f act s ( John l oves Mar y or Pet er hat es Paul ) , gener al f act s ( Ther e i s a man who l oves Mar y) , and so f or t h? A si t uat i on s cannot cont ai n such f act s, but i t can suppor t t hem, by vi r t ue of t he at omi c f act s i t cont ai ns. The ' suppor t ' r el at i on can be def i ned al ong t he f ol l owi ng l i nes: A si t uat i ons suppor t s anat omi c f act o ( i n symbol s: s
o) j ust i n case a
49
2. 3 The cont ext ual nat ur e of si t uat i ons Acr uci al f eat ur e of t he t heor y I wi l l devel op wi t hi n t he Aust i ni an f r amewor k i s t he cont ext ual nat ur e of t he si t uat i onal component . The si t uat i on whi ch a r epr esent at i on concer ns i s a hi ghl y cont ext - dependent aspect of i t s cont ent ; l i ke t he r ef er ence of a demonst r at i ve, i t depends upon t he i nt ent i ons mani f est ed by t he speaker . The si t uat i onal component i s even mor e heavi l y dependent on cont ext t han t he r ef er ence of a demonst r at i ve. I n t he t heor y I wi l l devel op, t he si t uat i onal component i s unar t i cul at ed: i t i s a pur el y cont ext ual component of cont ent . Not e how st r ong t he cl ai mI ammaki ng i s. I f t he cont ent of ever y ut t er ance cont ai ns a si t uat i onal component whi ch i s pur el y cont ext ual , t hen cont ext - dependence i s gener al i zed i n a ver y r adi cal manner . The t heor y t o be devel oped i s i ndeed r adi cal l y ' cont ext ual i st . ' 4 But I wi l l not el abor at e t hat poi nt i n t hi s paper . The cl ai mt hat t he si t uat i on t al ked about i s ( al ways) a pur el y cont ext ual aspect of cont ent can he obj ect ed t o, on t he gr ounds t hat t he si t uat i on can he made expl i ci t i n t he sent ence i t sel f . Consi der t he f ol l owi ng pai r : I t i s r ai ni ng I n Par i s, i t i s r ai ni ng Assume t hat t he f i r st sent ence i s ut t er ed i n Par i s and concer ns Par i s. The obj ect i on goes as f ol l ows: Bot h r epr esent at i ons ar e about Par i s. I n t he second case, however , t he si t uat i on t he r epr esent at i on i s about ( Par i s) i s l i n-
A si t uat i on s suppor t s a di sj unct i ve f act a v a' j ust i n case s o or s o' . A si t uat i on s suppor t s an exi st ent i al f act l x a ( x) i f f , f or some obj ect a, s suppor t s o( a) .
gui st i cal l y encoded; i t i s not a pur el y pr agmat i c, cont ext ual component . Hence t he si t uat i onal component i s not al ways cont ext ual . But I wi l l deny t hat t he t wo r epr esent at i ons concer n one and t he same si t uat i on. As we wi l l see, as soon as t he si t uat i on ( Par i s) i s made expl i ci t i n t he r epr esent at i on i t sel f , anot her si t uat i on r epl aces i t i n t he cont ent of t he
The set of al l si t uat i ons whi ch suppor t a f act a ( whet her a i s at omi c or not ) def i nes t he si t uat i on t ype T( . To say t hat a si t uat i on i s of t he t ype T0 i s t o say t hat i t bel ongs t o { s s a} . I nsof ar as i t st at es a f act o, a sent ent i al r epr esent at i on ( be i t l i ngui st i c or ment al ) denot es a cer t ai n si t uat i on t ype, namel y
new t hought t hus gener at ed. I t f ol l ows t hat basi c ver t i cal i nt er pr et i ng ent ai l s a change of cont ext . Thi s i s r emi ni scent of Gar di ner ' s t r eat ment of i l l ocut i onar y f or ce ( whi ch he cal l s ' sent ence- qual i t y' ) . Sent ence- qual i t y i s el usi ve, Gar di ner says. As soon as i t i s made expl i ci t by means of such phr ases as ' I assur e you " wi t h t hemcomes a new i mpor t at i on of sent ence- qual i t y, t he nat ur e of whi ch i s not decl ar ed. Thus t he at t empt t o asser t t he qual i t y of a
Es.
T0. Accor di ng t o Aust i ni an semant i cs, t he r epr esent at i on i s t r ue i f and onl y i f t he si t uat i on whi ch i t concer ns bel ongs t o t he si t uat i on t ype whi ch i t denot es ( Aust i n 1950) .
4.
' Cont ext ual i sm' i s my name f or t he sor t of posi t i on ar gued f or by Tr avi s ( passi m) and Sear l e ( 1978, 1980) .
50
Eur opean Revi ewof Phi l osophy
The Dynami cs of Si t uat i ons
sent ence wi t hi n t hat sent ence i t sel f does but i nvol ve us i n an i nf i ni t e r egr ess" ( Gar di ner 1932: 199l ) . Aust i n r ej ect ed t hat l i ne of ar gument and cl ai med t hat i l l ocut i onar y f or ce can be made expl i ci t r ef l exi vel y i n t he ut t er ance i t sel f ( Aust i n 1962) . Be t hat as i t may, t he si t uat i onal component of cont ent ar guabl y has t he same el usi ve pr oper t y whi ch Gar di ner ascr i bes t o i l l ocut i onar y f or ce: i t i s not possi bl e t o encode t he si t uat i onal component wi t hout shi f t i ng it. 3. 5- st r uct ur es 3. 1 I nt r oduci ng t he not i on 6- st r uct ur es ar e sent ence- l i ke r epr esent at i ons whi ch cont ai n ot her sent ence- l i ke r epr esent at i ons. To si mpl i f y mat t er s, I wi l l t al k of ' sent ences' i nst ead of ' sent ence- l i ke r epr esent at i ons. ' 8- st r uct ur es t her ef or e ar e compl ex sent ences consi st i ng of a sent ence ( t he ' r adi cal ' ) and an expr essi on ( t he ' 5- par t ' ) whi ch makes a sent ence out of a sent ence. Exampl es of such compl ex sent ences i ncl ude:
I n John' s mi nd / ci gar et t es ar e cheap W hen 1 was young, i t was t he case t hat / ci gar et t es ar e cheap wi l l consi der i n t hi s r adi cal hol ds. 6 A senanot her . W i t h r espect
cheap; but t hat i s not t r ue i n ot her si t uat i ons ( e. g. , i n Fr ance) . I n t he f i r st exampl e above, i t i s t he j ob of t he expr essi on ' I n Spai n' t o i ndi cat e t he speci f i c si t uat i on i n whi ch t he r adi cal ' ci gar et t es ar e cheap' i s sai d t o hol d. The 6- par t s ' i n John' s mi nd' and ' when I was young' pl ay t he same r ol e i n t he ot her exampl es, except t hat t he si t uat i ons r ef er r ed t o ar e r at her di f f er ent .
