VYTAUTO DIDŽIOJO UNIVERSITETAS

(VYTAUTAS MAGNUS UNIVERSITY)

Mykolas Jurgis Drunga

THE TIME-GAP ARGUMENT: DOES IT REFUTE DIRECT REALISM?

(AR LAIKO ATOTRŪKIO ARGUMENTAS SUGRIAUNA TIESIOGINĮ REALIZMĄ?)

Daktaro disertacija Humanitariniai mokslai, filosofija (01 H)

Kaunas, 2011

2

Disertacija ginama eksternu Mokslinis konsultantas: prof. dr. Gintautas Mažeikis (Vytauto Didžiojo universitetas, humanitariniai mokslai, filosofija – 01 H)

3

VYTAUTO DIDŽIOJO UNIVERSITETAS

MYKOLAS JURGIS DRUNGA

THE TIME-GAP ARGUMENT: DOES IT REFUTE DIRECT REALISM?

(AR LAIKO ATOTRŪKIO ARGUMENTAS SUGRIAUNA TIESIOGINĮ REALIZMĄ?)

Daktaro disertacijos santrauka

Humanitariniai mokslai, filosofija (01 H)

Kaunas, 2011

4

Disertacija ginama eksternu Mokslinis konsultantas:

prof. dr. Gintautas Mažeikis (Vytauto Didžiojo universitetas, humanitariniai mokslai, filosofija – 01 H) Disertacija ginama Vytauto Didžiojo universiteto Filosofijos mokslo krypties taryboje: Pirmininkas

prof. dr. Leonidas Donskis (Europos parlamentas, humanitariniai mokslai, filosofija – 01 H)

Nariai:

prof. dr. Dalius Jonkus, humanitariniai mokslai, filosofija 01 H, Vytauto Didžiojo universitetas prof. dr. Timo Airaksinen, humanitariniai mokslai, filosofija 01 H, Helsinkio universitetas prof. dr. Heta Gylling, humanitariniai mokslai, filosofija 01 H, Helsinkio universitetas doc. dr. Elena Lisanyuk, humanitariniai mokslai, filosofija 01 H, Sankt Peterburgo valstybinis universitetas

Disertacija bus ginama viešame Filosofijos mokslo krypties tarybos posėdyje 2011 m. gruodžio mėn. 9 d. 11 val. Vytauto Didžiojo universiteto Humanitarinių mokslų fakulteto prof. M. Gimbutienės (211) auditorijoje. Adresas: K.Donelaičio 52, Kaunas, Lietuva Disertacijos santrauka išsiuntinėta 2011 m. lapkričio mėn. 9 d. Disertaciją galima peržiūrėti Vytauto Didžiojo universiteto ir Lietuvos nacionalinėje Martyno Mažvydo bibliotekose.

5

Ar laiko atotrūkio argumentas sugriauna tiesioginį realizmą?

1. Įvadas

Tai Lietuvos filosofinėje literatūroje dar nesvarstytas klausimas, kaip apskritai tik mažai

nagrinėta ir visa sritis (pažinimo teorija, kitaip vadinama epistemologija ar gnoseologija), kuriai

jis priklauso. Todėl tam, kad šį klausimą adekvačiai nušviestume, sunku atsiremti į kokią nors

gyvybingą lietuviško filosofavimo tradiciją1. Tuo tarpu užsienio literatūroje, ypač anglakalbėje,

pažinimo teorijos problematika aptariama nepalyginamai gausiau ir nuodugniau. Tai viena iš

priežasčių, kodėl šis darbas – anglų kalba. Antra priežastis ta, kad autorius užaugo ir filosofiją

(visų pirma analitinę) studijavo Jungtinėse Amerikos Valstijose ir ten pradėjo šią disertaciją

rašyti, bet jos nebaigė. Todėl šis darbas pirmiausia priskirtinas prie anglakalbių epistemologijos

istorijos bandymų. Tačiau dėl to, kad jis ginamas Lietuvos universitete, autorius šiuo darbu

norėtų įspausti pėdą ir Lietuvos filosofiniame gyvenime. (Trečią priežastį paminėsiu pačiame

šio rašinio gale.)

Šios disertacijos tikslas yra duoti originalų įnašą į šiuo metu analitinėje filosofijoje itin

„nemadingo“ teorinio idealizmo gynybą. Teorinį idealizmą (kurį netrukus tiksliau apibūdinsiu)

reikia skirti nuo praktinio idealizmo, kuris yra etinė laikysena, reikalaujanti siekti tam tikrų

idealų. Šiame darbe kalbėsiu tik apie teorinį idealizmą, vadinsiu jį tiesiog „idealizmu“ ir

susitelksiu į vieną jo versiją, būtent į subjektyvųjį idealizmą. Jo centrinę tezę galima

suformuluoti taip: būti – tai (juslėmis ar protu) suvokti arba būti (juslių ar proto) suvoktam.

Vadinasi, būti kitaip nei suvokėjui arba suvoktam daiktui – neįmanoma.

Šiai tezei paremti nagrinėsiu tai, ką lietuviškai būtų galima pavadinti „laiko atotrūkio

argumentu“. Apžvelgdamas šio argumento istoriją nuo G. W. F. Leibnitzo iki dabarties laikų,

sukonstruosiu tokią šio argumento versiją, kurią būtų galėjęs panaudoti subjekyviojo idealizmo

tėvas George‘as Berkeley‘jus, idant dar aiškiau parodytų savo doktrinos teisingumą.

Berkeley‘jiškos dvasios laiko atotrūkio argumentas siekia pirmiausia įrodyti vaizdinių buvimą ir

po to sugriauti ne tik (šiuo metu dominuojančio) tiesioginio, bet ir (mažiau populiaraus)

netiesioginio realizmo neidealistinę versiją ir šitokiu būdu, kadangi kitų alternatyvų nėra, įrodyti

1 Sunku, bet ne absoliučiai neįmanoma, nes trupučiuką atsišlieti vis dėlto galima į Vosyliaus Sezemano (kone vienintelio) raštus – jis ne tik pats buvo europinio lygio filosofas, pasireiškęs kaip pažinimo teorijos ir estetikos žinovas, bet ir ženkliai prisidėjo prie atitinkamos lietuviškosios filosofinės terminijos sukūrimo. Užtat šiame rašinyje daugeliu jo pasiūlytų terminų ir naudosiuosi. Tik ne visais: pvz., vietoje „gnoseologija“ visuomet, kaip anglo-amerikietiškoje aplinkoje įprasta, rašau „epistemologija“ (išskyrus, kai cituoju). Šiaip dauguma V. Sezemano naudojamų terminų – puikūs, pvz.., „pagava“ kaip „percepcijos“ sinonimas.

6

(visiškai nepopuliarų angliškai rašančiame pasaulyje) subjektyvųjį metafizinį bei epistemologinį

idealizmą kaip iš visų pozicijų vienintelę galimą priimti.

Tad trumpas atsakymas į titulinį disertacijos klausimą būtų toks: taip, laiko atotrūkio

argumentas didele dalimi sugriauna tiesioginį realizmą, jei pastarasis suvokiamas, kaip paprastai

tai daroma, neidealistiškai, tačiau jis ne griauna, o, priešingai, palaiko tiesioginį realizmą, jei tas

suvokiamas idealistiškai.

Dabar platėliau aptarsiu kelių šiame darbe vartojamų pagrindinių terminų reikšmę.

2. Kai kurios esminės sąvokos

Analitinės filosofijos tradicijoje skiriamos trys doktrinos, atsakančios į klausimą, kokie yra

percepcijos, kitaip sakant, juslinės pagavos (regos, klausos, lytėjimo, uoslės, skonio) objektai

arba, dar kitaip tariant, kokie dalykai yra juslių pagaunami, t. y., matomi, girdimi, lytėjimu

jaučiami, užuodžiami, ragaujami. Šios trys doktrinos tai – (1) tiesioginis (betarpiškasis)

realizmas (kartais jis pats arba viena jo atmaina dar vadinama naiviuoju realizmu), (2)

netiesioginis realizmas (kartais vadinamas atstovaujamuoju, kartais kritiniu realizmu) ir (3)

idealizmas (kartais arba visiškai tapatinamas su fenomenalizmu, arba atskiriamas nuo

fenomenalizmo, bet laikomas jam giminingu).

Idant paaiškintume, kuo šios doktrinos skiriasi, paranku pasitelkti dar vieną skirtį – tarp

epistemologinių ir metafizinių (kitaip sakant, ontologinių) realizmo ir idealizmo rūšių.

Tiesioginis ir netiesioginis realizmai yra epistemologinės doktrinos, kurios dažniausiai remiasi

metafiziniu (ontologiniu) realizmu, bet gali remtis ir metafiziniu idealizmu.

Metafizinis realizmas teigia, jog už sąmonės (vadinamojo „vidinio“ pasaulio) ribų visiškai

savarankiškai egzistuoja nuo jos nepriklausomas materialusis pasaulis (vadinamas „išoriniu“),

kuris yra fundamentalus (nesuvedamas į nieką kita) ir kurį, - čia jau prasideda epistemologiniai

realizmai, - sąmonė gali pagauti arba tiesiogiai (betarpiškai), arba netiesiogiai (tarpiškai). Tuo,

kad sąmonė jį gali pagauti betarpiškai, įsitikinę tiesioginiai realistai. Anot jų, kai žiūrime į baltą

sieną, tai ją – ir tik ją – iškart ir matome. O tai, jog materialaus pasaulio daiktus galima pažinti

tik netiesiogiai, pirmiausia betarpiškai pagavus jiems atstovaujančius, juos pavaduojančius

sąmonės duomenis (kurie tik sąmonėje ir teegzistuoja) ir tik paskui per juos, taigi netiesiogiai

pažinus pačius materialiuosius daiktus (tų sąmonės duomenų priežastis), teigia netiesioginiai

realistai. Anot pastarųjų, kai, pvz., matome baltą sieną, iš tiesų matome du dalykus – sienos

formos baltą sąmonės duomenį ir jį sukėlusią baltą fizinę sieną.

Šiuos du epistemologinius (tiesioginį ir netiesioginį) realizmus grindžiantį metafizinį

realizmą atmeta metafizinis idealizmas. Pastarasis teigia, pirma, kad fundamentali, savarankiška

7

būtis yra tik sąmonė ir, antra, kad (o tai galime vadinti idealistine išlyga) materialusis pasaulis

yra išvestinis, egzistuojantis tik kaip loginė sąmonės konstrukcija. Vis dėlto su metafiziniu

idealizmu yra sutaikomi ir (kitais atvejais metafizinio realizmo ramstomi) epistemologiniai

realizmai, pripažįstantys, jog galima (tiesiogiai arba netiesiogiai) pažinti materialius daiktus, bet

prileidžiantys taip pat galimybę, jog šitie daiktai, laikantis idealistinės išlygos ir priešingai

metafiziniam realizmui, neturi fundamentalios, nuo sąmonės nepriklausomos būties.

Pats idealizmas gali būti ne tik metafizinis, bet ir epistemologinis. Jeigu epistemologinis

realizmas teigia, kad sąmonė sugeba pagauti (tiesiogiai ar netiesiogiai) materialius daiktus, tai

epistemologinis idealizmas teigia, kad sąmonė sugeba pagauti savo pačios turinius. Tačiau šitaip

šias dvi doktrinas apibrėžus, neaišku, kodėl jos laikomos viena kitai priešingos. Be to, pagal

tokias apibrėžtis jos abi atrodo akivaizdžiai teisingos: juk kas gi ginčytų, kad sąmonė sugeba

pažinti ir materialius daiktus, ir savo pačios turinius2? Tai reiškia, jog norint atspindėti šių

doktrinų priešingybę, taip pat išlaikyti principą, jog tik viena jų gali būti teisinga, reikia

duotuosius apibrėžimus šiek tiek pakeisti. Viena galimybė – priimti Sezemano siūlymą

epistemologiniu idealizmu vadinti teoriją, „pasak kurios mūsų pažinimas yra apribotas sąmonės

sferos, t. y. pažinimo objektai tėra tik sąmonės turiniai (idealioji būtis platesne prasme)“, o

epistemologiniu realizmu atitinkamai vadinti požiūrį, kuris pripažįsta, „kad pažinimui prieinama

ir nepriklausoma nuo sąmonės būtis (pvz., savarankiškai egzistuojąs išorinis pasaulis)“3.

Vadinasi, pagal tai mūsų pirminės apibrėžtys pakeistinos taip: „Jeigu epistemologinis realizmas

teigia, kad sąmonė sugeba, be savo pačios turinių, pagauti (tiesiogiai ar netiesiogiai) dar ir

materialius daiktus, tai epistemologinis idealizmas teigia, kad sąmonė sugeba pagauti vien tik

savo pačios turinius“. Vis dėlto pastarasis teiginys nereiškia, jog, pagal epistemologinį

idealizmą, sąmonei visiškai negalima pagauti materialių daiktų, o tik tai, kad tokie daiktai nėra

už sąmonės sferos, kad jie tiesiog iš sąmonės turinių sukonstruoti.

Sezemanas tatai gerai supranta ir tą jo supratimą rodo kad ir šios citatos. Pirmiausia jis teigia,

jog metafizinis idealizmas yra „teorija, tvirtinanti, kad visai būčiai pamatu eina dvasinė

(psichinė), arba imateriali, būtis“. Tai aiškus apibrėžimas: būtis, pasireiškianti sąmonės forma,

yra pamatas, fundamentas, iš kurio gali išdygti ir materialūs daiktai. „Šia prasme idealizmas yra

materializmo, laikančio medžiagą būties pradu, priešybė ir sutampa maždaug su spiritualizmu“.

Toliau Sezemanas dėsto štai ką (leisdamas suprasti, kad bent šiuo atveju jis epistemologijai

teikia pirmenybę prieš metafiziką):

2 Ginčytų tie, kurie nepripažįsta materialių daiktų buvimo, ir tie, kurie nepripažįsta sąmonės turinių buvimo. Tačiau šiuos nihilistinius požiūrius galima iš karto atmesti kaip mažai pagrįstus. 3 Vosylius Sezemanas, Raštai. Filosofijos istorija, kultūra. Vilnius, „Mintis“, 1997, p. 348-349.

8

„Kadangi klausimą: kas iš tikro yra? galima kritiškai nušviesti tik išsiaiškinus, kas galima

pažinti, tai gnoseologinis idealizmas dažniausiai siejasi su ontologiniu idealizmu“, nors

„ontologinis (metafizinis) idealizmas gali jungtis ne tik su gnoseologiniu idealizmu, bet ir su

gnoseologiniu realizmu“4. Čia Sezemanas labai taikliai nusako filosofinių doktrinų tarpusavio

ryšius.

3. Kodėl reikia pripažinti vaizdinius?

Šiame darbe aš ir imuosi apginti epistemologinį bei metafizinį idealizmą, sujungtą su

epistemologiniu realizmu. Konkrečiau kalbant, siekiu apginti tą subjektyvųjį idealizmą

(fenomenalizmą), kuriam ryškiausiai atstovauja George‘as Berkeley‘jus5, bet kuriam šiek tiek

pritariančių išraiškų galima rasti ir Gottfriedo Wilhelmo Leibnizo, Immanuelio Kanto bei

Johanno Gottliebo Fichtės ir ypač Arthuro Schopenhauerio raštuose6. Šis idealizmas yra

subjektyvus, nes pabrėžia pažįstantįjį subjektą kaip viso pažinimo išeities tašką bei teigia, jog tik

sąmoningi, individualūs subjektai gali ką nors pažinti, ir jis tolygus fenomenalizmui, nes teigia

(čia remiuosi V. Sezemano pateiktu apibrėžimu, tik dviem sakiniais jį papildau), „kad visas

mūsų žinojimas, jo turinys“ – tiesa, ne visas apskritai, o tik tas, kuris apima materialųjį pasaulį

(tai pirmas papildymas) – „susideda iš sąmonės būsenų (įspūdžių, idėjų) arba sąmonės

reiškinių“7, kurie esamomis sąlygomis čia ir dabar tikrai egzistuoja arba čia ir dabar egzistuotų,

jeigu esamos sąlygos būtų buvusios kitokios (tai antras papildymas).