5. 6.
3. 2 Fact ual and count er f act ual 6- st r uct ur es I n f act ual 6- st r uct ur es, t he 6- par t i ndi cat es t hat t he r adi cal hol ds i n a ( possi bl y maxi mal ) por t i on of t he act ual wor l d, or i n a set of such por t i ons ( si t uat i on t ype) . The por t i ons of @r ef er r ed t o i n f act ual 8- st r uct ur es can be var i ousl y ci r cumscr i bed. I n some cases- as i n t he t hi r d exampl e above- t he si t uat i on i s t empor al l y ci r cumscr i bed. Tempor al 5- st r uct ur es i ncl ude: ( At t / when John get s back) i t wi l l he t he case t hat p ( At t / when John got back) i t was t he case t hat p I t somet i mes happens t hat p The si t uat i on t al ked about can al so be spat i al l y ci r cumscr i bed: Two mi l es f r omher e, i t i s t he case t hat p W her e John l i ves, i t i s t he case t hat p I n Chi cago, p Ot her por t i ons of @ ar e ci r cumscr i bed ar ound i ndi vi dual s. ( An i ndi vi dual l i ke John, or an event l i ke W or l d W ar I I , can bot h be consi der ed as ' si t uat i ons' i n t he t echni cal sense, i . e. , as a set s of at omi c f act s. ) Ther e ar e, I
I n Spai n / ci gar et t es ar e cheap
I n t ypi cal 6- st r uct ur es ( t he onl y 5- st r uct ur es I paper ) t he 6- par t i ndi cat es t he si t uat i on i n whi ch t he t ence may be t r ue of ( or at ) a si t uat i on, but nor of / at t o t he si t uat i on i n Spai n, i t i s t r ue t hat ci gar et t es ar e
51
On Gar di ner , Aust i n and t he expl i ci t mar ki ng of i l l ocut i onar y f or ce, see Recanat i ( 1987: 31f f ) . The negat i on oper at or ' i t i s not t he case t hat ' i s an except i on ( hence negat i ve sent ences ar e ' at ypi cal ' h- st r uct ur es) : i t makes a sent ence out of a sent ence, but i t does not i ndi cat e t he si t uat i on i n whi ch t he r adi cal hol ds. For mor e about t ypi cal and at ypi cal 6- st r uct ur es, see Recanat i f or t hcomi ng, par t 3, wher e an expanded ver si on of t hi s mat er i al i s pr esent ed.
bel i eve, st i l l ot her met hods of i ndi vi duat i ng por t i ons of @ , but I cannot go i nt o t hi s mat t er and I wi l l be cont ent wi t h t he exampl es I have gi ven. I t ur n t o count er f act ual 5- st r uct ur es. The def i ni ng char act er i st i c of
6- st r uct ur es i s t hat t hey t al k about count er f act ual si t uat i ons or si t uat i on t ypes. Her e I am not usi ng ' count er f act ual si t uat i on' i n t he sense of a si t uat i on whi ch does not obt ai n i n t he act ual wor l d. By count er f act ual si t uat i on I mean a ( possi bl y maxi mal )
counl cr f act ual
. Fact ual si t uat i ons ar e por por t i on of an al t er nat i ve t o t he act ual wor l d @ t i ons of @ ; count er f act ual si t uat i ons ar e por t i ons of al t er nat i ves t o 0i magi nar y wor l ds, et c. A count er f act ual si t uat i on t ype i s a set of si t uat i ons whi ch cont ai ns some count er f act ual si t uat i on. Two t ypes of count er f act ual 6- st r uct ur es ar e wor t h di st i ngui shi ng:
Met a- r epr esent at i onal b- st r uct ur es pr esent t he r adi cal as hol di ng i n t he si t uat i on descr i bed by some r epr esent at i on whi ch t he 6- par t r ef er s t o. Ment al st at es, pi ct ur es, st or i es, ut t er ances, et c. al l count as ' r epr esent at i ons' i n t he r el evant sense. Exampl es of met a- r epr esent at i onal 5st r uct ur es i ncl ude:
52
The Dynami cs of Si t uat i ons
Eur opean Revi ewof Phi l osophy I n John' s mi nd, p I n t he pi ct ur e, p John bel i eves t hat p John sai d t hat p Accor di ng t o t he Anci ent s, p At t he end of Gone wi t h t he wi nd, p I n hypot het i cal 6- st r uct ur es, t he si t uat i on ( or t ype of si t uat i on) r ef er r ed t o i s count er f act ual , but i t i s not pr esent ed as t he si t uat i on
r esul t i ng 6- st r uct ur e, t hus gener at i ng an i t er at ed 6- st r uct ur e ' t he l andl or d t hi nks t hat ! i n 1996 i t wi l l be t he case t hat ! Pet er i s penni l ess. ' A t hi r d 6- par t , vi z. t he ' past ' oper at or , appl i es t o t hat i t er at ed 6- st r uct ur e. Four r epr esent at i ons can t her ef or e be di scer ned wi t hi n ( 1) : an at omi c r epr esent at i on, a si mpl e 6- st r uct ur e, and t wo i t er at ed 6- st r uct ur es of i ncr easi ng compl exi t y. The most compl ex i t er at ed 6- st r uct ur e, vi z. ( i ) , cor r esponds t o t he pr oposi t i on expr essed by ( 1) : (i )
depi ct ed by some r epr esent at i on. Exampl es of hypot het i cal 5- st r uct ur es i ncl ude:
(i i ) (i i i )
I t mi ght have been t he case t hat p I t may be t he case t hat p I f John had come, i t woul d have been t he case t hat p I f John has come, t hen i t i s t he case t hat p
( i v)
Necessar i l y, p I t i s l i kel y t hat p7
3. 3 I t er at ed and si t uat ed 6- st r uct ur es 6- st r uct ur es ar e compl ex r epr esent at i ons wi t h t wo const i t uent s: a r adi cal and a 6- par t . The r adi cal can be ' si mpl e or i t can i t sel f he a 6- st r uct ur e. I f t he r adi cal i t sel f i s a 6- st r uct ur e, t he r esul t i ng r epr esent at i on i s an i t er at ed 6- st r uct ur e such as t he f ol l owi ng: The l andl or d t hought t hat i n 1996, Pet er woul d be penni l ess The st r uct ur e of t hi s r epr esent at i on i s 6( 6( 6p) ) . A si mpl e sent ence, ' Pet er i s penni l ess, ' i s t ur ned i nt o a 6- st r uct ur e by appl yi ng t o i t t he 6- par t ' I n 1996 i t wi l l be t he case t hat . ' A second 6- par t , vi z. ' t he l andl or d t hi nks appl i es t o t he
53
I t was t he case t hat 6( SSp) ) t he l andl or d t hi nks t hat 6( Sp) i n 1996 i t wi l l be t he case t hat 6p Pet er i s penni l ess p
Semant i cal l y, each 6- st r uct ur e r epr esent s some si t uat i on ( i ndi cat ed by t he 6- par t ) as suppor t i ng a cer t ai n f act ( expr essed by t he r adi cal i n t hat 6st r uct ur e) . Let f he a f unct i on f r om 6- par t s t o t he si t uat i ons t hey i ndi cat e; and l et s1, s2 and s3 be t he val ues of t hat f unct i on f or t he ar gument s ' i n 1996 i t wi l l be t he case t hat ' t he l andl or d t hi nks t hat ' and ' i t was t he case t hat ' r espect i vel y. Fi nal l y, l et a be t he f act expr essed by ( i v) . The si mpl e 6- st r uct ur e ( i i i ) expr esses t he f act t hat s1 o; ( i i ) expr esses t he f act t hat 2 >; and ( i ) - t he gl obal 6- st r uct ur e- expr esses t he f act t hat s3 >. W hat f ur t her compl i cat es t he pi ct ur e i s t he f act t hat 6- st r uct ur es, however compl ex, ar e r epr esent at i ons, and as such t hey must t hemsel ves be i nt er pr et ed wi t h r espect t o some si t uat i on. As we have seen, a r epr esent at i on r epr esent s a st at e of af f ai r s, but i t s cont ent cannot be r educed t o t he st at e of af f ai r s i t r epr esent s. The compl et e ( or ' br oad' ) cont ent of a r epr esent at i on r i nvol ves not onl y t he st at e of af f ai r s a whi ch r r epr esent s, but al so t he si t uat i on s i n whi ch t he r epr esent ed f act hol ds. The compl et e cont ent of t he r epr esent at i onal act i s an ' Aust i ni an pr oposi t i on, ' of t he f or m:
Some r eader s may be sur pr i sed by my cl ai m t hat an i ndi cat i ve condi t i onal i s a count er f act ual i - st r uct ur e. I s t her e not a di st i nct i on bet ween i ndi cat i ve condi t i onal s and count er f act ual condi t i onal s? But I have def i ned a count er f act ual si t uat i on t ype as one t hat cont ai ns some count er f act ual si t uat i on; and I mai nt ai n t hat an i ndi cat i ve condi t i onal t al ks about such a si t uat i on t ype. Bot h i ndi cat i ve and count er f act ual condi t i onal s t hus t al k about count er f act ual si t uat i on t ypes. The di f f er ence i s t hat i n a count er f act ual condi t i onal , t he si t uat i on t ype cont ai ns onl y count er f act ual si t uat i ons, whi l e t her e i s no such r est r i ct i on i n t he case of an i ndi cat i ve condi t i onal .
S
A 6- st r uct ur e expr esses a f act whi ch has pr eci sel y t hat f or m; hence 6st r uct ur es expr ess Aust i ni an pr oposi t i ons. But , qua r epr esent at i ons, 6- st r uct ur es t hemsel ves must be i nt er pr et ed wi t h r espect t o si t uat i ons. The comi nvol ves not pl et e cont ent of t he r epr esent at i onal act of usi ng a 6- st r uct ur e
op
onl y t he Aust i ni an f act whi ch i t expr esses, but al so a si t uat i on whi ch i s pr esent ed as suppor t i ng t hat f act . The st r uct ur e of t he compl et e cont ent i s:
54
Eur opean
s'
Revi ew of
The Dynam i cs of Si t uat i ons
Phi l osophy
Hzzsky>>
Her e s i s t he si t uat i on r epr esent ed by t he 6- st r uct ur e as t hat i n whi ch t he f act expr essed by t he r adi cal hol ds; whi l e s' i s t he si t uat i on wi t h r espect t o whi ch t he 6- st r uct ur e i t sel f i s i nt er pr et ed. s and s' pl ay qui t e di f f er ent r ol es: I wi l l say t hat s' i s ' exer ci sed' i n i nt er pr et i ng t he 8- st r uct ur e, whi l e s i s ' ment i oned' by a const i t uent of t he h- st r uct ur e. To di st i ngui sh t he t wo r ol es, I wi l l put t he exer ci sed si t uat i on i nt o br acket s: [ s' ]
<S =u>>
Thi s f or mul a cor r esponds t o what I cal l a si t uat ed 6- st r uct ur e. Let us r et ur n t o our compl ex exampl e of i t er at ed 6- st r uct ur e: " The l andl or d t hought t hat i n 1996, Pet er woul d be penni l ess. " That compl ex 6- st r uct ur e ment i ons t hr ee si t uat i ons 5 and 53 cor r espondi ng t o t he t hr ee 6-
35
don or t he count r y. Thi s i s a poi nt whi ch Eur opean ' st r uct ur al i sm' has much i nsi st ed on: what ever i s si ngl ed out i n speech i s ext r act ed f r oma ' par adi gm' or cont r ast i ve set . I f no ot her l ocat i on was i mpl i ci t l y consi der ed, t he speci f i cat i on of t he l ocat i on woul d pl ay no r ol e and coul d be omi t t ed. The f act t hat t he l ocat i on i s si ngl ed out shows t hat t he si t uat i on wi t h r espect t o whi ch t he l ocat i onal l y speci f i c r epr esent at i on i s i nt er pr et ed i ncl udes t he ment i oned l ocat i on and ot her s f r om t he same par adi gm. The si t uat i on mi ght be, f or exampl e, t he Cont i nent of Eur ope ( whi ch cont ai ns Par i s, Rome, Vi enna, et c. ) . But t he l ocat i onal l y non- speci f i c " I t i s r ai ni ng" can be i nt er pr et ed wi t h r espect t o a smal l er si t uat i on, vi z. Par i s i t sel f ( t o t he excl usi on of any ot her l ocat i on) . Thus we have t he f ol l owi ng cont r ast : ( 1) ( 2)
[ Eur ope] > I Par i s]
par t s, but i t i s cont ext ual l y i nt er pr et ed wi t h r espect t o yet anot her si t uat i on- some si t uat i on i n whi ch t her e i s one and onl y one l andl or d, f or exampl e t he r ent si t uat i on. 8 The r esul t i s a si t uat ed 8- st r uct ur e, wi t h a r at her
A l ocat i onal l y speci f i c ut t er ance such as " I n Par i s, i t i s r ai ni ng" cont ext ual l y expr esses a si t uat ed 6- st r uct ur e l i ke ( 2) . But i f someone, i n Par i s, ut t er s t he l ocat i onal l y non- speci f i c " I t i s r ai ni ng, " t he compl et e cont ent she t her eby
compl ex r epr esent at i onal component :
expr esses may wel l be si mpl er and cor r espond t o ( 1) . Fr omwhat I have sai d i t f ol l ows t hat basi c ver t i cal i nt er pr et i ng i nvol ves a change of cont ext . I n t he t r ansi t i on f r om " I t ' s r ai ni ng" t o " I t ' s r ai ni ng i n
I The r ent si t uat i on]
=> -
{s]
p.