4 ten pat, p. 349. 5 Rašau taip, o ne „Berkeley‘s“, „Berkeley‘is“ ar „Berklis“, nes „Berkeley‘jus“ geriausiai atitinka anglišką jo pavardės rašybą su lietuviška galūne. Lietuviškai reiktų tarti „Barklėjus“. 6 žr., pvz., Parerga ir paralipomena, Vilnius, „Pradai“, 2001. Čia Schopenhaueris pagrįstai teigia, kad Berkeley’jus „nuosekliai ėjo dekartininkų keliu ir dėl to tapo tikrojo idealizmo pradininku, t. y. tos sampratos, kad viskas, kas yra ištįsę erdvėje ir ją pripildo, taigi apskritai regimasis pasaulis, savo būtį kaip pačią savaime gali turėti absoliučiai tik mūsų vaizdinyje ir kad yra absurdiška, net prieštaringa priskirti jam dar būtį už vaizdinio ribų ir nepriklausomą nuo pažįstančiojo subjekto ir tuo remiantis pripažinti savaime egzistuojančią materiją. Tai yra labai teisinga ir gili įžvalga – joje glūdi visa Berkeley‘jaus filosofija“, p. 20-21. Tuoj pat po žodžio „materiją“ Schopenhaueris prideda tokią išnašą: „Filosofams diletantams, prie kurių priklauso daug jos daktarų, žodį „idealizmas“ reikėtų visai išplėšti iš rankų, nes jie, nežinodami, ką jis reiškia, krečia su juo visokias išdaigas: idealizmas jiems kai kada yra spiritualizmas, kai kada – kažkas panašaus į filisteriškumo priešingybę, ir šiai jų pažiūrai pritaria bei ją remia vulgarieji literatai“. Tačiau Schopenhaueris nepateikia jokių argumentų, kodėl viena savo reikšme idealizmas negali būti tapatus spiritualizmui, o kita – praktiniam idealizmui, taigi „filisteriškumo priešingybei“. Toje pačioje išnašoje jis toliau aiškina taip: „Žodžiai „idealizmas“ ir „realizmas“ nėra benamiai klajūnai – jie turi savo aiškias filosofines reikšmes; jei kas mano ką nors kitka, tevartoja kitą žodį. Idealizmo ir realizmo priešprieša apibūdina tai, kas pažinta, objektą, o spiritualizmo ir materializmo priešprieša – tą, kuris pažįsta, subjektą. (Nieko neišmanantys šių laikų postringautojai idealizmą ir spiritualizmą tiesiog painioja.)“. Bent jau Sezemanas juos ne „painioja“, o tiesiog sutapatina pagal labai seną tradiciją, kad idealizmas viena reikšme (epistemologine) yra priešprieša realizmui, o antra (metafizine) – materializmui. Taigi nors cituotoje Schopenhauerio išnašoje beveik viskas – medus, deja, pakliuvo ir šaukštelis deguto. 7 Vosylius Sezemanas, Raštai. Gnoseologija. „Minties“ leidykla, Vilnius, 1987, p. 260.

9

O tam, kad apginčiau (šiuo metu anglo-amerikiečių filosofijoje labai mažai šalininkų turintį)

subjektyvųjį idealizmą, turiu sugriauti tarp minėtųjų filosofų šiuo metu vyraujantį tiesioginį

realizmą, taip pat ir ano šimtmečio pirmojoje pusėje vyravusį netiesioginį realizmą, bet, aišku,

sugriauti juos tik tiek, kiek jie neidealistiški ir remiasi metafiziniu realizmu.

Tuo tikslu visų pirma bandysiu įrodyti, kad egzistuoja ir pažinime kertinį vaidmenį vaidina

dalykai, kurie net ir lietuviškai yra ar gali būti įvardijami gana įvairiai. Tai – jutimai, (jutiminiai,

jusliniai) įspūdžiai, (jutiminės, juslinės) idėjos, jutimų (juslių) duomenys, jusliniai sąmonės

turinio duomenys (arba tik: jusliniai sąmonės duomenys), jusliniai sąmonės reiškiniai,

fenomenai, vaidiniai, vaizdiniai, ir taip toliau. Vadinkime visus šiuos dalykus be skirtumo

vaizdiniais. Ir nors dauguma šiuolaikinių analitinių filosofų ir beveik visi tiesioginiai realistai

vaizdinių buvimą neigia, juos idealistai ir netiesioginiai realistai įveda tam, kad paaiškintų

iliuzijas (tikrovės iškraipymus) ir haliucinacijas (nesamų daiktų tariamas pagavas).

Pvz., žmogus žiūri į baltą sieną, tačiau jam (dėl kokio nors jo paties organinio negalavimo,

dėl tam tikro sienos apšvietimo ar dėl kitos priežasties) ji atrodo mėlynai. Kaip šią iliuziją

paaiškinti? Arba, žmogus žiūri į tą pačią baltą sieną, tačiau jam (kadangi jis chroniškas

girtuoklis ar narkotikų prisirijęs) atrodo, kad ant tos sienos raitosi oranžinės spalvos driežai.

Kaip paaiškinti šią haliucinaciją? Pasak idealistų ir netiesioginių realistų, iliuzijos atveju žmogus

tiesiogiai pagauna mėlyną vaizdinį (jis tikrai mato kažką mėlyna), o haliucinacijos atveju –

driežo formos besiraitantį oranžinį vaizdinį (jis tikrai mato kažką besiraitančio oranžinio ir

driežiško). Tiesioginiai realistai tai neigia, nes, jų nuomone, pirmuoju atveju žmogus tiesiogiai

pagauna tik baltą sieną, antruoju – irgi tik baltą sieną, ir jokių vaizdinių, atskirų nuo baltos

sienos, pripažinti nereikia.

Tai kodėl (antruoju atveju) žmogui vis dėlto atrodo, kad jis mato ant baltos sienos kažinką

driežiškai besiraitančio oranžinio? Na, atsako tiesioginiai realistai, jeigu jo smegenys liguistai

paveiktos alkoholio ar narkotikų, tai jo sprendimo galia pažeista ir jo žodžiais negalima

pasitikėti. Jam tiesiog rodosi, ir jis kliedi! O kas darosi pirmuoju atveju, kuris daug normalesnis,

pasitaiko gerokai dažniau ir jo nesukelia jokios psichiką deformuojančios medžiagos? Čia

žmogui irgi tik rodosi, kad siena yra mėlyna – ji balta, ir nieko mėlyno joje realiai nėra, - sako

tiesioginiai realistai ir klausia: kokių čia dar reikia vaizdinių!?

Taip, sutinka idealistai ir netiesioginiai realistai: nieko mėlyno toje baltoje sienoje iš tiesų

nėra, tačiau kodėl jam vis dėlto atrodo, kad siena – mėlyna? Kuo šį faktą galima paaiškinti? Juk

nesiginčijama dėl sienos spalvos (visi sutinka, kad ji balta), tik klausiama, kokio fakto dėka

žmogui rodosi, kad ji mėlyna, o ne, pvz., balta (taip, kaip ji iš tikrųjų yra) ar žalia ar oranžinė ar

violetinė ar dar kokia kita? Teigdami, kad žmogui tik atrodo, kad siena mėlyna, tiesioginiai

realistai nieko nepaaiškina, ypač jie nepaaiškina paties fakto, jog balta siena atrodo mėlynai.

10

Tatai paaiškina tik idealistai ir netiesioginiai realistai teigdami, jog žmogus tiesiogiai mato

mėlyną vaizdinį, kurį jis arba (klaidingai) interpretuoja kaip išreiškiantį sienos spalvą, arba iš

kitų duomenų supranta, kad siena, nors atrodo mėlyna, tokia nėra. Kitaip sakant, jis iš tiesų mato

kažką mėlyna, bet tas „kažkas“ nėra siena, o tik sienos vaizdinys (jutimų duomuo). Jis mato ir

sieną, tačiau mato ją ne baltą, o mėlyną (čia ir suveikia mėlynasis vaizdinys) – taigi mato du

dalykus (nors nebūtinai suvokia juos kaip du), mato sieną ir jos mėlyną vaizdinį. Būtent šitaip

iliuzijas aiškina idealistai ir netiesioginiai realistai (apie tai, kur jie išsiskiria, kalbėsime šiek tiek

vėliau – 5-ojo šio straipsnio skyrelio antrojoje pastraipoje).

Panašiai idealistai ir netiesioginiai realistai sutaria ir dėl haliucinacijų. Iliuzijos atveju mums

rodosi tikras daiktas, bet ne toks, koks jis yra: jis rodosi turintis savybę, kurios stokoja, ar

stokojantis savybės, kurią turi. Tačiau kai patiriame haliucinaciją, tai mums rodosi visiškai

nesamas daiktas su savybėmis, kurias jis turėtų, jeigu egzistuotų. Taigi girtuoklis ar narkomanas

mato sienoje kažką kaip driežas besiraitančio oranžinio, kurio fizinėje realybėje visai nėra, o

daltonikas mato ne tą spalvą, kuri iš tiesų sienoje yra, bet kitą. Tačiau abiem atvejais jų

paklydimus paaiškina tai, jog jie realiai mato tik sąmonėje egzistuojančius vaizdinius.

Vadinasi, idealistai ir netiesioginiai realistai prikiša tiesioginiams realistams, kad pastarieji

neįstengia paaiškinti iliuzijų ir haliucinacijų. Savo disertacijoje šią mintį plėtoju ir vaizdinių

buvimą įrodinėju disertacijos skyriuose 2.2 ir 2.4.

Jeigu tokie dalykai kaip vaizdiniai (jutimų duomenys) iš tiesų egzistuoja, tai (metafiziniu

realizmu grindžiamas) tiesioginis realizmas – klaidingas, nes būtent jis teigia, kad išorinius

(materialius) daiktus pažįstame betarpiškai ir jokio jutimų duomenų tarpininkavimo nereikia. O

būtent tai, kad jų ne tik yra, bet ir reikia, įrodo tik ką minėti du argumentai – argumentas iš

haliucinacijų ir argumentas iš iliuzijų. Yra ir dar vienas, kauzalinis argumentas, nurodantis ne

tik į haliucinacijų buvimą (to jau savaime užtenka jutimų duomenų egzistavimui įrodyti), bet jį

dar papildantis tam tikru samprotavimu, iš kurio aišku, kad jutimų duomenys dalyvauja ne tik

haliucinacijose, bet ir kiekvienoje percepcijoje:

1. Teoriškai įmanoma, aktyvuojant kokį nors normalioje percepcijoje dalyvaujantį smegenų

procesą, sukelti haliucinaciją, kuri savo subjektyviu charakteriu būtų į tą percepciją visiškai

panaši.

2. Jei haliucinacija ir percepcija turi tą pačią neuralinę priežastį, reikia jas ir panašiai aiškinti.

Nebūtų pagrįsta sakyti, kad haliucinacijoje dalyvauja vaizdinys (jutimų duomuo), o percepcijoje

jis nedalyvauja, jei abi turi tą pačią betarpišką priežastį.

Vadinasi,

3. Visi percepcijos procesai smegenyse sukuria tam tikrą sąmonės objektą, kurio negalima

sutapatinti su kokiu nors išorinio pasaulio reiškiniu, t.y. jie sukuria vaizdinį (jutimų duomenį).

11

Šitie trys argumentai puikiai žinomi, disertacijos skyriuose 6.3.2., 6.3.3., 6.3.4. ir 6.3.5. juos

smulkiau aptariu ir paaiškinu, kodėl juos reikia priimti. Neblogai žinomas, tačiau rečiau

naudojamas dar vienas argumentas už jutimų duomenų buvimą, o būtent jis ir sudaro mano

disertacijos ašį.

4. Laiko atotrūkio argumentas

Šis argumentas, pirmą kartą išdėstytas G. W. Leibnizo, kai tik literatūroje minimas, tai

dažniausiai kritiškai. Aš tą kritiką nuodugniai apsvarstau ir prieinu prie išvados, kad ji visa

neteisinga ar bent nepakankamai pagrįsta. Bet yra kelios šio argumento versijos ir aš

sukonstruoju tokią, kurią būtų galėjęs panaudoti G. Berkeley‘jus, jeigu būtų manęs, kad jam jos

reikia. Tačiau jis, matyt, buvo patenkintas kitais argumentais, kuriais manė galįs pakankamai

pagrįsti savo įsitikinimą, jog materialusis pasaulis susideda iš vaizdinių ir jog (tik truputį

pakoreguojant A. Schopenhauerio formuluotę) „viskas, kas yra ištįsę erdvėje ir ją pripildo, taigi

apskritai regimasis pasaulis, savo būtį kaip pačią savaime gali turėti absoliučiai tik Dievo

sąmonėje ir mūsų vaizdinyje ir kad yra absurdiška, net prieštaringa priskirti jam dar būtį už

vaizdinio ribų ir nepriklausomą nuo pažįstančiojo subjekto ir tuo remiantis pripažinti savaime

egzistuojančią materiją“.

Vis dėlto, ypač nuo XX a. pradžios, vis daugiau filosofų pradėjo ne tik abejoti G.

Berkeley‘jaus teiginiais bei argumentais (kurių daugumos aš smulkiau nenagrinėju), bet taip pat

ir teigti, kad tokių dalykų kaip vaizdinių (jutimų duomenų) stačiai nėra. Tuo tikslu, daugiausia

gindami tiesioginį realizmą, jie kritikavo (mano minėtus) argumentus iš iliuzijų ir haliucinacijų

bei kauzalinį argumentą, kuriuos aš užstoju. Tačiau pagrindinė mano disertacijos tezė yra ta, kad

(1) laiko atotrūkio argumentas yra gerokai stipresnis negu daugumai jį svarsčiusių filosofų

atrodo ir kad (2) jis įrodo vaizdinių buvimą bei tuo pačiu visiškai ar didele dalimi įrodo

tiesioginio realizmo klaidingumą. O likusia dalimi (jei tokia išvis lieka) tą klaidingumą įrodo

kiti, tik ką minėti argumentai.

Savo disertacijos originalumą įžvelgiu ne tik tame, kad tai, kiek žinau, yra vienintelis tokios

apimties ir pobūdžio darbas, ištisai skirtas šiam Leibnizo ir Bertrando Russello vertintam

argumentui ir jau šimtmetį besitęsiančioms jo sukeltoms diskusijoms apžvelgti, bet ir tame, kad

šiuo argumentu ginu būtent jungtinę epistemologinio ir metafizinio idealizmo bei

epistemologinio realizmo poziciją, kuriai filosofijos istorijoje atstovavo tik Berkeley‘jus, o

dabartyje – tik neseniai miręs Johnas Fosteris ir gal dar pora kitų, įskaitant Howardą Robinsoną.

Aš taip pat bene vienintelis imuosi užduoties bandyti įrodyti, jog šitokiai pozicijai apginti

12

berkeley‘jiškasis laiko atotrūkio argumentas (jį išdėstau disertacijos skyriuje 1.4.) tinka

geriausiai.

Štai jis (išdėstytas naudojantis G. Berkeley‘jaus terminologija):

(1) Tarkime, kad žiūriu į tai, ką laikau esant Saule.

(2) Kadangi Saulės šviesos spinduliams, anot fizikų, trunka apie aštuonias minutes, kad mane

pasiektų, logiškai įmanoma, jog per tą laiką Saulė būtų sunaikinta ir jau nebeegzistuotų tada,

kada jos šviesa pasiekia mano akis.

(3) Tarkime, kad Saulė iš tiesų per tą laiką sunaikinta.

(4) Tai, ko jau nebėra, nebegali būti kažkas, ką aš tiesiogiai matau.

(5) Tačiau vis dėlto aš tiesiogiai (betarpiškai) matau kažką, ką laikau esant Saule.

Vadinasi,

(6) Tai, ką tiesiogiai matau, yra regimoji (akivaizdi) Saulė, ne fizinė Saulė, kurią tariame buvus

sunaikinta.

(7) Tačiau tiesiogiai matau tik vieną regimąją (akivaizdžią) Saulę ir nieko kito, kas „ją atitiktų ar

būtų panaši į ją“ – tik vieną spindintį apskritą šviesulį danguje, bet ne du.

(8) O be šitos tiesiogiai regimos Saulės nei tiesiogiai, nei netiesiogiai nematau jokio kito į ją

panašaus daikto.

(9) Jei netiesioginis (kitaip sakant, atstovaujamasis) realizmas teisingas, tai galiu matyti tai, kas

jau nebeegzistuoja, ir matau ne tik regimąją (akivaizdžią) Saulę, bet ir dabar tariamai sunaikintą

fizinę Saulę.

Vadinasi,

(10) Netiesioginis (kitaip sakant, atstovaujamasis) realizmas – neteisingas.

(11) Nors šios regimosios Saulės antrine priežastimi gal ir yra fizikos Saulė, jos, kaip ir visų

daiktų, pirminė ir tikroji priežastis yra Dievas.

(12) Tačiau regimoji Saulė yra šviesos, spalvos ir formos jutimų derinys arba iš šių jutimų

sukonstruotas „mūsų pačių smegenų kūrinys“.

(13) Šis derinys arba kūrinys egzistuoja arba už sąmonės ribų, arba sąmonėje.

(14) Jeigu jis egzistuoja už sąmonės ribų, jis galėtų radikaliai pasikeisti, man to nepastebėjus.

(15) Tačiau šitas šviesos, spalvos ir formos jutimų derinys arba iš šių jutimų sukonstruotas

„mūsų pačių smegenų kūrinys“ negali radikaliai pasikeisti, man to nepastebėjus.

Vadinasi,

(16) Mano matoma regimoji Saulė egzistuoja mano sąmonėje.

(17) Bet net jeigu fizikos Saulė nebūtų buvus sunaikinta (o tikrovėje ji niekada nebuvo – ir

žmonijos istorijoje tikriausiai niekada ir nebus – sunaikinta, vis tiek aš tiesiogiai matyčiau tik

regimąją Saulę – dėl 7-ajame teiginyje paminėtos priežasties.

13

(18) Tarkime, kad fizikos Saulė sunaikinta nebuvo, t. y. kad (3) yra neteisinga prielaida.

(19) Tuomet tuo pačiu aktu, kuriuo matau regimąją Saulę, aš taip pat tiesiogiai ar netiesiogiai

matau nesunaikintą fizikos Saulę, kuri yra dar didesnis šviesos, spalvos ir formos jutimų derinys

ar dar sudėtingesnis iš šių jutimų sukonstruotas „mūsų pačių smegenų kūrinys“ – ir tai toks

derinys ar kūrinys, kurio vieną dalį sudaro regimoji Saulė.

Vadinasi,

(20) Kai tik matau regimąją Saulę, matau ir didesnį jutimų, kurių bent kažkiek matau tiesiogiai,

derinį.

(21) Bet tai, kas galioja regimosios Saulės matymui, galioja atitinkamomis aplinkybėmis ir

kiekvieno kito regimo daikto atveju.

Vadinasi,

(22) Visi regimi daiktai yra jutimai arba jutimų deriniai, egzistuojantys tik sąmonėje.

Paskutinė išvada ir išreiškia berkeley‘jiškąjį idealizmą. Kadangi argumentas logiškai

sandarus, belieka tik įvertinti jo premisų teisingumą. Tačiau tai per didelis uždavinys vienam net

ir disertacijos ilgio darbui, todėl nuodugniai pagrįsiu tik porą, literatūroje gausiausiai priekaištų

sulaukusių, premisų, būtent, 4-ąją ir 12-ąją.

Daugiausia dėmesio skirsiu 4-osios premisos (vadinkime ją matymo įvykio ir matomo objekto

vienalaikiškumo teze) pagrindimui, nes ji dažniausiai puolama. Ją apgynus, pirmoji išvada

(išreikšta 6-ąja premisa) tvirtai įrodyta, nes premisas (1) – (4), atrodo, galima laikyti

neginčytinomis. Ši išvada teigia, jog hipotetiniu sunaikintos Saulės atveju matome „regimąją

Saulę“, o tai ir yra, ką Berkeley‘jus vadina „idėja“, o mes „vaizdiniu“. Jei ši išvada teisinga, tai

vaizdinių buvimas įrodytas – bent vienu, „regimosios Saulės“, atveju. Tam, kad įrodytume, jog

vaizdinius matome ir visais kitais, ne vien „regimosios Saulės“, atvejais, reikia papildomų

argumentų, kuruos pateikiu, remdamasis daugiausia Lenu Carrier‘iu, savo disertacijos skyriuje

4.9.

Antroji išvada (išreikšta 9-ąja premisa), atmetanti netiesioginį realizmą, remiasi empiriniu

pastebėjimu, kad nieko, išskyrus vieną į Saulę panašų daiktą, mes ir nematome, o pagal

netiesioginį realizmą turėtume matyti bent du (vieną betarpiškai, antrą netiesiogiai). Gali

atrodyti, kad su netiesioginiu realizmu šis argumentas apsidirba per lengvai, tačiau tuo, kad

netiesioginis realizmas visiškai neįtikinamas, neabejoja ir Davidas Smithas: jam atmesti jis

skiria labai trumpą argumentą, bet Johnas Fosteris su juo dorojasi kur kas išsamiau; abiejų, ypač

pastarojo, argumentus apžvelgiu skyriuose 5.3. ir 3.5.4. Vis dėlto netiesioginiam realizmui

pritaria Brianas O‘Shaughnessy‘is (3.4.), tačiau aš bandau įrodyti, kad jo argumentams

natūraliau būtų išsiskleisti idealizmu (3.5.3.). Morelandas Perkinsas taip pat gina netiesioginį

realizmą (3.5.2.). Be to, jis, kaip ir B. O‘Shaughnessy‘is, ryškina skirtumą tarp tiesioginės ir

14

netiesioginės pagavos (3.5.5.). Tą skirtumą taip pat ryškina, remdamasis Berkeley‘jumi,

James‘as Cornmanas ir kiek kitaip B. Russellas (3.5.1) bei J. Fosteris (3.5.4.). Būdai skirti

tiesioginę nuo netiesioginės pagavos gali būti keli ir vienodai pagrįsti (3.5.5.)

5. Ar galima matyti tik tai, kas egzistuoja dabar?

Grįžkime prie 4-osios premisos. Labai daug filosofų jai priešinasi. Jiems atrodo, kad galime

betarpiškai matyti ne tik tai, kas dabar dedasi, bet ir tai, kas praeityje įvyko – ne tik dangaus

kūnus ir pokyčius juose, bet ir tai, kas tik ką įvyko (bet daugiau nebevyksta) panosėje. Kai kurie

iš tų filosofų netgi teigia, kad viskas, ką dabar regime, yra grynai praeityje. Jų visų argumentus

išdėstau ir kritikuoju antrame ir trečiame disertacijos skyriuose, labiausiai nusitaikydamas į

tokius tiesioginius realistus kaip F. Dretske, A. Quintonas, D. M. Armstrongas, A. J. Ayeris, R.

Chisholmas, G. Pitcheris, J. Bengsonas, K. Waltonas. Anot jų, jau vien dėl mokslinio fakto, kad

daiktus matome tik tada, kai mūsų organizmus paveikia daiktų išleidžiama ar atspindima ir

baigtiniu greičiu keliaujanti šviesa, tas matymas įvyksta vėliau nei toji šviesa palieka matomąjį

daiktą. Todėl, esą, ir matome daiktą ne tokį, koks jis dabar (matymo akimirką) yra, o tokį, koks

jis buvo, kai išleido ar atspindėjo šviesą.

Netiesioginiai realistai, priimdami priešpaskutiniame sakinyje minėtą faktą, paskutinį sakinį

perrašo taip: „Todėl tiesiogiai matome ne daiktą, bet jam atstovaujančius vaizdinius, o

netiesiogiai išties matome daiktą ne tokį, koks jis dabar (matymo akimirką) yra, bet tokį, koks

jis buvo, kai išleido ar atspindėjo šviesą“. Idealistai, taip pat priimdami minėtąjį mokslinį faktą,

tą anos pastraipos paskutinį sakinį perrašo taip: „Todėl tiesiogiai matome vaizdinius, iš kurių, jei

daiktas matymo akimirką egzistuoja, jis ir sunkonstruotas ir taip pat jį tiesiogiai matome, o jei

nebeegistuoja, tai buvo sunkonstruotas ir jo dabar nei tiesiogiai, nei netiesiogiai nematome, bet

iš dabar tiesiogiai matomų vaizdinių bei kitų faktų galime daryti pagrįstas išvadas apie jo

buvusias savybes“.

Tačiau ne visi tiesioginiai realistai 4-ąją premisą neigia. Davidas Lewisas jai pritaria todėl,

kad, jo įsitikinimu, ne(be)samų daiktų matyti negalima, o ir žvaigždės, tol, kol jas matome, nėra

tiesiogiai praeityje, nes, jo nuomone, „šviesos laiduojama sąsaja turi tokią pat teisę, kaip ir

vienalaikiškumas tam tikroje atskaitos sistemoje, būti jau nebegaliojančios absoliutaus

vienalaikiškumo sąvokos teisėtu įpėdiniu“. Taigi jis teigia, jog visos mūsų matomos žvaigždės

egzistuoja tuo metu, kai yra matomos; tatai teigia ir Hanochas Ben-Yamis, plėtodamas „regimo

vienalakiškumo“ sampratą (6.1). Todėl jiedu neigia 2-ąją laiko atotrūkio argumento premisą.

Tada, žinoma, šis argumentas už vaizdinių buvimą subliūška.

15

Nesu pakankamai kompetentingas įvertinti D. Lewiso ir H. Ben-Yamio poziciją remiančius

fizikos argumentus, tačiau dalijuosi J. Bengsono nuomone, jog laikyti tai, kas įvyko prieš,

tarkim, 700 metų, dabarties įvykiu, grubiai „pažeidžia mūsų turimą dabarties sąvoką“.

Bet net jeigu nebūtų laiko atotrūkio tarp akimirkos, kada šviesa palieka žvaigždę, ir

akimirkos, kada ji pasiekia mūsų akis, yra kitas laiko plyšys tarp akimirkos, kada šviesa pasiekia

mūsų akis, ir akimirkos, kada mes žvaigždę pamatome. Juk nervų sistemai reikia mažytės dalies

sekundės, kad perkeltų signalą iš tinklainės į smegenis, ir per tą laiką žvaigždė arba bet koks

kitas objektas gali spėti išnykti. Tuo atveju, rodos, vėl iškyla galimybė laiko atotrūkio

argumentą panaudoti.

Tokiais trumpais laiko plyšiais iš tiesų rėmėsi, išdėstydamas savąjį laiko atotrūkio

argumentą, G. W. Leibnizas. Pavyzdžiu imdamas žiūrėjimą ne į Saulę, o į čia pat ant sienos

kabantį paveikslą, jis teigė, kad, kol šviesos spinduliai nuo paveikslo keliauja į mano akis,

paveikslas gali išnykti, todėl, jei taip įvyko ir kadangi tam tikru laiku galiu matyti tik tai, kas tuo

laiku egzistuoja, tai tuo laiku ir matau jau nebe paveikslą, o tik jo vaizdinį. Tačiau XX-ajame

amžiuje kai kurie filosofai (pvz., H. Robinsonas) šį argumentą atmetė – ne todėl, kad jie atmestų

matymo įvykio ir matomo objekto vienalaikiškumo tezę („tam tikru laiku galiu matyti tik tai, kas

tuo laiku egzistuoja“), bet todėl, kad šios tezės teisingumui, jų manymu, neturi reikšmės laiko

skirtumai, kurie per maži, kad žmogaus akis (ir sąmonė) juos pastebėtų. Kai tezėje kalbame

apie laiko atžvilgiu sutapti turinčius matymo įvykio laiką ir matomo objekto laiką, omeny

turime žmogaus juntamą, sąmonėje pagaunamą laiką, o ne tą, kurį įmanoma tik moksliškai

nustatyti.

Užtat tokiu atveju (kai laiko tarpas, nors jis fiziškai realus, tačiau jo neįmanoma žmogiškai

užčiuopti) laiko atotrūkio argumentas netenka pagrindo, nuo kurio jį būtų galima paleisti į darbą.

Todėl siekiant įrodyti, kad ir šiuo atveju matome ne fizinį daiktą, o jo vaizdinį, tenka kliautis

argumentais iš haliucinacijų ar iš iliuzijų ar kauzaliniu argumentu, kurių kiekvienas įrodo, kad

visais, taigi ir trumpo laiko atotrūkio, atvejais, pirmiausia matome vaizdinius.

Tačiau D. Lewisas ir H. Ben-Yamis beveik tikrai klysta, ir laiko atotrūkis tarp akimirkos, kai

šviesa palieka žvaigždę, ir akimirkos, kai ji pasiekia mūsų akis, yra pakankamai didelis, jog

laiko atotrūkio argumento visiškai pakaktų vaizdinių buvimui įrodyti.

4-ajai premisai pritaria ir kitas tiesioginis realistas L. Carrieris, bet ne todėl, kad jis, kaip tik

ką minėtieji du, manytų, jog dėl šviesos greičio tarp žvaigždės ir jos matymo atsiranda „regimas

vienalaikiškumas“, o todėl, kad, jei tam tikru laiku žvaigždė iš tiesų užgęsusi, tai tuo laiku

matome ne pačią (jau praeityje atsidūrusią) žvaigždę, o jos išleidžiamą šviesą, nors ir atrodytų,

kad matome žvaigždę. Bet šviesa, kaip ir pati žvaigždė, yra, jo nuomone, fizinis objektas, ne

vaizdinys, tad laiko atotrūkio argumentas įrodo ne tai, kad visada matome vaizdinius, o tai, kad

16

visuomet matome šviesą, bet tais atvejais, kai šviesą išleidęs fizinis objektas dar egzistuoja, tai

matome ir jį. Šiuo požiūriu dalijasi ir B. O‘Shaughnessy‘is (žr. 4.7., 4.8., 4.10., 4.12.).

Jie abu pritaria tezei, kad tam tikru konkrečiu laiku galime matyti tik tai, kas tuo laiku

egzistuoja, o ne anksčiau egzistavo ar tuo labiau vėliau egzistuos. Aš pats skyriuje 4.14. įrodau,

kad ši tezė yra būtina, loginė tiesa, kurios pagrindinis ramstis yra įžvalga, kad jeigu daiktas yra

matomas, tai turi būti jei ne fiziškai, tai bent logiškai įmanoma, kad matantysis galėtų tuo daiktu

manipuliuoti, jį paveikti, pakeisti ir t. t. Bet šitaip daryti įtaką galima tik dabar esančiam, o ne

praeityje buvusiam, daiktui. Todėl jei daiktas dabar matomas, tai jis turi ir dabar egzistuoti. Šią

įžvalgą pasiekti man padėjo diskusijos su J. J. Valbergu (žr. skyrius 4.3. ir 4.4.).

Beje, tezę, jog matome tik tai, kas dabar egzistuoja, paremia ir kai kurie tyrimus

atlikinėjantys neurobiologai. Tai aptariu skyriuje 6.4.