4. Pr oj ect i on
4. 1 Pr oj ect i on as si t uat i onal assumpt i on Gi ven a si t uat ed 8- st r uct ur e I s' ] a>>, consi st i ng of an exer ci sed si t uat i on and an i nner 6- st r uct ur e, pr oj ect i on consi st s i n ' assumi ng' t he ment i oned si t uat i on s whi ch f i gur es i n t he i nner 6- st r uct ur e: t he exer ci sed si t uat i on wi t h r espect t o whi ch t he 6- st r uct ur e i s i nt er pr et ed dr ops out of t he pi ct ur e, and t he ment i oned si t uat i on becomes t he exer ci sed si t uat i on. Pr oj ect i on t her ef or e t akes us f r om ( a) t o ( b): (a) ( b)
[ s' ] > [ sj . . .
I n t hi s sect i on 1 wi l l gi ve a f ew exampl es of pr oj ect i on. W e shal l see t hat pr ocan f r om 6st r uct ur e, and t hat i t can al so oper at e f r om j ect i on oper at e any ' si mpl e' r epr esent at i ons.
4. 2 Pr oj ect i on f r om hypot het i cal ( 5- st r uct ur es I magi ne t he f ol l owi ng di scour se, consi st i ng of f our sent ences Dl t o 134:
The Dynami cs of Si t uat i ons
( D)
John may be ear l y. I f he i s ear l y, he wi l l wai t . He wi l l r ead t he newspaper , or wi l l t hi nk about hi s next book. Ther e won' t be any pr obl em.
Thef i r st sent ence, " John may be ear l y, " posi t s a si t uat i on t ype: i t says t hat , i n some possi bl e si t uat i on, John i s ear l y t o hi s appoi nt ment : Dl
3
5: 5
Not e t hat "I t ' s r ai ni ng her e" i s cl oser t o " I t ' s r ai ni ng i n Par i s" t han t o "I t ' s r ai ni ng" i n bot h r espect s. (i ) The pl ace wher e i t i s r ai ni ng i s expl i ci t l y r epr esent ed bot h i n "I t ' s r ai ni ng i n Par i s" and i n "I t ' s r ai ni ng her e' The cont ent of "I t ' s r ai ni ng her e" i s an i t er at ed 6- st r uct ur e, l i ke t he cont ent of "I t ' s r ai ni ng i n Par i s" ( and unl i ke t he si mpl er cont ent of "I t ' s r ai ni ng" ). (i i ) I n bot h cases t hat pl ace i s vi r t ual l y cont r ast ed wi t h ot her s, hence t he exer ci sed si t uat i on i s l ar ger t han t hat i n ( I ).
>
That 6- st r uct ur e i t sel f i s i nt er pr et ed wi t h r espect t o a si t uat i on s1, a si t uat i on whi ch suppor t s t he f act t hat John maybe ear l y. That ( exer ci sed) si t uat i on i s di f f er ent f r omt he possi bl e si t uat i on i n whi ch John i s ear l y ( ment i oned si t uat i on) . The compl et e cont ent of Dl t her ef or e i s: C( D1) [ s1]
H>>
The second sent ence, D2, says somet hi ng of t he si t uat i on t ype posi t ed by t he f i r st sent ence: t hat i t suppor t s >: D2
TJOhn i s ear l y
>
That 8- st r uct ur e al so must be i nt er pr et ed wi t h r espect t o a si t uat i on. Si mpl i f yi ng somewhat , I wi l l t ake t hat si t uat i on t o be t he same as pr evi ousl y, vi z. .s The compl et e cont ent of t he second sent ence t her ef or e i s:
C( D2) [ s1]
>
At t hi s poi nt pr oj ect i on oper at es, and t he si t uat i on t ype Tj ohn i s S ear l y assumed. The t hi r d sent ence, whi ch descr i bes what John wi l l do, i s i nt er pr et ed wi t h r espect t o t hat si t uat i on t ype. C( W ) [ Tj 0i
i s ear l y]
>
Sent ence ( 4) al so i s i nt er pr et ed wi t h r espect t o t hat si t uat i on: C( D4) [ Tj 0i
10.
57
i s ear l y]
. Nowa par anoi d i s someone who bel i eves hi msel f t o be i n a cer t ai n t ype of si t uat i on. Let us cal l t he si t uat i on John bel i eves hi msel f t o he i n hi s ' bel i ef - wor l d, ' Bel ; 0h1. The second sent ence of ( I ) i s i nt er pr et ed di r ect l y wi t h r espect t o t hat bel i ef wor l d, whi ch i s assumed. The speaker does not ser i ousl y asser t t hat ever ybody spi es John or want s t o ki l l hi m: she expect s t he hear er t o under st and t hat f act as hol di ng i n John' s bel i ef wor l d. Such a shi f t i n poi nt of vi ew i s const i t ut i ve of ' f r ee i ndi r ect speech, ' of whi ch ( 1) i s a t ypi cal i nst ance. I n ( 2) Twai n' s f i ct i on ( a por t i on of t he act ual wor l d) i s ment i oned i n t he f i r st sent ence. NowTwai n' s f i ct i on has a cer t ai n cont ent , t hat i s, i t descr i bes a cer t ai n si t uat i on. The second sent ence of ( 2) i s di r ect l y i nt er pr et ed wi t h r espect t o t hat count er f act ual si t uat i on. I t i s i n t hat si t uat i on t hat t her e ar e a l ot of kni ght s. Si mi l ar l y, i n ( 3) , t he si t uat i on ment i oned i n t he f i r st sent ence
60
The Dynami cs of Si t uat i ons 61
Eur opean Revi ewof Phi l osophy
i s Ber kel ey, and t he second sent ence i s i nt er pr et ed wi t h r espect t o t hat si t uat i on: ' ever ywher e' r anges over l ocat i ons i n Ber kel ey. The di f f er ence bet ween ( 1) - ( 2) and ( 3) wi l l be deal t wi t h i n §5. W hat i s common t o t hose exampl es i s t hat t hey al l i nvol ve pr oj ect i on: t he si t uat i on ment i oned i n t he f i r st sent ence i s ' exer ci sed' and used as cont ext f or t he i nt er pr et at i on of t he second sent ence. Exampl es ( 1) t o ( 3) can be anal ysed as f ol l ows:
t he wor l d of a pi ct ur e, of a novel , of an ut t er ance, or t he bel i ef wor l d of some par t i cul ar per son. Met a- r epr esent at i ons ar e known t o i nvol ve a doubl e l ayer of cont ent . They ar e about some obj ect - r epr esent at i on, but t he obj ect - r epr esent at i on i t sel f i s about somet hi ng, hence t wo l evel s ar e i nvol ved and pr oj ect i on becomes possi bl e. Exampl es l i ke ( 3) , however , showt hat t hat dual st r uct ur e i s an i nst ance of a mor e gener al phenomenon. Ber kel ey i s a ni ce pl ace. Ther e ar e bookst or es and cof f ee shops ever ywher e.