Lygiai taip pat ir tai, ką girdime dabar, egzistuoja dabar. Tai teigė A. O. Lovejoy‘jus ir

daugelis kitų filosofų, nors neigė G. Pitcheris, tačiau jo argumentus galima įtikinamai

sukritikuoti, tai ir darai skyriuje 3.2. Žodžiu, ir garsams galioja principas, kad esse est audiri.

Bet ar tikrai žvaigždės ar bet kokio kito daikto skleidžiama ar atspindima šviesa yra fizinis

objektas? Skyriuose 4.8., 4.10., 4.12 ir 6.5. argumentuoju, kad nors fotonai yra fizinės dalelės ir

gali būti tapatinamos su šviesa, kita prasme šviesa yra pojūtis arba jutimas, kurį fotonai

sąmonėje sukelia, o šis šviesos jutimas ir yra vaizdinys. Taigi, jeigu, pasak L. Carrierio ir B.

O‘Shaughnessy‘io, kiekvieną kartą, kai kokį nors daiktą matome, matome šviesą, tai visada

matome ir vaizdinį.

O kaip regimi vaizdiniai sudaro regimą fizinį objektą? Atsakyti į šį klausimą – reiškia

pagrįsti 12-ąją premisą. Tai didele dalimi darau, atmesdamas pagrindinius konkurentus –

netiesioginį neidealistinį realizmą (skyriuose 5.3. ir 3.5.4.) ir, žinoma, tiesioginį neidealistinį

realizmą (jam atmesti skiriama visa disertacija). Tačiau kaip konkrečiai fiziniai daiktai

konstruojami iš vaizdinių – šį klausimą, pateikęs kelias alternatyvas, palieku atvirą (skyriuje

6.5.). Tai uždavinys ateities tyrinėjimams. Jie būtini, siekiant galutinai įrodyti idealizmą.

6. Kitos problemos

Disertacijoje daugiausia dėmesio skiriu matymo objektams. Tačiau galima kelti ir klausimą, kas

per dalykas yra pats matymas. Dažniausiai matymas yra laikomas aktu arba būsena, kuriuo arba

kuria tam tikras subjektas (žmogus ar gyvūnas) siejamas su tam tikru objektu (vaizdiniu arba

daiktu). Pvz., sakinį Jonas mato Jonę galima suvokti kaip teigiantį Jono aktą Jonės atžvilgiu

pagal modelį Jonas vizualiai pagauna Jonę arba tą sakinį galima suvokti kaip teigiantį Jono

būseną Jonės atžvilgiu pagal modelį Jonas yra vizualinio Jonės pagavimo būsenoje. Yra ir trečia

17

galimybė. Galima matymą laikyti grynu santykiu tarp matančiojo subjekto ir matomojo daikto,

pagal modelį Joną ir Jonę sieja matymo (pirmasis mato antrąją) santykis. Šitie trys modeliai

subtiliai skiriasi. Mano disertacijos išvados suderinamos su visais trimis.

Bet yra dar dvi teorijos, bandančios nusakyti matymo loginę esmę. Pagal vieną jų, matyti

daiktą – tai reprezentuoti (specifine ir nelabai aiškia šio žodžio prasme) jo savybes. Tai šiuo

metu labai madinga reprezentacinė teorija. Kuo ir kiek ji skiriasi nuo atstovaujamojo (kitaip

sakant, netiesioginio) realizmo – kontroversiškas klausimas. Šį klausimą svarstau ir apie pačią

reprezentacinę teoriją neigiamai atsiliepiu skyriuose 5.2, 5.3. ir 5.4. Tačiau skyriuje 5.6. prieinu

prie išvados, kad reprezentacinė teorija, nors ir klaidinga, bet ji netrukdo berkeley‘iškajam laiko

atotrūkio argumentui.

Visai kitaip su dar viena teorija. Tau vadinamoji prieveiksminė (adverbialinė) teorija, kurią

lietuviškai išdėstyti lengviau negu angliškai. Grįžkime prie ankstesnio iliuzijos pavyzdžio, kai

žiūrime į baltą sieną ir mums atrodo, kad ji mėlyna. Daug kas (įskaitant mane) prieiname prie

išvados, kad čia matome kažką mėlyna – būtent, mėlyną vaizdinį. O prieveiksmininkai

(adverbialistai) sako – nieko panašaus: matome baltą sieną, tik ji atrodo mėlynai. Taip pats tai

išreiškiau, kai šią iliuziją (pirmusyk) išdėsčiau. Prieveiksmininkai nori išvengti vaizdinių, jų

vietoje kalbėdami apie matymo būdus, išreikštus prieveiksmiais.

Deja, ši teorija neatlaiko kritikos. Tiesa yra ta, kad baltą sieną matome ne „mėlynai“ –

tiesiog matome ją mėlyną, matome mėlyną objektą, ir tas objektas yra vaizdinys, nors iliuzinis.

Kad prieveiksminė teorija klaidinga, pagrindžiu skyriuje 5.5.

7. Baigiamosios pastabos

Berkeley‘iškąjį idealizmą gerokai sunkiau apginti anglo-amerikietiškosios analitinės filosofijos

negu lietuviškosios filosofinės tradicijos kontekste. Tai dar viena priežastis, kodėl, siekdamas

iššūkio, šį darbą rašiau ir ginuosi angliškai. Amerikoje ir šiaip anglakalbiame pasaulyje šiandien

mažai kas traktuoja idealizmą rimtai (bet tai nereiškia, kad niekas taip nedaro, tik tokius

filosofus galima ant vienos rankos pirštų suskaičiuoti, kelis iš jų suminėjau ir dviem savo

disertacijoje ypač rėmiausi), o Europoje (tuo pačiu ir lietuviškojoje tradicijoje) – padėtis visai

skirtinga. Nebūtina čia minėti kadaise dažnai ir įkyriai linksniuotą, o dabar lyg susitarus beveik

visuotinai pamirštą Leniną, laikiusį Berkeley‘jų ir jo įtakoje buvusius filosofus pagrindiniais

savo priešininkais ir jų kritikai paskyrusį visą savo Materializmą ir empiriokriticizmą. Užtenka

tik pažymėti, jog Berkeley‘jų gerbė ir pozityviai vertino Edmundas Husserlis, kurio

fenomenologijos tradicija Lietuvoje tikrai gyva, jei ne dominuojanti. Husserlis ne tik

Berkeley‘jų vertino, bet ir buvo savo įdėjomis gerokai į jį panašus. Anot Hermano Philipse‘o,

18

„kaip ir Husserlis, Berkeley‘jus manė, kad materialaus pasaulio, egzistuojančio nepriklausomai

nuo sąmonės, sąvoka yra absurdiška. 55-ame savo Idėjų I skyriuje Husserlis neigė esąs

berkeley‘iškas idealistas dėl to, kad jis neneigia pasaulio buvimo. Tačiau šis bandymas

nesusisieti su nepopuliariuoju vyskupu nevykęs, nes remiasi Berkeley‘jaus imaterializmo

nesupratimu. Kaip ir Husserlis, Berkeley‘jus teigia tik pašalinantis absurdišką pasaulio

egzistavimo interpretaciją, bet neneigiantis paties jo buvimo. Kaip Husserlis, Berkeley‘jus yra

redukcinis, bet ne nihilistinis idealistas. Ir kaip Husserlis, jis teigia, jog materialų pasaulį

sukonstruoja protas iš jam imanentiškų jutimų ar idėjų“8.

Beje, tai, kad ne tik Husserlis, bet ir Maurice‘as Merleau-Ponty buvo idealistai, teigia ir mano

disertacijoje minimas Davidas Smithas, parašęs anglo-amerikiečiams analitikams tinkamą įvadą

į Husserlio filosofiją. O kad lietuviškojoje tradicijoje idealizmas gyvas, aišku iš to, kiek laiko ir

energijos jam aptarti bei kritikuoti skiria V. Sezemanas. Skirtingai nei pastarasis, idealizmą

paniekinančiai traktuoja P. Kuraitis, o anksčiau – Angelas Daugirdas. Tačiau savo disertacijoje

aš šių ir kitų lietuviškų šaltinių neminiu – ne tik todėl, kad jie Vakarų pasaulyje mažai žinomi,

bet pirmiausia todėl, kad tai, ką iš jų galima išmokti, nublanksta prieš tai, ką idealizmo klausimu

galima pasisemti iš anglakalbės literatūros. Kaip tik todėl, manau, ir svarbu, jog tai, ką apie

idealizmą neigiamo ar teigiamo sako anglakalbiai filosofai, ypač analitikai (kurių požiūriais jau

domimasi ne tik visur kitur, bet ir žemyninėje Europoje, nors mažiausia – mūsų šalyje), būtų

išgirsta ir Lietuvoje.

8 žr. Herman Philipse, “Transcendental Idealism”. In: The Cambridge Companion to Husserl, 2011.

19

Abstract

The Time-Gap Argument (TGA), apparently first used by Leibniz and here examined

systematically and historically, is an attempt to show that whenever we see anything, there is

something we see that is mental, i.e., a sense-datum. This conclusion is derived from four

empirical premises and one held to be a conceptual truth. The empirical premises are that (a) in

order for a physical object to be seen, it is causally necessary for this object to have emitted or

reflected light that reaches our eyes; that (b) light travels at a finite speed; that (c) by the time

light emitted by it reaches our eyes, a certain star has already ceased to exist; and that (d) at that

time we nonetheless see something star-like at or near the place where there used to be a star.

The alleged conceptual truth is that (e) in order for any object or event to be seen at time t, it

must exist at t. The argument from (a), (b), (c), (d), and (e) to the just-mentioned conclusion still

needs a further premise asserting a continuity between the case of the star and any other cases of

possible perception. Since this premise as well as (a), (b), (c), and (d) can be shown to be well-

supported and since the argument from them in conjunction with the disputed (e) to the

conclusion is valid, critics of the TGA have attacked (e) with the aim of showing that the

argument is unsound and that the conclusion is false.

In this dissertation I defend the TGA and its conclusion by arguing that (e) is indeed a

conceptual truth and that these attacks on it, waged throughout the last hundred years, are one

and all unsuccessful. They involve one or more of the following claims: that present perception

only appears to, but does not really, require the present existence of the object perceived; that

there is no radical break between perception and memory; that we can directly “see into the

past”; that we always do “see only into the past,” if we see anything at all; that if we see

anything it can only be physical objects or events; that any light we see is physical; that there are

no good reasons to posit mental sense-data; and that no mental sense-data in fact exist. I attempt

to show that all these claims are false or inconclusively supported.

In the course of doing so, I also argue against the Representational and Adverbial Theories of

Perception insofar as these deny the existence of sense-data. Finally, since the conclusion of the

TGA (as stated above) leaves open the possibility of our also seeing physical things in virtue of

our seeing sense-data, I argue for an Idealist (Phenomenalist) rather than a Representative

Realist construal of the relation between our seeing sense-data and our seeing physical objects.

This construal allows me to recognize and defend a Direct Realist version of Idealism, according

to which what we directly perceive are presently existing sensory records of physical events,

which events, even if they are in the past and are not themselves perceived, were or are logically

constructed out of such records. This kind of Idealism is superior to a Representative Realism

20

that claims we directly perceive presently existing records of past events that are ontologically

entirely distinct from these events; and both are superior to a Non-Idealist Direct Realism which

claims we perceive past events directly.

21

Contents

Preface

Chapter One: Introduction

1.1. Aiming towards Idealism

1.2. The Crux of the Time-Gap Argument: Must a Thing Exist When

It Is Perceived?

1.3. Other Objections to the TGA

1.4. A Berkeleian TGA

1.5. The Importance of the TGA

Chapter Two: Denying the Simultaneity-Requirement (I)

2.1. The Existence-Requirement

2.2. Dretske on the TGA and the Simultaneity-Requirement

2.3. An Excursus: Physical vs. Phenomenological Simultaneity

2.4. Dretske and Direct Awareness (I)

2.5. Dretske and Direct Awareness (II)

2.6. Quinton on the TGA

2.7. Armstrong on the TGA

2.8. Conclusion about Dretske, Quinton, and Armstrong

Chapter Three: Denying the Simultaneity-Requirement (II)

3.1. Can Our Eyes Range into the Past (Only)?

3.2. The TGA and the Velocity of Sound: Pitcher et al.

3.3. Perception, photography, television, and memory (I)

3.4. Perception, photography, television, and memory (II)

3.5. Direct vs. Indirect Perception

3.5.1. Cornman, Russell, Berkeley

3.5.2. Perkins

3.5.3. O’Shaughnessy

3.5.4. Foster

3.5.5. Conclusion about Direct vs. Indirect Perception

3.6. Conclusion about Seeing into the Past

Chapter Four: Affirming the Simultaneity-Requirement

4.1. Cornman and the TGA

4.2. Dancy’s Objection: ‘Presence’ Is Ambiguous

4.3. Valberg on Presence

4.4. The Puzzle of Experience

22

4.5. Berkeley on Perceivings and Perceiveds

4.6. Berkeley and the Simultaneity-Requirement

4.7. O’Shaughnessy and the Simultaneity-Requirement

4.8. Carrier et al. on Seeing Light and the Simultaneity-

Requirement (I)

4.9. The Continuity Argument

4.10. Carrier et al. on Seeing Light and the Simultaneity-

Requirement (II)

4.11. Seeing distant objects vs. seeing objects close-by: Mandelbaum and Noë

4.12. Carrier et al. on Seeing Light and the Simultaneity-Requirement (III)

4.13. Presentism vs. Eternalism

4.14. Conclusion about the Simultaneity-Requirement

Chapter Five: The Representational and Adverbial Theories

5.1. Dretske and the Sensory Core

5.2. Travis and Alston on Representationalism

5.3. Jackson on Representationalism

5.4. Valberg and Noë on Representationalism

5.5. The Adverbial Theory of Perception

5.6. Conclusion about the Representational and Adverbial

Theories

Chapter Six: What We See

6.1. Apparent Simultaneity

6.2. Intra-Organismic Time-Gaps

6.3. Other Arguments for Sense-Data

6.3.1. The Simplest Arguments

6.3.2. Valberg on the Argument from Illusion

6.3.3. Smith on the Argument from Illusion

6.3.4. Robinson’s Response to Smith on the Argument from Illusion

6.3.5. The Causal Argument

6.4. Seeing the Present

6.5. Conclusion about What We See

23

Preface

My short answer to the title question of this work is: Yes and No. The Time-Gap Argument does

refute (or at least it goes a long way toward refuting) Direct Realism if the latter is construed, as

it standardly is, non-idealistically. It doesn’t refute it—on the contrary, it supports it—if Direct

Realism is understood idealistically. My long answer is set forth and defended on the following

pages.

This dissertation is the result of work I started and should have finished decades ago. Soon

after catching the bug of philosophy in the 1960s I read some passages which have held me in

their grip ever since. Bertrand Russell had written in his Human Knowledge: Its Scope and

Limits (1948) that “though you see the sun now, the physical object to be inferred from your

seeing existed eight minutes ago; if, in the intervening minutes, the sun had gone out, you would

still be exactly seeing what you are seeing. We cannot therefore identify the physical sun with

what we see; nevertheless what we see is our chief reason for believing in the physical sun.”

In a similar vein A. J. Ayer wrote in The Problem of Knowledge (1956) that an “instance

which Russell often cites is that of the sun which we only see as it was eight minutes before;

when it comes to remote stars the difference may amount to thousands of years. It may even

happen that by the time we see it the star has ceased to exist. But if the star no longer exists, we

cannot, so it is argued, now be seeing it; and since in every case in which the light has had an

appreciable distance to travel it is possible that the object which we are seeing has gone out of

existence in the interval, we cannot ever identify it with what we see: for our present experience

will be the same, whether the object still exists or not.”