C( 1)
[ a' ] > [ Bel . i ] [ w( Connect i cut Yankee) ] =11
C( 3)
F USA]
Ber kel ey i s not a r epr esent at i on, and t he f i r st sent ence of ( 3) i s not a met ar epr esent at i on. Yet we f i nd t he dual st r uct ur e whi ch makes pr oj ect i on possi bl e. The f i r st sent ence of ( 3) t al ks di r ect l y about Ber kel ey consi der ed as an obj ect : i t st at es a f act of whi ch Ber kel ey i s a const i t uent . But t he second sent ence of ( 3) does not t al k about Ber kel ey i n t hi s manner . The second sent ence of ( 3) expr esses a f act i nt er nal t o Ber kel ey r at her t han a f act ' about ' Ber kel ey i n t he sense of i ncl udi ng Ber kel ey as a const i t uent . The second sent ence of ( 3) i s i nt er pr et ed wi t h r espect t o t he Ber kel ey si t uat i on wi t hout ment i oni ng t hat si t uat i on. That i s possi bl e because t he Ber kel ey si t uat i on has al r eady been ment i oned- i n t he f i r st sent ence. Thi s i s a st andar d i nst ance of pr oj ect i on. ' Si t uat i ons' ar e ent i t i es wi t h a dual char act er . They ar e obj ect s whi ch ent er i nt o r el at i ons and can be t hought and t al ked about ; but at t he same t i me t hey have a ' cont ent , ' whi ch i s a sor t of mi cr o- wor l d, cont ai ni ng t he f act s i nt er nal t o t he si t uat i on. W hen we ment i on a si t uat i on, we evoke i t s cont ent and t hi s i s suf f i ci ent t o make pr oj ect i on possi bl e: ment i oni ng a si t uat i on makes i t s cont ent avai l abl e as a ' cont ext ' f or t he i nt er pr et at i on of ot her r epr esent at i ons whi ch do not ment i on t hat si t uat i on. ( See Sper ber and Wi l son 1986 f or t he r el evant not i ons of ' cont ext ' and ' i nt er pr et at i on. ' ) I t f ol l ows t hat pr oj ect i on i s possi bl e whenever a si t uat i on wi t h a cer t ai n cont ent i s ment i oned; what i s ment i oned need not be a r epr esent at i on.
[ Ber kel ey] 5. Pr oj ect i on f r om met a- r epr esent at i ons 5. 1 Pr oj ect i on, met a- r epr esent at i ons, and t he dual st r uct ur e of si t uat i ons
I n exampl es ( 1) and ( 2) f r om t he pr evi ous subsect i on ( r epeat ed bel ow) , t he base of t he pr oj ect i on i s a met a- r epr esent at i on. (1) (2)
John i s t ot al l y par anoi d. Ever ybody spi es hi mor want s t o ki l l hi m, i ncl udi ng hi s ownmot her . I di d not know you wer e so much i nt er est ed i n kni ght s. You shoul d r ead A Connect i cut Yankee i n Ki ng Ar t hur ' s cour t , by Mar k Twai n. Ther e ar e a l ot of kni ght s.
I n ( 2) t he base r ef er s t o a cer t ai n book, and i n ( 1) i t ( i ndi r ect l y) r ef er s t o John' s ment al st at es. Bot h t he book and t he ment al st at es ar e ' r epr esent at i ons' whi ch have a cer t ai n cont ent , i . e. , whi ch descr i be a cer t ai n si t uat i on. I n met a- r epr esent at i onal pr oj ect i on t he base r ef er s t o a r epr esent at i on, and t he si t uat i on whi ch i s pr oj ect i vel y assumed i s t he si t uat i on whi ch t he ment i oned r epr esent at i on descr i bes. I r ef er t o t hat si t uat i on as w( r ) : t he wor l d of t he r epr esent at i on. Dependi ng on t he sor t of r epr esent at i on r i s, w( r ) wi l l be
11. On t he meani ng of ' w' see bel ow, §5. 1.
5. 2 Si t uat i onal cont ent vs. r epr esent at i onal cont ent
That i s not t o deny t hat t her e ar e cl ear di f f er ences bet ween met ar epr esent at i onal cases l i ke ( l ) - ( 2) , on t he one hand, and non- met a- r epr esent at i onal cases l i ke ( 3) on t he ot her hand. The met a- r epr esent at i onal exampl es, especi al l y ( 1) , cl ear l y i nvol ve somet hi ng l i ke ment al pr oj ect i on.
62
Eur opean Revi ewof Phi l osophy
Fewst udent s of ' f r ee i ndi r ect speech' woul d deny t hi s. The same i nt ui t i on i s commonl y appeal ed t o by t heor i st s of f i ct i on, who mai nt ai n t hat ment al pr oj ect i on i s i nvol ved i n exampl es l i ke ( 2) . Not e t hat t he i nt ui t i on i s weaker i n t he second case t han i t i s i n t he f i r st case; t hat may be because t he pr oj ect i on i s ver y r out i nel y accompl i shed i n our convent i onal way of t al ki ng about f i ct i on. Be t hat as i t may, when we t ur n t o exampl e l i ke ( 3) , i t becomes
The Dynami cs of Si t uat i ons
somet hi ng si mi l ar but mor e compl ex happens. A si t uat i on s i s al so ment i oned dur i ng t he f i r st st age, but t hat si t uat i on s i s a r epr esent at i on and, qua r epr esent at i on, i t descr i bes a si t uat i on w( s) . Nowi t i s t he descr i bed si t uat i on w( s) whi ch i s assumed dur i ng t he second st age, i n i nt ensi onal pr oj ect i on. So we have t he f ol l owi ng cont r ast : Ext ensi onal pr oj ect i on [ s' I [ sj b. . .
hi ghl y cont r over si al t hat ment al pr oj ect i on i s i nvol ved ( as I cl ai m) : we do not have st r ong i nt ui t i ons t o t hat ef f ect , as we do i n t he case of ( 1) . I t hi nk t he i nt ui t i ve di f f er ence bet ween ( l ) - ( 2) and ( 3) poi nt s t o an i mpor t ant di f f er ence bet ween r epr esent at i ons and ot her ent i t i es. Let us compar e Ber kel ey and Twai n' s book. Bei ng a r epr esent at i on, Twai n' s book has somet hi ng whi ch Ber kel ey hasn' t , namel y, a r epr esent at i onal cont ent . I sai d t hat Ber kel ey, qua si t uat i on, has a cont ent , namel y t he set of f act s i nt er nal t o t he si t uat i on; but t hat si t uat i onal cont ent i s not t he same sor t of t hi ng as t he cont ent of Twai n' s book. The si t uat i onal cont ent of s i s t he set of f act s i nt er nal t o s. The di st i nct i on bet ween si t uat i onal cont ent and r epr esent at i onal cont ent i s i mpor t ant because r epr esent at i ons t hemsel ves ar e ' si t uat i ons. ' Qua si t uat i ons, t hey have a dual char act er : t hey ar e obj ect s whi ch ent er i nt o r el at i ons, and t hey have a si t uat i onal cont ent . I n t he same way i n whi ch Ber kel ey has a si t uat i onal cont ent , namel y t he set of f act s whi ch hol d i n Ber kel ey, Twai n' s book has a si t uat i onal cont ent , consi st i ng of ' i nt er nal ' f act s such as t he f act t hat t her e ar e f or t y f i ve chapt er s ( t he f i r st one of whi ch i s not count ed as a chapt er ) , or t he f act t hat a par t i cul ar passage i s bor r owed, l anguage and al l , f r om t he Mor t e d' Ar t hur . W hat i s di st i nct i ve of Twai n' s book, qua r epr esent at i on, i s t hat i t has a r epr esent at i onal cont ent over and beyond i t s si t uat i onal cont ent . The r epr esent at i onal cont ent of Twai n' s book cont ai ns e. g. , t he f act t hat a per son cal l ed Hank Mor gan, cr acked on t he head by a cr owbar i n ni net eent h- cent ur y Connect i cut , wakes t o f i nd hi msel f i n Ki ng Ar t hur ' s Engl and.
Because r epr esent at i ons have a r epr esent at i onal cont ent over and above t he si t uat i onal cont ent whi ch ever y si t uat i on has, t hey al l ow f or a speci al t ype of pr oj ect i on whi ch I cal l ' i nt ensi onal pr oj ect i on. ' Let us f i r st char act er i ze t he ot her t ype of pr oj ect i on: ext ensi onal pr oj ect i on. I n ext ensi onal pr oj ect i on, a si t uat i on s i s ment i oned, and t her eaf t er assumed: dur i ng t he f i r st st age, t he si t uat i on s i s consi der ed as an obj ect and t al ked about f r om t he per spect i ve of a br oader , exer ci sed si t uat i on s' , whi l e t he second st age sees s i t sel f pl ayi ng t he r ol e of t he exer ci sed si t uat i on. I n i nt ensi onal pr oj ect i on,
63
I nt ensi onal pr oj ect i on: [ s' j f f [ w( s) ] H. . The di f f er ence i s i l l ust r at ed by: ( 4)
I ' ve j ust r ead Twai n' s hook. Ther e ar e a l ot of chapt er s.
vs. ( 5)
I ' ve j ust r ead Twai n' s book. Ther e ar e a l ot of kni ght s.
I n bot h cases pr oj ect i on oper at es: t he second sent ence i s i nt er pr et ed wi t h r espect t o an exer ci sed si t uat i on di st i nct f r omt he exer ci sed si t uat i on whi ch gover ns t he i nt er pr et at i on of t he f i r st sent ence. I n ( 4) t he exer ci sed si t uat i on f or t he second sent ence i s t he si t uat i on ment i oned i n t he f i r st sent ence ( Twai n' s book) . I n ( 5) t he exer ci sed si t uat i on f or t he second sent ence i s w( Twai n' s book) : t he st or y t ol d i n t he book. W henever i t i s a r epr esent at i on whi ch i s ment i oned i n t he base, t he t wo t ypes of pr oj ect i on can t ake pl ace; whi l e onl y ext ensi onal pr oj ect i on i s possi bl e when t he ment i oned si t uat i on i s not a r epr esent at i on. For an i l l ust r at i on of t he cont ext ual ambi gui t y t hus gener at ed, consi der t hi s pi ece of di scour se: I ' ve j ust r ead Twai n' s book. Ther e ar e a l ot of pages. ' Page' can mean ei t her one si de of t he l eaf of a book, or a young mal e at t endant on ki ngs, nobl es et c. I n t he cont ext of " I ' ve j ust r ead Twai n' s hook, " bot h r eadi ngs make sense. I n t he ' l eaf ' r eadi ng t he second sent ence st at es a f act concer ni ng Twai n' s book ( ext ensi onal pr oj ect i on) ; i n t he ' young mal e' r eadi ng i t st at es a f act concer ni ng t he st or y t ol d by Twai n' s book ( i nt ensi onal pr oj ect i on) .