When I was ready to begin writing my doctoral dissertation at M.I.T. in 1971, I chose the

Time-Gap Argument as my topic, with the immediate approval of my thesis advisors Baruch A.

Brody and James F. Thomson. At that time I didn’t yet know what conclusions I’d come to; for

a while I even flirted with radical positions such as that only sense-data but no physical objects

exist and that even if the latter existed we would never see them. Judith J. Thomson firmly

dissuaded me from entertaining such extreme views; but even though I never rejected her advice

(and found myself increasingly reluctant to contradict certain fundamental commonsense claims

as enunciated by G. E. Moore) I couldn’t for the life of me see my way clearly to any

satisfactory treatment of the argument. As a result of this philosophical perplexity, and the

pressure as well as the lure of competing personal interests, I never completed my dissertation at

M.I.T.

But the bug stayed with me; and although I had now entered the journalism profession, in my

spare time I continued dabbling in philosophy. Since the early 2000s the perceptual time-gap

24

problem was the main topic of electronic communication with Fred Dretske, John Foster,

Howard Robinson, David Smith, and Jerry Valberg. Concurrently I pursued two even older

interests. One, emerging in secondary school, was Christian philosophy, especially that of an

existentialist bent, together with philosophy of religion in general. The other interest, first

aroused at the University of Chicago in the mid-1960s, centered around George Berkeley; much

later this led not only to new collegial friendships—with Timo Airaksinen, Bertil Belfrage, and

Tom Stoneham, among others—but also to invitations to read papers at International Berkeley

Society conferences in Helsinki and Karlsruhe. It was my thinking about how George Berkeley

would have responded to the Time-Gap Argument that finally gave me the idea of how to

continue and complete my dissertation.

Luckily, this topic is not yet exhausted: it wasn’t a dead horse when I started my dissertation,

and it isn’t one now. To be sure, much of the literature on the Time-Gap Argument dates from

the early and middle decades of the last century, but significant papers have been written on it in

the last two decades as well. There still is virtually no unanimity in the approaches taken. More

importantly, very little of this most recent or even earlier work defends the point of view I do,

and none of it rules it out with arguments that would make my endeavor silly. (I now, but didn’t

always, share the attitude expressed by Brand Blanshard when, in his Autobiography, he

avowed: “I have never been able to accept the realist view that the objects of direct experience

are independent of consciousness. Indeed everything we sense or feel seems to me to exist only

in consciousness.”) Thus my text is, I hope, neither otiose nor antiquated.

I am grateful to Vytautas Magnus University, where I now work in the Public

Communications Department and the Lithuanian Emigration Studies Center, for letting me

defend this dissertation as an external student. I wrote most of it in the last decade in Prague and

Kaunas in the severely limited time left over from my teaching, research, broadcasting, and

translating duties. If accepted, it will most likely be one of the very few theses in analytic

philosophy to have been defended in English at a Lithuanian university.

My arguments and views owe much to reading the works of, and/or to corresponding with,

the people mentioned above (Foster, Robinson, and Valberg, most of all). I especially thank

Professor Gintautas Mažeikis for consenting to be my Consultant and Professors Timo

Airaksinen and Heta Gylling as well as Dalius Jonkus and Elena Lysaniuk for kindly serving on

my dissertation committee. Hearty thanks go to Professor Leonidas Donskis not only for his

splendid friendship (both personal and philosophical) but also for his unflagging help at VMU

early on in helping me both psychologically and materially to overcome many bureaucratic

hurdles on my non-standard academic path. For their assistance in enabling me to fulfill all

requirements I’m also grateful to the intellectually enthusiastic Professor Egdijus

25

Aleksandravičius, head of the Lithuanian Emigration Institute, who secured me my first job at

VMU; to VMU University Rector Professor Zigmas Lydeka, who with his colleagues is doing

his utmost to help make this school into a modern European university with an ambitious

tradition in Baltic philology, historiography, and philosophy; and to other members of the VMU

community, including Kristina Jūraitė and Alisija Rupšienė, for all their support. (If I don’t

mention all it’s only for fear, in my “senior moments,” of forgetting some.) Last but certainly

not least I wish to express my immense gratitude to Professor Auksė Balčytienė: she not only

enabled me to put to profitable use my journalistic and teaching talents but also invaluably

guided me in taking the final steps required toward completion of my dissertation.

I dedicate this work to my wife, Vida, who truly is my better half.

26

Chapter One: Introduction

1.1. Aiming towards Idealism

Ever since the time of Descartes the “nature of perception—the sensory perception of items in

the physical world by human subjects”9 has been a live, and continues to be an unresolved,

philosophical problem. Though there always have been broad consensuses on the ontological

nature of the perceived object and on the nature of our epistemological access to it, these have

not lasted forever, sometimes taking centuries to change, sometimes far less. Thus for the last

few decades, at least in the Anglo-American tradition, the majority opinion has increasingly

favored Direct Realism, which gradually took over its present-day leading position from the

Indirect or Representative Realism (sometimes also called Causal Realism as well as, less often,

Critical Realism) that reigned supreme (though not unchallenged) in the early and middle parts

of the last century; while still earlier it was Idealism that had been predominant (though, again,

not exclusively so) for more than a century, having replaced the older forms of Representative

Realism that had largely prevailed throughout the early modern period. But at no time was any

of these three basic positions either utterly dominant or utterly dead.

For the sake of orientation, let me say that I will understand these views in the way that John

Foster has characterized them at the beginning of his fairly recent The Nature of Perception:

First, there is Direct Realism. This accepts a Realist view of the physical world: it takes

the physical world (the world of physical space and material objects) to be something

whose existence is logically independent of the human mind, and something which is, in

its basic character, metaphysically fundamental. And, within this realist framework, it

takes our perceptual access to the physical world to be direct. Second, there is the

Representative Theory (or Representative Realism). This too accepts a Realist view of

the physical world. But it sees this realism—in particular, the claim of mind

independence—as putting the world beyond the reach of direct perception. Thus, in

place of the claim that our perceptual access to the physical world is direct, it insists that

the perceiving of a physical item is always mediated by the occurrence of something in

the mind which represents its presence to us. Finally, there is Idealism. This agrees with

the Representative Theory in holding that direct perceptual awareness does not reach

beyond the boundaries of the mind, but manages to combine this with the insistence that

9 The concept is ordinary; the words are John Foster’s. Much of what follows is either under the direct influence of, or, to a lesser extent, in conscious opposition to, this seminal modern Idealist philosopher.

27

our perceptual access to the physical world is nevertheless direct. What enables it to

combine these seemingly irreconcilable views is that it abandons Physical Realism. Thus

it takes the physical world to come within the reach of direct perceptual awareness by

taking it to be something which is logically created by facts about human sensory

experience, or by some richer complex of facts in which such sense-experiential facts

centrally feature.10

In the rest of his book Foster undertakes to refute both Direct Realism and Representative

Realism and thereby to prepare the ground for a defense of Idealism. My dissertation, too, is an

attempt to help clear the way for Idealism. However, it is much more limited in that I devote

myself to discussing and defending just one argument against Direct Realism, although I attempt

to do so in a way that favors Idealism rather than Representative Realism.11

1.2. The Crux of the Time-Gap Argument: Must a Thing Exist When It Is Perceived?

Both Foster and I aim to do something that David Smith, in an even more recent book, claims to

be impossible. There he argues that “Direct Realism cannot be shown to be false … by

reflecting on the nature of perception”12. In aid of this, Smith extensively and carefully discusses

the Argument from Illusion and the Argument from Hallucination, and attempts to show that

they do not pose insuperable threats to Direct Realism. However, he says very little about, and

thinks very little of, the Time-Gap Argument (henceforward, TGA), which I believe has a much

greater chance of discrediting Direct Realism than Smith admits. The over-all aim of my

dissertation is to make this belief plausible.

10 John Foster, The Nature of Perception (OUP, 2000): 1. This classification (meant to be exhaustive) of views on the nature of the perceived object and of our access to it will be the main focus of my discussion. However, since views are also classifiable with respect to what kind of item (relation, state, or event) the perceiving itself is taken to be, I will engage this related issue too. 11 Currently Idealism is a thoroughly unpopular position. The only philosophers I know of in the Anglo-American tradition nowadays championing versions of it are, I believe, Robert M. Adams, John Bolender, A. C. Grayling, Peter B. Lloyd, Howard Robinson, and, first and foremost, the unfortunately recently departed John Foster. To the list of recent idealists we might add T. L. S. Sprigge and, a bit further back in time, my teacher Roderick Firth and the latter’s teacher C. I. Lewis. Several currently active philosophers, including Jerry Valberg, Peter Unger, and David Smith, are in one respect or another seriously sympathetic to it. Thus in the present philosophical climate defenders of Idealism have their job cut out for them. My dissertation is intended to be but a small contribution. 12 A. D. Smith, The Problem of Perception (Cambridge, Mass. and London, England: Harvard University Press, 2002): 5-6. However, Smith doesn’t mention that Francis Bacon might be considered to have prefigured the TGA. It was Arthur O. Lovejoy who in The Revolt Against Dualism (p. 23) quoted certain passages from Novum Organum, II, 46 to show that the “probability” of “a temporal sundering, and therefore an existential duality, of the content given and the reality made known to us through that content” had “suggested itself” to Bacon, though the latter then unfortunately discounted that probability “for several bad reasons.” Lovejoy went on to claim that the conjecture of “the images or rays of the heavenly bodies” taking “some appreciable time in travelling to us” had been “not only propounded but embraced and defended” already by Francis Bacon’s “subtler medieval namesake” Roger Bacon “three centuries earlier.”

28

What is valuable in the little that Smith does say about the TGA is his reminder, relegated to

a footnote, that Leibniz was (one of) the first to state it “concisely.”13 He did this in his New

Essays on Human Understanding, in the context of claims (by Philalethes, a spokesman for

Locke) that “the ideas which are received by sensation are often altered by the judgment of the

mind in grown people” and that “we substitute the cause of the image for what actually appears

to us, and confound judging with seeing.” Assenting to this, Leibniz (in the guise of Theophilus)

says that when

we are deceived by a painting our judgments are doubly in error. First, we substitute the

cause for the effect, and believe that we immediately see the thing that causes the image.

For strictly we see only the image, and are affected only by rays of light. Since rays of

light need time—however little—to reach us, it is possible that the object should be

destroyed during the interval and no longer exists when the light reaches the eye; and

something which no longer exists cannot be the present object of our sight. Secondly, we

are further deceived when we substitute one cause for another and believe that what

comes merely from a flat painting actually comes from a body.14

It is, of course, in the course of describing the first deception that Leibniz so succinctly laid out

this pioneer version of the TGA. In the body of his own book Smith implies that this argument

can “either be reflected ambulando or be independently answered without too much

difficulty”.15 When asked about how he would proceed to deflect it, he replied he would do so

by rejecting its crucial premise that, necessarily, any object of direct awareness exists at the time

of the awareness. According to Smith, this premise can be rejected “without absurdity.”16 But

can it?

Many philosophers share Smith's view on this issue but many others do not. The latter, unlike

the former, have a strong intuition that the notion of direct perceptual awareness of something in

the past just doesn't make any sense. Clearly, the point can rationally be decided only by

argument. Therefore, let me, following Howard Robinson,17 set out an argument for this premise

as follows:

(1) The direct perceptual awareness of a star, for example, is a psychological state that

has content.

13 ibid., 275, n. 27. 14 Leibniz, G. W. New Essays on Human Understanding, tr. P. Remnant and J. Bennett (Cambridge: Cambridge University Press, 1982): 134-135. 15 Smith, The Problem of Perception, 9. 16 Personal communication. 17 Here I am expanding an argument given on p. 81 of his Perception. Robinson is another Idealist philosopher whose work has inspired the present essay.

29

(2) The content of a psychological state is what (in whole or in part) constitutes that

state, what (in whole or in part) makes it the state that it is.

(3) What constitutes the direct perceptual awareness of a star, what makes it the state that

it is, is, in part, the star one is directly perceptually aware of.

(4) Anything which (in whole or in part) constitutes something else must exist at the

same time as, or overlap in time with, the thing which it wholly or partly constitutes.

Therefore,

(5) The content of a direct perceptual awareness of a star cannot cease to exist at an

earlier time than the direct perceptual awareness of which it is the content.

I think that this argument is sound,18 but most philosophers seem to be unaware of it. In an

early work on perception the Direct Realist Fred Dretske objected to its conclusion (i.e., that any

object of perceptual direct awareness must exist at the time of the awareness) on the grounds

that it is neither desirable nor warranted. I will respond to his elaborate and revealing

argumentation, as well as that of Anthony Quinton and David Armstrong, two other Direct

Realists, in Chapter Two of my dissertation.

Jonathan Dancy, likewise a Direct Realist, has claimed that this conclusion masks—and

illegitimately trades on—an ambiguity (and also relies on an inadmissible infallibility). I will

discuss his arguments together with John Foster’s, who perhaps surprisingly is no fan of the

TGA. The latter’s arguments are as follows.

In comments to me in an e-mail exchange, John Foster has objected to (4) on the grounds that

it isn’t “obvious that anything which partly constitutes something else must exist at the same

time as, or overlap in time, this other thing,” and anyway, that my claim would be taken by the

Direct Realist as begging the question. In answer to my question, “If ‘this other thing’ exists

from time t onwards to time t+1, then how can the thing that partly constitutes it (and thereby is

a part of it) exist entirely before time t?,” Foster replied by saying that the Direct Realist “won't

think that the earlier object is part of the later perceptual episode, but only that the perceptual

state (type-state) which that episode realizes is a state of being aware of that earlier item, and so

a state that could not be realized if that item did not exist (my emphasis). It is only in this sense

that he will take the earlier item to be partly constitutive of the later perception.”19

But why could that state not be realized if that (earlier) item did not exist? It is, I think, only

because this would violate the conceptual requirement that in order to be perceived, an item

must exist at some time or other—perhaps (leaving this issue provisionally open) not necessarily

18 Speaking on the same page about the “claim that content and act must be simultaneous” Robinson says he finds “its intuitive appeal overwhelming.” That’s exactly my view. 19 Personal communication.

30

the same time as that at which it is perceived. Let us call this the existence-requirement for

perception (which few if any philosophers nowadays would be willing to question), and

distinguish it from the simultaneity-requirement. The latter is something that both the Direct

Realist (or rather, to be more exact, most Direct Realists, though emphatically not all) and

Foster (who tries to refute Direct Realism on other grounds) reject. In Chapter Four, I will

defend it further against Foster’s just-mentioned misgivings as well as those of Dancy and of

many other like-minded philosophers writing at the beginning or middle of the last or the

beginning of the present century.

But, first, in Chapter Three I will present several of the other arguments of these like-minded

philosophers. These are surprising arguments in that they deny the existence of a radical

separation between perception and memory and claim that one can gain genuine perceptual

access to the past not only by remembering things, but also by looking at past things as well as

looking at photographs of them or watching television broadcasts.

In addition to objecting to the simultaneity-requirement, Dretske, in a later exchange, has

objected to (3) as well by claiming that the content of a psychological state, e.g., what makes the

direct perceptual awareness of a star the state that it is, is not the star, but “is completely given

by the properties that the state represents” the star “as having.” The content, says Dretske, “is

just like a measuring instrument. A thermometer is in the same representational state (has the

same content) if it registers 95 degrees in location L (representing L as being 95 degrees) as it

does when it registers 95 degrees in location L' (representing L' as 95 degrees). Likewise, I

think, an experience of a star (a distant object that no longer exists) as a bright shiny object has

the same content as an experience of a nearby object (which still exists) as a bright shiny

object.”20

In Chapter Five, I will give reasons for thinking that Dretske’s account (called the

Representationalist Theory of Perceptual Experience or Representationalism or

Representationism21 for short, and currently quite popular) is not at all compelling. But even if

I’m wrong and the Representationalist Theory of Perceptual Experience is correct, I will argue

that this need not militate against the simultaneity-requirement for perception.