64
Eur opean Revi ewof Phi l osophy
The di f f er ence bet ween ext ensi onal and i nt ensi onal pr oj ect i on account s f or t he di f f er ence i n our i nt ui t i ons concer ni ng ( 1) , ( 2) and ( 3) above; i t account s f or t he f act t hat , i n t he Ber kel ey exampl e, t he pr oj ect i on i s bar el y not i ced, whi l e i n t he ' par anoi d' exampl e t he shi f t of poi nt of vi ew i s ver y mani f est . Ext ensi onal pr oj ect i on i s l i ke zoomi ng a det ai l s f r omsome or i gi nal scene ( exer ci sed si t uat i on) s' i s f ocused on i n such a way t hat i t becomes t he scene ( t he exer ci sed si t uat i on) . The di f f er ence bet ween t he f i r st st age ( when s i s seen agai nst t he backgr ound of ot her t hi ngs i n s' ) and t he second st age ( when s i t sel f i s t he backgr ound agai nst whi ch t hi ngs i n s ar e seen) i s a di f f er ence of gr ai n. I n i nt ensi onal pr oj ect i on, however , s i s not a por t i on of s' whi ch has been f ocused on, but an al t er nat i ve uni ver se. The shi f t i s much mor e dr amat i c and not i ceabl e. The i nt ui t i ve di f f er ence bet ween t he t wo t ypes of case shoul d not be over est i mat ed, however . As I al r eady ment i oned, exampl e ( 2) i s i nt er medi at e bet ween ( 1) and ( 3) as f ar as t he st r engt h of our i nt ui t i ons i s concer ned. And exampl es of i nt ensi or i al pr oj ect i on i n whi ch t he pr oj ect i ve st ep i s hat f ol l ows i s an exampl e t aken har dl y not i ced ar e not di f f i cul t t o f i nd. W al most at r andom. 5. 3 An exampl e of i nt ensi onal pr oj ect i on Open t he f i r st encycl opedi a of phi l osophy t hat comes t o hand. I n a t ypi cal ent r y f r om such an encycl opedi a, you wi l l f i nd sent ences expl i ci t l y about t he phi l osopher ' s l i f e and vi ews ( e. g. , sent ences begi nni ng wi t h " Accor di ng t o Lei bni z. ), f ol l owed by sent ences di r ect l y pr esent i ng t hose vi ews. The f ol l owi ng passage f r om Ur mson' s ' Conci se Encycl opedi a' pr ovi des an exampl e: Lei bni z' s [ met aphysi cs) i s compl et ed by hi s pr oof s of t he exi st ence of God. The syst emof cr eat ed monads i s, i n a sense, compl et e i n i t sel f , t hat i s, i t i s necessar i l y as i t i s gr ant ed t hat any par t of i t exi st s. But no one par t of i t cont ai ns t he r eason f or i t s own exi st ence, so t hat t he r eason f or i t s exi st ence must l i e i n a bei ng whi ch does cont ai n i t s own r eason f or exi st ence, t hat i s, i n a necessar y bei ng, whi ch we cal l God. Thi s ar gument , t he ' cosmol ogi cal ar gument ' , appear s i n t he Monadol ogy and t her e i s not hi ng i n i t whi ch i s pecul i ar t o Lei bni z ( Ur mson, ed. , 1975; 156) .
The Dynami cs of Si t uat i ons
65
who speaks, as i t wer e. The f i r st and l ast sent ence r ef er t o Lei bni z and hi s vi ew, and expr ess t he bel i ef s of t he aut hor of t he ar t i cl e, r at her t han Lei bni z' s bel i ef s. I n t he si t uat i onal f r amewor k, t he passage can be descr i bed as f ol l ows. The aut hor of t he ar t i cl e t al ks about one par t i cul ar aspect of t he hi st or y of West er n phi l osophy and phi l osopher s, vi z. Lei bni z and hi s met aphysi cal syst em. That i s t he exer ci sed si t uat i on f or t he f i r st and l ast sent ences of our passage: t he aut hor of t he ar t i cl e st at es t wo f act s concer ni ng t hat por t i on of West er n phi l osophy, namel y t he f act t hat Lei bni z' s doct r i ne of monads i s compl et ed by hi s pr oof s of t he exi st ence of God ( f i r st sent ence) , and t he f act t hat t he Cosmol ogi cal ar gument can be f ound i n t he Monadol ogy and i s not pecul i ar t o Lei bni z ( l ast sent ence) . I n cont r ast , t he exer ci sed si t uat i on f or t he i nt er medi at e sent ences i s not t he por t i on of t he wor l d whi ch i ncl udes Lei bni z and hi s wor ks, but a count er f act ual si t uat i on, vi z. t he si t uat i on whi ch Lei bni z' s met aphysi cal wr i t i ngs descr i be: t he ' Lei bni zi an uni ver se, ' as we may cal l i t . That i s a ver y di f f er ent si t uat i on f r omt he por t i on of t he act ual wor l d whi ch t he aut hor of t he encycl opedi a ar t i cl e t al ks about . The Lei bni zi an uni ver se i ncl udes God, monads, et c. ; t he por t i on of @ whi ch t he aut hor of t he ar t i cl e t al ks about i ncl udes Lei bni z, Ar naul d, and t he Royal Li br ar y at 1- l anover . The i nt er medi at e sent ences ar e meant as ( di r ect ) descr i pt i ons of t he Lei bni zi an uni ver se; t hey ar e not an el l i pt i cal descr i pt i on of Lei hni z descr i bi ng t he Lei bni zi an uni ver se. I n ot her wor ds, pr oj ect i on oper at es. But pr oj ect i on i s such a per vasi ve mechani smt hat i t i s bar el y not i ced even t hough, i n t hi s par t i cul ar case, i t bel ongs t o t he mor e dr amat i c, i nt ensi onal var i et y. 6. Si t uat i ons and cogni t i ve cont ent
6. 1 Pr oj ect i on vs. el l i psi s Let us go back t o t he Ber kel ey exampl e: Ber kel ey i s a ni ce pl ace. Ther e ar c bookst or es and cof f ee shops ever ywher e.
Ther e i s a cl ear cont r ast bet ween t he f i r st and l ast sent ence of t hat passage on t he one hand, and t he i nt er medi at e sent ences on t he ot her . The i nt er me-
Most l anguage t heor i st s woul d say t hat " Ther e ar e bookst or es and cof f ee shops ever ywher e" i s el l i pt i cal f or " I n Ber kel ey, t her e ar e bookst or es and cof f ee shops ever ywher e. " I n what sense i s my cl ai mabout pr oj ect i on di f f er ent
di at e sent ences st at e Lei bni z' s own vi ew of t he mat t er : i t i s Lei bni z hi msel f
f r om t he t r adi t i onal el l i psi s cl ai m? W hat i s t he di f f er ence bet ween sayi ng
66
Eur opean Revi ew of Phi l osophy
The Dynami cs of Si t uat i ons
67
t hat t he sent ence i s el l i pt i cal f or " I n Ber kel ey t her e ar e bookst or es and cof f ee shops ever ywher e" and sayi ng t hat t he Ber kel ey si t uat i on i s cont ext ual l y assumed? I agr ee t hat Ber kel ey i s an aspect of t he cont ent of " Ther e ar e bookst or es
I n t hi s t heor y t her e i s a semant i c di f f er ence bet ween " Ther e ar e bookst or es and cof f ee shops ever ywher e, " whi ch st at es a f act concer ni ng Ber kel ey, and " I n Ber kel ey t her e ar e bookst or es and cof f ee shops ever ywher e, " whi ch st at es a di f f er ent f act ( a f act about Ber kel ey) , concer ni ng a wi der por t i on of
and cof f ee shops ever ywher e' i n t he f ol l owi ng sense. The speaker ( or t he t hi nker ) has Ber kel ey i n mi nd when she says " Ther e ar e bookst or es and cof f ee shops ever ywher e. " She i s t al ki ng of Ber kel ey, not of Par i s or any ot her ci t y. I n Per r y' s t er mi nol ogy, Ber kel ey i s an ' unar t i cul at ed const i t uent ' of t he
t he uni ver se. That semant i c di f f er ence i s t he r eason why I r ej ect t he not i on t hat one ut t er ance i s mer el y ' el l i pt i cal ' f or t he ot her ( hence synonymous wi t h i t ) .