In the same chapter I will also briefly discuss James W. Cornman’s presumptive objection to

(3) on the grounds that “my present perceptual experience of a distant star consists in my being

in a state of star-sensing now as a result of my eyes being suitably affected by stimulus

20 Personal communication. 21 Not to be identified with the Representative Theory (Representative Realism) of Foster’s classification, although the difference isn’t straightforwardly clear. See sections 5.2. and 5.4.

31

previously transmitted from the star...”22 A star-sensing, according to him, is neither a star nor

the awareness of a star but an objectless, adverbial sensing event. I will give reasons (largely

unoriginal ones) for saying that no such events exist.

1.3. Other Objections to the TGA

Not all philosophers, not even all Direct Realists, who attack the TGA do so by denying the

simultaneity-requirement. Len Carrier,23 for example, is one Direct Realist who accepts it. He

can do so because he rejects the Leibnizian version’s implied conclusion that what we

immediately see is only the image rather than the thing that causes the image, i.e., the star. He

rejects this conclusion because he accepts (1) the view that though we do not immediately see

either the star or its image, we do immediately see the star’s light and (2) the view that both the

star and its light are physical objects rather than images of them. These views, I will argue

contra Carrier, to the extent they are correct actually strengthen the TGA’s force against Direct

Realism.

Other Direct Realists (David Lewis and Hanoch Ben-Yami) suggest an interpretation of the

scientific facts that tries to block the TGA altogether by denying that a star can go out of

existence before light from it reaches the seeing subject. If it cannot, it follows, in the words of

Lewis, that “...the stars, as I now see them, are not straighforwardly past; for lightlike

connection has as good a claim as simultaneity-in-my-rest-frame to be the legitimate heir to our

defunct concept of absolute simultaneity.”24

This view poses what is perhaps the greatest challenge to the TGA. Still, as I will try to

show, I think it can be overcome. The discussion will involve distinguishing two time gaps: (1)

the one between the time light ostensibly leaves a star and the time it arrives at the observer’s

eye, and (2) the much shorter one between the time it arrives at the observer’s eye and the time

(a certain processing in the brain having taken place) the star or its light or its image is seen. I

will argue that even if the first time gap should (contrary to fact, I believe) pose no problem for

the Direct Realist (as Lewis and Ben-Yami25 argue), the second time gap (an intra-organismic

22 J. W. Cornman, Perception, Common Sense, and Science (New Haven and London: Yale University Press, 1975): 49-50. 23 L. S. Carrier, “The Time-Gap Argument,” Australasian Journal of Philosophy, Vol. 47, No. 3 (December, 1969): 263-272. 24 D. Lewis, “Veridical Hallucination and Prosthetic Vision.” In: J. Dancy, ed. Perceptual Knowledge (OUP, 1988): 83. 25Hanoch Ben-Yami, “Apparent Simultaneity.“ In [2007] EPSA07: 1st Conference of the European Philosophy of Science Association (Madrid, 15-17 November, 2007).

32

one) certainly does (contra Ben-Yami again, but now basing himself on Daniel Dennett26); but it

does so, I admit, only if certain other arguments against Direct Realism work.

1.4. A Berkeleian TGA

Although all versions of the TGA initially aim to refute Direct Realism, some go on to try to

prove Indirect Realism, while others proceed in an opposite direction, first trying to refute

Indirect Realism and then going on to prove Idealism. Particularly pertinent here are some of the

claims and arguments of George Berkeley, whose thought will play a major role in what

follows.

This is a good place to introduce an absolutely essential distinction between two kinds of

Direct Realism. As Berkeley wrote towards the end of his Three Dialogues, “My endeavors tend

only to unite and place in a clearer light that truth, which was before shared between the vulgar

and the philosophers: the former being of opinion, that those things they immediately perceive

are the real things; and the latter, that the things immediately perceived, are ideas which exist

only in the mind. Which two notions put together, do in effect constitute the substance of what I

advance.”27

The first of these “notions” is a Direct Realism that is distinct from the one the TGA aims to

refute. The latter Direct Realism, to put it in the words of David Smith, combines (1) an

epistemological component, the view that we “directly perceive the physical world”, with (2) a

metaphysical component, the view “that the physical world has an existence that is not in any

way dependent upon its being ‘cognized’—that is, perceived or thought about.” Smith goes on

to say that according to Direct Realism’s metaphysical component, “[t]he physical world is not,

as it is usually put, dependent on ‘consciousness,’ at least not finite consciousness.” This gives

us a Realism that “is opposed to Idealism: the view that whatever seems to be physical is either

reducible to, or at least supervenient upon, cognitive states of consciousness”.28

Berkeley’s Direct Realism, too, has an epistemological component, and it is the same as the

one in the Direct Realism discussed by Smith; but its metaphysical component is just the

opposite. The metaphysical component of Berkeley’s Direct Realism is captured in the second

“notion” that he mentions, which, of course, implies Idealism as characterized by Smith. Thus

we may call Berkeley’s Direct Realism Idealist, and Smith’s Direct Realism Non-Idealist.

26 Daniel Dennett, Consciousness Explained (Boston: Back Bay Books, Little Brown and Company, 1991): 101-170. 27 George Berkeley, Works, 3D, 262 28 Smith, The Problem of Perception, 1-2.

33

Now while Berkeley did not in fact use the TGA, he could have used it. There is nothing in

his philosophy that would make it inadmissible or inadvisable for him to argue for Idealist

Direct Realism as follows:

(1) Suppose I look at what I take to be the sun.

(2) Since rays of light from the sun, according to the physicists, need about eight

minutes’ time to reach me, it is possible that the sun should be destroyed during that

interval and no longer exist when its light reaches my eye.

(3) Now just suppose that the sun actually has been destroyed during this interval.

(4) But something that no longer exists can’t be something that I now immediately see.

(5) Yet I do see something immediately that I take to be the sun.

Therefore,

(6) What I immediately see is the visible sun, not the sun of physics that I suppose to

have been destroyed.

(7) But I immediately see only one visible sun and nothing else having a “conformity or

resemblance to”29 it; only one dazzlingly bright round thing in the heavens rather than

two.

(8) And in addition to the immediately seen visible sun I do not immediately or

mediately see any other thing resembling it.

(9) If Indirect or Representative Realism is true, then I can see something that no longer

exists and I do see both the visible sun and the physical sun I suppose to have been

destroyed.

Therefore,

(10) Indirect or Representative Realism is false.

(11) This visible sun, though perhaps secondarily caused by the sun of physics, has, like

all things, strictly and truly been caused by God.

(12) But the visible sun is a collection of, or a “fiction of our own brain”30 dependent on,

ideas of color, light, and figure.

(13) This collection of, or this “fiction of our own brain” dependent on, ideas of color,

light, and figure, either exists without the mind or within the mind.

(14) If it exists without the mind, then it could radically change without my noticing it.

(15) But this collection of, or this “fiction of our own brain” dependent on, ideas of

color, light, and figure, could not radically change without my noticing it.

Therefore,

29 Berkeley, Works, 203. 30 Berkeley, Three Dialogues, Preface

34

(16) The visible sun I see exists within the mind.

(17) Even if the sun of physics had not been destroyed (as in reality it never has been and

most likely never will be in human life-time), I still would be immediately seeing the

visible sun only—for exactly the reason mentioned in (7).

(18) Suppose that the sun of physics has not been destroyed, i.e., that (3) is false.

(19) Then in the very act of immediately seeing the visible sun I am also perceiving,

whether immediately or mediately, the undestroyed sun of physics, which is a larger

collection of, or a more complex “fiction of our own brain” dependent on, ideas of which

collection or fiction the visible sun is a part.

Therefore,

(20) Whenever I see the visible sun I perceive a larger collection of ideas, some of which

I perceive immediately.

(21) But what is true of perceiving the visible sun is true, mutatis mutandis, of

perceiving any sensible object whatsoever;

Therefore,

(22) All sensible objects are ideas or collections of ideas which exist only in the mind.

QED.

Although I find this Berkeleian TGA entirely congenial, in this work I mostly engage with

only a part of it—the part that helps to refute Non-Idealist Direct Realism. And that, to begin

with, comes right at the beginning.

Thus all the steps before the first conclusion (6) are either legitimately supposed or

uncontroversial except for (4), which is a corollary of the simultaneity-requirement. Once (4)

has been fully defended, (6) is firmly secure. But since “the visible sun” is what later

philosophers would have called “an array of sun-like sense-data” or “a complex of sense-data of

the sun” or something similar with the expression “sense-data” (or synonyms thereof) in its

name, the argument for (6) establishes the existence of sense-data31 and thereby the falsity of

Non-Idealist Direct Realism at least in the case of seeing the sun.

In order to get the same result for all cases of seeing, defenders of the TGA typically resort to

a Continuity Argument. Premise (20) represents the conclusion of such an argument but the

argument itself is not contained in the above Berkeleian TGA. Yet there are considerations not

31 I assume that the many arguments voiced over the years against positing sense-data can be overcome. For strong defenses of this assumption see, inter alia, Foster, Ayer, 144-191; Robinson, Perception; Foster, The Nature of Perception, 147-170, 186-195; and O’Shaughnessy, Consciousness and the World, 439-514. The last three books especially make any claim that the sense-datum theory should be relegated to the philosophical dustbin seem hollow, pretentious, and closed-minded.

35

explicitly offered by Berkeley out of which such an argument may readily be constructed. This,

too, is something I will try to do.

Once the existence of sense-data is secure, Non-Idealist Direct Realism has been defeated,

thus leaving us to choose between Indirect (Representative) Realism and Idealist Direct

Realism. Although justifying this choice goes beyond what this dissertation minimally aims to

accomplish, I will, in the course of discussing the Berkeleian and other versions of the TGA,

argue in favor of considerations that support the rejection of Indirect Realism. And my defense

of Idealist Direct Realism will largely (though not exclusively) be limited to the point that if

Non-Idealist Direct Realism and Indirect Realism have been eliminated, Idealist Direct Realism

is the only option left.

1.5. The Importance of the TGA

Although the TGA might have originated with Leibniz, the first extended discussion of it

occurred much later—almost a hundred years ago in a symposium called “The Time Difficulty

in Realist Theories of Perception” and published in the Proceedings of the Aristotelian Society

11 (1911-12), with H. W. Carr, F. B. Jevons, W. Brown, and G. Dawes Hicks participating.

From at least that time onwards it always found people willing to defend it or to refute it. Of

major early 20th century British philosophers, two—Bertrand Russell and C. D. Broad—

endorsed it strongly (Broad) or even wholeheartedly and repeatedly (Russell). Their

contemporary G. E. Moore, who did not, is nonetheless reported to have remarked that to many

people it appeared “absolutely conclusive.”32 One of these, whether or not Moore had exactly

him in mind, was the American philosopher and historian of ideas A. O. Lovejoy (1873-1962).33

His presentation and defense both of (1) the epistemological dualism of sensed data and

perceived objects—with the TGA receiving pride of place in this connection—and of (2) the

metaphysical dualism of mind and body still seem to me to be not far off the mark and well

worth reading. Other important philosophers who have thought highly of the TGA include

Maurice Mandelbaum (1908-1987) and our contemporary J. J. Valberg. The latter has produced

and defended what in effect is a short version of the TGA or, rather, a strongly related argument

that he thinks leads to a “puzzle of experience” deemed by him to be insoluble but which I will

try to show can be solved by adopting a Berkeleian Idealist Direct Realism.

Nevertheless, a much greater number of other philosophers interested in perception have

discussed the TGA very little or not at all, or else they have criticized it. Of the latter, some have

32 In Lectures on Philosophy, edited by C. Lewy (London, 1966): 67. 33 See The Revolt Against Dualism, especially pp. 23-25, 77-85.

36

thought it important enough to deserve a very thorough examination. These include David

Armstrong, James Cornman, Jonathan Dancy, Fred Dretske, Moltke Gram, George Pitcher, and

Anthony Quinton, all writing in the second half of the last century. But even in the first decade

of the 21st century the argument has excited interest, though one young philosopher criticizing it

no longer calls it the TGA or speaks of a perceptual “time-gap” or “time-lag”; instead he hails

what he takes to be the actuality of “interstellar perception.”

I will discuss the TGA-related views of all these philosophers, including a few others perhaps

no less important but not mentioned so far. It is my hope that even if my dissertation falls short

of its goal of showing that the TGA proves the existence of sense-data and disproves Non-

Idealist Direct Realism, it will at least constitute a reasonably complete critical survey of all the

positions so far taken on what Mandelbaum has called “one of the strongest of all the arguments

which can be used against direct realism.”34

But how, in addition to working toward the same conclusion, is the TGA related to these

other arguments (including the Argument from Illusion and the Causal Argument)? In what

ways, if at all, are these arguments inferior to the TGA? On the other hand, can the TGA really

stand on its own, or is it (as I’ve already suggested at the end of 1.3.) in need of their assistance,

after all? No discussion of it would really be enlightening and complete without answers to these

questions. These will also emerge in the following pages.

Finally, a concluding remark about my procedure. In Chapters Two, Three, and Four I will

provide a full defense of the simultaneity-requirement, first by presenting and trying to deflect

attacks on it and then by giving positive reasons to uphold it. In Chapter Five I will criticize

Dretske’s Representationalist Theory and Cornman’s Adverbial Theory of Perceptual

Experience. Chapter Six will feature a discussion of the second and smaller time-gap

(mentioned toward the end of 1.3. above), of some other arguments for sense-data, and a

conclusion in which the key insights of Valberg, Cornman, Carrier, Foster, and Berkeley are

joined together. Brief discussions of, or references to, other philosophers writing on the TGA

will occur throughout the text. Hopefully, by the end, Direct Realism of the Non-Idealist variety

will, insofar as it is Non-Idealist, not seem to be the most attractive option; and Direct Realism

of the Idealist kind will not seem as unattractive as it is still commonly supposed to be.

34 M. Mandelbaum, Philosophy, Science, and Sense Perception (Baltimore: The John Hopkins Press, 1964): 174.

37

Chapter Two: Denying the Simultaneity-Requirement (I)

2.1. The Existence-Requirement

If anything whatsoever is seen, that thing must exist and it must exist at the time it is seen. The

consequent of the foregoing conditional has two clauses, the first of which states the existence-

requirement and the second the simultaneity-requirement. As I’ve already indicated, the first is

now largely undisputed while the second is not. On my view, both conditions have to be

satisfied.

First, a few words about the existence-requirement. The issue of whether, with respect to this

requirement, the verb “to see” has one sense or more than one used to be vociferously argued

about some fifty or more years ago. Three positions have been taken on this issue: (1) there is

one sense (use) of “to see” and this sense (use) logically requires that the thing seen exist; (2)

there is one sense (use) of “to see” and this sense (use) does not require the existence of the

thing seen; (3) there are two senses (uses) of “to see,” one of them requiring the existence of the

thing seen and the other not requiring it.

As far as I’m aware, this issue has been settled (as far as anything in philosophy can be

“settled”) by Fred Dretske in his 1969 book Seeing and Knowing, a seminal work in third-

quarter-century philosophy of perception. There he lengthily presents what to my mind is a fully

satisfactory defense of the view that there is only one sense of “see,” a sense governed by the

existence-requirement. It is what he calls a “non-epistemic” sense whose correct application to

any agent does not require that agent to have any concepts, beliefs or even minimal human

understanding. In this sense a dog as well as a human being, a child as well as a scholar can see

a floppy disk even though he, she, or it may not understand, or be capable of understanding,

what it is they see. And it is in this sense that, as Dretske puts it, “one cannot truly say of S

(whether this be some other person or yourself) that S sees a D unless there exists a D, a real D,

which S sees.”

Now, of course, sometimes this existence condition may be suspended by the words in which

a perceptual report is couched or by the context in which it is given. Thus, if the report “George

B. saw the female Pope praying to Allah” is preceded by the words “It looked to George B. as

though” or the words “In George B.’s dream,” or the words “There’s a story that goes as

follows:. . . ,” the existence-requirement is on that occasion cancelled without its thereby being

implied that there is a sense of “to see” not governed by the existence-requirement or that there

are two senses one of which is so governed and the other isn’t. As Dretske observes,

38

[t]he fact that there is only one sense of the verb in question here (in relation to the

existence condition) is supported by the fact that unless we signal our suspension of the

existence condition by some manner or means, our listener can object to our claim to

have seen a D on the grounds that there are no D’s (or no D’s in the vicinity). If there

were two senses of the verb, one of which was not governed by an existence condition,

this type of objection would not be generally available. For, then, it would be possible,

with a straight face and in all sincerity, to truly say that one saw Harold at the scene of

the crime without even believing he was there, and indeed without his having been

there.35

This is all I will say in defense of the existence-requirement for now; to my knowledge,

Dretske’s argument on this point has been neither effectively challenged nor refuted.

The simultaneity-requirement (the simultaneity principle, the simultaneity-thesis: I use all

three names interchangeably) is a different matter. Soon after defending the existence-

requirement he goes on to discuss “the question of when D must exist to be seen. That is, can

we see stars that no longer exist?” Dretske claims that the latter question must be answered

affirmatively and that the simultaneity-requirement does not apply to seeing; he spends even

more time on arguing against the simultaneity-requirement than he does on arguing for the

existence-requirement. Since his and my intentions are opposed here and since I believe his

negative argument to be instructive but flawed, I will use his argument against the

simultaneity-requirement as a way of eventually getting into my argument for it.36

2.2. Dretske on the TGA and the Simultaneity-Requirement

Though Dretske rather thoroughly prepares the ground for the TGA, unfortunately he chooses

an ill-suited hinge on which to make the argument turn, viz., a coffee pot. (Why I think it is

poorly chosen will emerge a bit later.) Dretske starts out by retelling the well-known causal

story:

If the statement ‘S sees a coffee pot’ is true, then, as a matter of well-established

scientific fact, light is being reflected from the coffee pot into S’s eyes, the light is

stimulating certain sensitive cells which, in turn, are transmitting electrical impulses via

35 Dretske, Seeing and Knowing, 49-50. Dretske’s defense of the existence-requirement occurs on pp. 43-50. 36 My quoting and discussing it at such great length is additionally justified by its having helped persuade David Smith (as I learned in a personal note from him) to give the TGA exceedingly short shrift. That alone makes a careful engagement with Dretske’s argument worthwhile. I should mention that Dretske’s 1969 text represents one of the last sustained endeavors to refute the TGA before it was taken up again in the early 1990s by Valberg and Robinson (favorably) and subsequently by a larger number of other philosophers (mostly unfavorably).

39

the cortical nerve to the cerebral cortex where a pattern of excitation occurs. As a

consequence of this chain of events, S undergoes an experience, an experience which we

might ordinarily describe by saying that S (first) sees the coffee pot. Let us refer to this

typical or standard sequence of events with the notation e1, e2, e3, ...; the terminal event

or stage, en, will be that last event or state of affairs which, until it happens, S has not yet

seen the coffee pot. ... I shall simply refer to en as the terminal stage en in the standard

causal sequence associated with ‘seeing a coffee pot’.

If the sequence e1, e2, ... en is interrupted at any stage, by inserting a barrier between S

and the coffee pot, by S’s closing his eyes, or by some severe damage to S’s nervous

system, then the latter members of this sequence will not occur and S will not (as we

commonly say) see the coffee pot. Nonetheless, the terminal stage in this sequence of

events, or something very much like the standard terminal stage, that which does not

normally occur until e1 through en -1 have occurred, can occur without these particular

causal antecedents. We might, theoretically at least, stimulate S’s cerebral cortex by a set

of electrodes and thereby create for S a visual experience which was, from its subjective

side, indistinguishable from the one that was initiated by light from the coffee pot in the

standard or normal case. Or, if S is suffering from some kind of hallucination, the last

member (or something very much like the last member) of our standard sequence may

occur, or the latter members of it may occur, without the usual antecedents; S might (if

he makes a mistake) describe himself as seeing a coffee pot when there is no coffee pot

which is reflecting light into his eyes.

Finally, . . . there is a brief temporal interval, measurable perhaps only in

microseconds, between the time when the light leaves the coffee pot and the occurrence

of the terminal stage (i.e., between e1 and en). When we speak of seeing the moon, the

sun, or the stars, the temporal interval becomes appreciable. In some of these cases one

can legitimately speak of the causal sequence as beginning many years prior to the

occurrence of the terminal event itself. This is sometimes expressed by saying that when

we see a star we see it as it was many years ago, light requiring that long to reach us

from the star.37

From this uncontested statement of facts, says Dretske,

it seems to follow that the terminal stage itself is the most significant stage of the causal

sequence; if it occurs, whatever its causal antecedents, then the percipient will undergo

an experience which, in its subjective aspect, is indistinguishable from what he

37 Dretske, Seeing and Knowing, 68-70. Subsequent quotations from this book are all from pp. 70-75.

40

experiences when he (as we ordinarily say) really sees a coffee pot. But when he really

sees a coffee pot, we want to say that the coffee pot looks some way to him. And this

seems to suggest that within the terminal state itself there is something describable as a

coffee-pot-looking-something, an element which he takes to be the coffee pot in the

standard case and which he might mistakenly take to be a real coffee pot in the non-

standard case. When this terminal event has the standard causal antecedents we

ordinarily say that S sees a coffee pot (the coffee pot looks some way to S); when it does

not have the standard causal antecedents, we say he is hallucinating or whatever—in this

case it definitely is not a coffee pot which he might mistakenly take to be a coffee pot.

But there is no significant difference between these two events aside from their causal

antecedents. We seem forced to the view that in both the standard and the non-standard

case an event occurs which may be described by saying that S is visually aware of

something which he may or may not take to be a (real) coffee pot. In the non-standard

case it is certainly not a coffee pot. But even in the standard case it cannot be said to be

the coffee pot itself since, if for no other reason, the coffee pot may no longer exist when

this terminal event occurs. Hence, we may say that a percipient is directly aware of the

element which is contemporaneous with the terminal event; when this terminal event has

the standard causal antecedents we say that S sees the coffee pot, but this should now be

understood to mean that S sees the coffee pot indirectly since it is not the coffee pot

which looks some way to him. The coffee pot is merely associated with that causal

sequence which terminates in an event (en) which, whether it has these antecedents or

not, may be described by saying that something looks to him as we would ordinarily say

the coffee pot looks to him. It is this something which he sees directly.

Something like this is behind every attempt to introduce a direct vs. an indirect way of

seeing things by analyzing the causal process of perception. The idea is that we are

always, even in cases that are normally described by saying that we see an object of

some sort, directly aware of an ‘impression’, ‘percept’, or ‘sensation’ which is aroused in

us by the causal action of light. This is an extremely seductive argument, and I do not

wish to belittle its persuasiveness.

After generously admitting the force of this argument Dretske nonetheless claims “there is

one, perhaps two, mistakes in it.” What are they? He explains the first (alleged) mistake as

follows:

In hallucination (or artificial stimulation of the brain) we may have a visual experience,

call it en’, which is indistinguishable from the experience we have when we are said to

see a real coffee pot—call this en . Now, since en and en’ are (or may be)

41

indistinguishable and in the case of en’ it is clear that we are aware of something which

is not a coffee pot (although we may take it to be a real coffee pot), then also in en, the

so-called veridical case, we must be aware of something which is not a real coffee pot.

The pattern of this argument is obviously fallacious. I may not be able to distinguish

between S’s handing me a genuine one-dollar bill (en) and S’s handing me a counterfeit

one-dollar bill (en’). If the counterfeit is good, the two events may be indistinguishable.

Surely, however, we cannot conclude that because I am being handed a counterfeit bill in

the one case, I must therefore be receiving a counterfeit bill in the other case. It may well

be that in what we commonly refer to as the experience of seeing a coffee pot, there are

other (hallucinatory) experiences which are indistinguishable from it. But this does not

imply that the two sorts of experience involve an awareness of the same sorts of things.

We can quite consistently maintain that in the standard case we are visually aware of the

coffee pot, it is the coffee pot itself which looks some way to us, although this

experience is indistinguishable (subjectively) from other experiences in which we are not

aware of a coffee pot.”

In my view, however, what is fallacious here is not the argument that Dretske attacks, but

Dretske’s attack on it. The last sentence quoted raises the question, If in the non-hallucinatory

case it is the coffee pot itself which looks some way to us and if this experience is subjectively

indistinguishable from another experience in which there’s no awareness of a coffee pot, then

what is it that makes the latter experience indistinguishable from the former? Dretske nowhere

so much as even tries to answer this obvious question. I will return to this point in a moment.

For now let me note that even though I may not be able to distinguish between getting a

genuine bill and getting a counterfeit one, in both cases I am really receiving a one-dollar bill;

and it is a matter for further investigation whether that bill is genuine or counterfeit. Surely

Dretske is right in saying that just because I am being handed a counterfeit bill in the one case, it

doesn’t follow I must be receiving a counterfeit bill in the other, just as he would have been

right in saying that merely because I am being handed a genuine bill in the one case, it doesn’t

follow I must be receiving a genuine bill in the other; for in both cases it must still be

determined whether that one-dollar bill is genuine or not; and just merely looking at them

(however intently) is often not enough for telling which is which.38

That latter claim is exactly the point of those philosophers who argue for the view that in

perception we are always directly aware of what Locke and Berkeley prefer to call “ideas” and

others call “appearances,” “presentations,” “phenomena,” “sensations,” “images,”

38 That is why, when bills are larger, cashiers and bankers routinely subject them to more searching tests.

42

“impressions,” “sense-impressions,” “percepts,” “prehensa,” “qualia,” “sensa,” “sense-data,” or

“sense-qualia.”39 These philosophers (let us call them “sense-datum theorists,” for convenience)

would not start out by saying that visual awareness is something that is analogous to being

handed a counterfeit one-dollar bill; if they made use of the analogy at all, they’d start out by

saying that visual awareness (of, say, something looking like a coffee pot) is analogous to being

handed a bill; and then they’d go on to claim that this visual awareness could be either an

awareness of a real coffee pot or an awareness of a hallucinated one, both possibilities being so

far theoretically open.

Now, of course, any given bill is either genuine or counterfeit, depending on its causal

history (i.e., whether or not it has been printed according to certain procedures in a certain

government press and/or vetted by certain authorities). Finding out which is the case would

involve either tracing this history or, what is usually the case, submitting the bill to more severe

inspection tests than merely looking at it with the naked eye; for the latter alone, to repeat, is

often insufficient for distinguishing the genuine article from a good counterfeit.

Suppose that a certain bill handed to me by S turns out to be counterfeit. It doesn’t follow—

and no one in his right mind would argue—that an indistinguishable bill handed to me by S (or

anyone else) will turn out to be counterfeit, too. But nor will anyone in the visual awareness

case argue (as Dretske says they will) that “since en and en’ are (or may be) indistinguishable

and in the case of en’ it is clear that we are aware of something which is not a coffee pot

(although we may take it to be a real coffee pot), then also in en, the so-called veridical case, we

must be aware of something which is not a real coffee pot”. What the above-mentioned

philosophers will argue, rather, is that any visual awareness of something looking like a coffee

pot is either an awareness of a coffee pot or an awareness of something else (something looking

like, without being, a coffee pot) and that both these types of awareness have a common

component, namely, an awareness of a complex of coffee-pot-like sense-data. The presence of

this component is what explains why the experience of the coffee pot’s looking some way is

subjectively indistinguishable from a similar experience in which there’s no coffee pot looking

any way at all. (This picks up the point raised a few paragraphs back.)

39 I, along with other philosophers accepting their existence, will understand these objects of perception to have the following features: (1) they are things of which perceivers are directly aware; (2) they are mental in the sense of being mind-dependent (which leaves open the possibility that they are also physical if, for example, the mind-body identity theory should be true, which it probably isn’t); (3) they actually possess standard sensible qualities; (4) they are not intrinsically “of” something (they possess no intrinsic intentionality) although they are usually taken or interpreted as being “of” something; and (5) they are either sensory particulars (in which case they are usually called sense-data) or sensory universals (in which case they are usually called sense-qualia or just qualia). I apply the term “sense-data” to both, and mark the distinction, when required, explicitly. I call something “physical” if and only if it exists in space-time, is a particular, and is perceivable by more than one observer.

43

Therefore, contra Dretske, it seems that these two types of experience do, after all, involve

an awareness of the same sorts of things.40 But this common component need not be all there is

to both experiences; it might not exhaust both of their contents. It might exhaust the content of

the hallucinatory experience, but not that of the veridical one. Moreover, there is an all-

important question left open that still has to be answered. Is my experience veridical or

hallucinatory? Is my visual awareness of something looking like a coffee pot an awareness of a

coffee pot or an awareness of something else (something that looks like though it isn’t a coffee

pot)? Now finding out the answer to the this question involves, at a minimum, determining what

preceded my experience and what follows it. This is something that the experience by itself does

not disclose.

So much for the first “mistake” allegedly found by Dretske. The second “mistake” in that

“extremely seductive argument” occurs in what he calls “its strongest form,” the TGA. But

before we get to that “mistake” let us finally look at Dretske’s own infelicity (or is it a sly

stratagem?) in presenting the TGA in terms of a coffee pot.

2.3. An Excursus: Physical vs. Phenomenological Simultaneity

Dretske claims that

(w)hen we see the coffee pot, the terminal stage in the causal sequence could occur at a

time when the coffee pot no longer exists. This is easier to appreciate in the case of

distant objects (stars), but even in routine cases the same feature is present although the

temporal interval which makes this possible becomes extremely small. Let us consider,

then, a standard causal sequence with this minor alteration: when the terminal event, en,

occurs, the object which initiated the sequence, the object which we ordinarily say we

see, has ceased to exist. Certainly this catastrophic incident will not affect the character

of en since the object has already (before it ceases to exist) exercised all the causal

influence on en of which it is capable. Now, when en occurs the percipient has the

experience which is ordinarily described by saying that he sees the object. Yet, in this

case, the object no longer exists. The percipient is aware of something which looks like a

coffee pot, but this something cannot be the coffee pot itself. And if this is so, it

indicates that the percipient is always directly aware of something other than the coffee

pot; for whether or not the coffee pot ceases to exist makes no difference to the terminal

40 At least, this is a highly plausible conclusion. But it has been disputed--if not by Dretske himself in this book–then by those who subscribe to the so-called disjunctive view of perception. I think this view has been persuasively criticized by, among very many others, Robinson (in his Perception, p. 152-159) and Valberg (in The Puzzle of Perception, p. 98-100).

44

event, and we have just shown that this terminal event may be described by saying that

something other than a coffee pot looks some way (e.g. like a coffee pot) to the

percipient.

My seeing a coffee pot is certainly a routine case of seeing; but what Dretske unfortunately

fails to acknowledge is that it’s actually impossible that in such a case the terminal event, en,

occurs after the coffee pot has ceased to exist. To see this, let’s again consider the Leibniz

version of the TGA. While Dretske uses a coffee pot as his sample of an object seen, Leibniz

uses a picture; but in both cases the distance between perceiver and object is roughly the same—

and equally minuscule—compared to the distance between any perceiver on earth and any

celestial objects she might perceive.