cont ent expr essed by t he second sent ence. St i l l , t her e ar e di f f er ent ways of const r ui ng t hi s not i on of an unar t i cul at ed const i t uent . To say t hat " Ther e ar e bookst or es and cof f ee shops ever ywher e" i s el l i pt i cal f or " I n Ber kel ey t her e ar e bookst or es and cof f ee shops ever ywher e" i s t o opt f or a par t i cul ar const r ual , one whi ch I t hi nk must be r ej ect ed. I n t he Aust i ni an f r amewor k t he compl et e cont ent of a r epr esent at i on i s t wo- f ol d: i t consi st s of t he f act whi ch t he r epr esent at i on expr esses, pl us t he si t uat i on whi ch t hat f act concer ns. The st r uct ur e of t he compl et e cont ent i s: [ 51
H
Si nce t he compl et e cont ent of an ut t er ance or t hought i s an Aust i ni an pr oposi t i on wi t h t wo component s, t her e ar e t wo opt i ons f or unar t i cul at ed const i t uent s: t hey can bel ong ei t her t o t he si t uat i onal component , or t o t he ' r adi cal ' ( i . e. , t he r i ght - hand si de of t he Aust i ni an pr oposi t i on) . Accor di ng t o my t heor y of pr oj ect i on, t he si t uat i on ment i oned by t he base det er mi nes t he ' exer ci sed si t uat i on' wi t h r espect t o whi ch t he r epr esent at i on t hat f ol l ows i s pr oj ect i vel y i nt er pr et ed. The compl et e cont ent of " Ther e ar e bookst or es and cof f ee shops ever ywher e" t her ef or e i s: [ Ber kel ey] H I n ot her wor ds, t he f act whi ch i s st at ed ( on t he r i ght - hand si de) does not i ncl ude Ber kel ey as a const i t uent ; i n t hat r espect i t i s ver y di f f er ent f r om t he f act st at ed by t he base, namel y, t he f act t hat Ber kel ey i s a ni ce pl ace. For t hat f act i s about Ber kel ey i n t he st r ong sense: i t i ncl udes i t as a const i t uent . The compl et e cont ent of t he base i s: [ The US si t uat i on]
Her e we f i nd Ber kel ey on t he si de of t he r adi cal , r at her t han on t he si t uat i onal si de.
6. 2 ' Concer ni ng' and ' bei ng about ' The di f f er ence bet ween ' concer ni ng' and ' bei ng about ' has been el abor at ed by Per r y i n hi s paper " Thought wi t hout r epr esent at i on" ( Per r y 1986b) . That di f f er ence comes out most cl ear l y i n t he case of t he Z- l ander s, a gr oup of peopl e who " do not t r avel t o, or communi cat e wi t h r esi dent s of , ot her pl aces" and have no name f or t he pl ace t hey l i ve i n. W hen a Z- l ander sees r ai n, he wi l l say t o ot her s not i n a posi t i on t o l ook out door s, I t i s r ai ni ng. Hi s l i st ener s t hen act appr opr i at el y t o t her e bei ng r ai n i n Zl and: t hey cl ose t he wi ndows i n Z- l and, cancel pl ans f or 7- l and pi cni cs, and gr ab umbr el l as bef or e goi ng i nt o t he Z- l and out - of - door s. They have no ot her use f or " i t i s r ai ni ng. " They do not cal l t hei r sons i n f ar - of f pl aces, or l i st en t o t he weat her news, or r ead newspaper s wi t h nat i onal weat her r epor t s. ( Per r y 19866: 212) . As Per r y poi nt s out , Z- l and i s an unar t i cul at ed const i t uent of t he cont ent expr essed by t he Z- l ander ' s ut t er ance " I t ' s r ai ni ng. " The ut t er ance i s t r ue i f and onl y i f i t i s r ai ni ng i n Z- l and. But t he Z- l ander s do not have a concept or i dea of Z- l and as opposed t o ot her pl aces. Thei r weat her t hought s ' concer n' 7- l and, not by vi r t ue of cont ai ni ng a r epr esent at i on of Z- l and ( i n whi ch case t hey woul d be ' about ' Z- l and) , but by vi r t ue of t hei r bei ng i n Z- l and. I n ot her wor ds, t he unar t i cul at ed const i t uent i s unar t i cul at ed not onl y l i ngui st i cal l y but al so ment al l y: i t ' s a const i t uent of cont ent di r ect l y pr ovi ded by t he envi r onment . I n such cases t he ment al r epr esent at i on, consi der ed i n abst r act i on f r om t he envi r onment whi ch i t concer ns, expr esses l ess t han a compl et e pr oposi t i on. The Z- l ander s t hi nk " I t i s r ai ni ng: " t he cont ent t hus ar t i cul at ed i s not f ul l y pr oposi t i onal - i t i s a pr oposi t i onal f unct i on, whi ch i s t r ut h- eval uabl e onl y wi t h r espect t o a par t i cul ar pl ace. Let us cal l t hat ar t i cul at ed cont ent t he ' cogni t i ve' or ' nar r ow' cont ent of t he r epr esent at i on. I t i s onl y when t he envi r onment i s t aken i nt o account t hat a compl et e, ' br oad' cont ent i s expr essed, whi ch i ncl udes Z- l and as an unar t i cul at ed const i t uent . I n t he
68
The Dynami cs of Si t uat i ons
Eur opean Revi ewof Phi l osophy
Aust i ni an f r amewor k, t he di st i nct i on bet ween t he ( ' nar r ow' ) ment al component and t he envi r onment al component cor r esponds t o t he di st i nct i on bet ween t he exer ci sed si t uat i on and t he r adi cal . The compl et e cont ent of t he Z- l ander ' s t hought or ut t er ance i s: [ Z- l and]
> [ Eur ope] H<Par i s =