Now in his TGA Leibniz claims that we see something all right and believe it’s the picture

we see—but it can’t be the picture because it has been destroyed, so what we see is something

very much like it41 (otherwise we wouldn’t think we still see it), and that is the image. There is a

strong objection to this reasoning which applies to Dretske’s version of the TGA as well.

We may presume that Leibniz knew that light travels at such a high finite velocity and that

the distance between any picture and human eyes is so small that the picture’s destruction one or

two nanoseconds after light-rays had left it is something we could not consciously notice.

Since, however, he famously had a doctrine of unconscious perception, this would not have

impeded Leibniz from making a meaningful distinction between our seeing the picture and our

seeing its image. So Leibniz would probably have claimed, in defense of his time-gap argument

(and here I state this defense in the words of Howard Robinson who wrote them without explicit

reference to Leibniz),

that the fact that we cannot discriminate very small time intervals is simply irrelevant.

Whether or not I can discriminate such an interval, I do in fact perceive [the picture as it

was some minuscule part] of a second ago; and it could, conceivably, have changed in

some perceptible respect just in that time, so that by the time I in fact perceive it it is

different from how I perceive it to be. The context is relevantly extensional and not

intentional: my cognitive abilities in minute time discriminations are not to the point.

Speaking for himself now, Robinson then goes on to say that “the response to this defense [of

the TGA] would be that it misses the point.” This, I think, is exactly right; and it is a point that I

believe must be insisted on. “The point,” Robinson goes on to say, “is that the content must be

phenomenologically contemporaneous, which means that it must not be possible to discriminate

41Hence the word “proxy” (used by Smith) and the expression “perceptual deputy” (used by Gram) are very appropriate to describe the function of what Leibniz calls “images,” Berkeley calls “ideas,” and most subsequent philosophers call “sense-data.”

45

any time distinction between perception and content by direct experience. This condition is

satisfied provided that the physical time difference is insufficient to be noticed.”

Something like this is probably what Berkeley too would have said had he thought or heard

of Leibniz’s TGA. Even if it is a matter of some controversy whether he accepted the existence

of insensible corpuscles, he certainly did not believe there could be insensible perceptions.42 So

he would not have used this version of TGA to show anything, much less that we only see a

picture-like thing that is caused by the picture rather than the picture itself.

For Berkeley the TGA could have gotten off the ground only if we knew, on independent

grounds, that a certain visible object had ceased to exist at time t and yet a perceptibly

appreciable moment later we still seemed to see it, i.e., it appeared to us as though it were still

there. This is certainly not the case with coffee pots or any other familiar, nearby things for most

human observers in most circumstances.

That’s why Berkeley, had he used the TGA, would have stated it in terms of seeing a star, the

sun, or even the moon. And that’s what most philosophers who discuss the TGA have in fact

done. Except for Dretske who, I will now contend, misjudges the argument and underestimates

its strength largely because of that inappropriate example of his. More importantly, he not only

misgauges the argument, but also tries to mask this fact and thereby inadvertently plays into the

TGA’s hands.

2.4. Dretske and Direct Awareness (I)

Immediately following my last quote from his text (section 2.3), Drestke takes up direct

awareness only eventually to shortchange it—in a way that I find both crafty on Dretske’s part

and ultimately self-defeating. Showing this will take some time. He first says this:

Being directly aware of something is a state of affairs which implies that the element of

which one is directly aware must exist at the time one is directly aware of it. Since

science has shown us that coffee pots need not exist at the time when, as we ordinarily

say, we see them, we must conclude that we are never directly aware of coffee pots—nor

anything else the perception of which involves a causal sequence involving a temporal

interval.

42 It is a separate question whether or not Berkeley was right in believing so. There is much to be said for Leibniz’s view. Certainly Dretske is one in a long list of contemporary philosophers who believe that unconscious perceptions are both possible and actual. Still, in this work we are concerned mainly with conscious perceptions. Insensible perception will only become a topic in Chapter Six, where we will also take a new look at phenomenological contemporaneity.

46

This conclusion, I remark, is actually false for those cases where the temporal interval is too

small to be humanly perceptible. Science, as I’ve just argued, has not shown what Dretske

claims it has shown; i.e., it has not shown that all the temporal differences that scientists can

determine to obtain are relevant to what humans can perceive to exist or not to exist. Still, at this

point it seems that Dretske actually endorses the general conclusion (for all cases whatsoever),

because he goes on to say, disparagingly:

Some philosophers have resorted to heroic methods to avoid this conclusion. They have

suggested, for instance, that ‘seeing D’ is an activity which begins at the moment when

the light rays (which eventually enter our eyes) are reflected from D.43 That is, the

concept of ‘seeing’ is stretched far enough backward in time to ensure the existence of

the objects which we se—at least they must exist when we ‘begin’ to see them. I call this

method ‘heroic’ because it seems to be an outright sacrifice of common sense in order to

save common sense. For if we adopt this alternative, we must say that we begin to see

the star many years before we (as we ordinarily say) see it. In fact, we may begin to see

the stars before we are born (or even conceived) and finish seeing them when we are

four years old.44 There is nothing preventing one from talking this way, I suppose, but I

do not think one will be talking about ‘seeing something’ as this is ordinarily

understood. For I think it is part of what we mean when we say of S that he saw D that

he began to see D, or he first saw D, when and only when D began to look some way to

S or, to put it in terms of the causal analysis, only when the causal sequence first reached

the terminal stage, en.“

I fully agree with Dretske on this criticism of Hirst.45 He then continues as follows:

I think we must conclude, then, that if someone builds into the notion of ‘direct awareness’

the idea that the object of which we are directly aware must (logically must) exist at the time

we are directly aware of it, then (given the finite velocity of light and the transmission of

impulses) we are not directly aware of such things as coffee pots when, as we ordinarily say,

we see them. I do not see how this conclusion can be avoided.

43 Here Dretske refers in a footnote to R. J. Hirst’s The Problems of Perception, p. 308. 44 Here Dretske, in another footnote, says that “Hirst concedes the oddity of these consequences but shrugs them off with the observation that our ordinary ways of speaking are ‘theoretically very unsatisfactory’.” 45 The same criticism is leveled against Hirst by C. W. K. Mundle in his 1971 book Perception: Facts and Theories. On p. 44 he first quotes Hirst’s statement (from p. 133 of The Problems of Perception) that “[w]e must either say that seeing or hearing the object is the whole causal process or that it is the end-stage of it.” Then Mundle observes that “Hirst chooses the former alternative. This commits him, however, to some embarrassing conclusions. For example, that my seeing the sun takes about eight minutes, the time taken for light to travel from sun to earth. Worse still, Hirst would have to say that my seeing a distant star takes a much longer time than I have lived. Russell’s statement that we cannot suppose that the last effect jumps back to its starting point, has special force here.”

47

Well, again, this conclusion can (and must!) be avoided by noting that coffee pots ordinarily

are physically too close to us for them to be able to cease to exist before our seeing of them

ceases. Thereupon, however, Dretske avows surprise in a way that is surprising to me:

But proponents of this view [i.e., that the object of which we are directly aware must

(logically must) exist at the time we are directly aware of it] have gone on to draw a

much stronger conclusion than the one I have just conceded, a conclusion that is totally

unwarranted by anything that has already been admitted. They have gone on to draw the

conclusion that we are directly aware of something when we see a coffee pot. Where

does this conclusion come from?

This, I think, is the “second mistake” he had in mind in the lengthy quote cited in section 2

above. But before I say where I think this conclusion comes from and why it is fully warranted

(and also why I find his surprise surprising), let me quote Dretske still further:

Consider an analogy. We all know what is involved in receiving a message from D. D

may signal us with a flag, send us a letter, call us on the telephone, or speak to us on the

radio. Suppose, now, that I define a new notion, ‘directly receiving a message from D’,

such that this is only possible on the condition that D must (logically must) exist at the

time I directly receive his message. It follows immediately that I do not directly receive

messages, either by radio, telephone, or letter, from distant friends. In each case my

friend could have ceased to exist before I received his message. Of course, if someone

hands me his message in person, we might say that I received his message directly. But

in cases where I did not receive D’s message directly, does it follow that there is

someone or something from whom I did receive a message directly? Is there anything to

which I must be related by this newly defined relation? I do not see why there should be.

When the postman hands me the letter from my distant cousin, I am not directly

receiving a message from my cousin. From whom am I directly receiving a message?

The postman? But if we say this, then we are completely altering the original meaning of

‘receiving a message from’. Originally this meant that the person from whom I received

the message composed it; it is now being taken to mean ‘the person who hands me the

composed message’. Or suppose Martians signal us by radio. We are not directly

receiving a message from them. From who or what are we directly receiving a message?

From the radio waves which eventually reach us? If so, and I do not know of any other

suitable candidate for ‘the thing which must exist at the time we receive the message’,

then we are interpreting the phrase ‘receiving a message from’ in such an eccentric way

that it is positively misleading to continue using it in the phrase ‘directly receiving a

message from’. For we are now interpreting the latter phrase in such a way that we can

48

be related by it to such things as electromagnetic waves. In short, we can directly receive

messages from things which it does not even make sense to suppose we received a

message from (in anything like the original sense of these words).

But to what, we might ask, is receiving a message from D supposed to be analogous? Dretske

has been discussing S’s seeing a coffee pot; but is receiving a message from D really analogous

to S’s seeing something? Well, in Berkeley’s view it is: he thinks that when S sees the coffee

pot, it is God who has put ideas of the coffee pot into S’s mind as if He were sending a message

to S (we will return to this point in a later chapter). But that’s not Dretske’s thought at all; he

shows no inclination to believe that our seeing something is due to God’s sending us messages.

Of course, there may be other ways in which receiving messages is analogous to seeing but

Dretske offers no hint, in this text, as to what they might be. Therefore his supposed analogy

seems pointless.46

Furthermore, the difference between somebody’s handing me his message in person and his

having somebody else do it is perfectly obvious, as is the naturalness of saying that in the former

case I received the message directly and in the latter I did so indirectly: Dretske is without a

doubt right here. But if someone introduces a new notion, “directly receiving a message from

D,” such that this is only possible on the condition that D must exist at the time someone

directly receives his message, then, as Dretske again correctly implies, this would mean that

quite a few messages are only indirectly received without any of those being directly received at

all.

Yet how is all of this supposed to be relevant to the issue of whether the coffee pot (or

anything else) is perceived, whether directly or indirectly?

First of all, Dretske is wrong in suggesting that the contemporaneity of perception and thing

perceived is what marks off direct perception from perception as such (perception simpliciter).

Of course, one may stipulate that “S directly perceives D =def S perceives D and D exists at the

46 In response to my objection Dretske wrote to me as follows: “The point of my mail example was to agree that something must exist at the time you get a letter from x (even if x no longer exists), but the thing that exists at this later time need not be something from which you are receiving a letter. You need not stand to this proximal event (mailman dropping a letter in your mailbox) in the relation (or anything like the relation) you stand to x. Likewise, I agree that when the star appears to you at a time the star no longer exists something must exist (in your head) to explain your experiencing the star as still existing, but this needn't be something of which you are aware. You say it must. What is the argument? It seems to me you beg the question by saying something must look some way to you and it can't be the star. Why not?” I say it can’t be the star because the star no longer exists; it’s no longer there to be looking some way to me; what’s looking some way to me (indeed looking to me as though there were a star there) is a sense-datum, of which I am aware by virtue of being aware of its looking to me as though there were a star there (although I might not be aware of it as being a sense-datum; I’m just aware of the sense-datum); and I must be aware of this sense-datum in order to explain why I have an experience as of the star as still existing rather than not experiencing anything (relevant) at all or just thinking of the star or just imagining it or just remembering my seeing of it. And this sense-datum is probably not in my head, but it certainly is in my mind.

49

time S perceives D.” Under this definition, the vast majority of things perceived would indeed

trivially turn out to be directly perceived. But this is not at all the way that philosophers who

have found the concept of direct perception useful have defined the term. What, after all, would

be the point of such a stipulative definition? It seems entirely unmotivated.

For these philosophers, that things directly perceived exist at the time they are perceived is

not a matter of arbitrarily defining “directly perceive” in the way just described. Rather, they

have been led to distinguish between direct and indirect perception in order to solve certain

problems arising in the philosophy of perception. For this reason, beginning with Locke,47 they

have suggested and/or constructed conceptions of direct or immediate perception so as to

distinguish either two types of knowledge, or two ways of getting to know (the same or different

kinds of) things. One of these conceptions has been developed in a Berkeleian, the other in a

Russellian, way. Both ways, though distinct, crucially involve the claim that what makes

perception indirect is some kind of mediation. This, and not the contemporaneity of perceiving

and perceived, is the crux of the matter. We will discuss this issue in detail in 3.5.

2.5. Dretske and Direct Awareness (II)

Secondly, Dretske’s wrong turn becomes fully evident in the next two paragraphs of his text.

Here’s the first:

If the term ‘awareness’ (‘sees’ or ‘perceives’) is to have anything like its normal sense,

anything like the sense it has in ‘S is aware of (sees or perceives) the coffee pot’, then it

does not follow that S must be directly aware of something or see something directly

when he sees the coffee pot. If one wishes to insist that he must be directly aware of

something when he sees the coffee pot, then one must be prepared to admit that the

statement ‘S is directly aware of D’ does not imply that D is even the sort of thing which

it makes sense to suppose someone aware of (in the original sense of this term). D need

not be the sort of thing which can be colored, or even look colored—anymore than radio

waves need be the sort of thing which can compose messages. The only condition D

must satisfy is that D must necessarily exist when we (as we ordinarily say) see the

coffee pot. And if we cast around for what this might be, it is not difficult to find the

natural candidate: it is simply the coffee pot’s looking some way to S.48 For this state of

47 See An Essay Concerning Human Understanding, Book IV, Chapter IV, Section 3. 48 The locution of D’s looking some way to S is introduced by Dretske on pp. 20-21 of his text when he construes “S seesn D” as “D is visually differentiated from its immediate environment by S” and then says that the “phrase ‘visually differentiated’ is meant to suggest” that “S’s differentiation of D is constituted by D’s looking some way to S and, moreover, looking different than its immediate environment.” He goes on to explain that when “I say that D

50

affairs is necessarily involved in S’s seeing the coffee pot, it is simultaneous with his

seeing the coffee pot (indeed, it is just an expression for the terminal stage of the causal

sequence-- en), and it can occur when the coffee pot no longer exists. This, in fact, is

precisely what the scientific facts show us: that the coffee pot can continue to look some

way to us after it has ceased to exist. The fact that the coffee pot’s looking some way to

S does not sound like the sort of thing of which S can be aware (it certainly is not colored

or shaped like a coffee pot) should not disturb us; for there is nothing which should lead

us to believe that being directly aware of this sort of thing is anything like being aware of

something.

This last sentence is disconcerting, in two ways. First of all, if I can be aware of my seeing the

coffee pot (and Dretske is not, or at least shouldn’t be, denying that), then why can't I be aware

of its looking some way to me (even though that way, or its looking so to me, is obviously not

itself colored or shaped in any way)? Surely, in the very process of my coming to be aware that

I’m seeing the coffee pot I can also become aware of how it’s looking to me, can’t I?49

Secondly, the second clause of that sentence seems to me to be an outright howler: how can

being directly aware of anything whatsoever fail to be like being aware of something? How can

looks some way to S, I do not wish this to be understood to mean that, for some character C, it looks to S as though (as if) D (or something) were C. The locution ‘looks as though (or as if) it were’ usually signifies that the percipient believes, or is inclined to believe, or is prepared to cautiously put it forward that, what he sees is a certain sort of thing or possesses a certain property. That is, it suggests or implies something about the belief attitude of the percipient, and I want the construction ‘D looks some way to S’ to be free of this implication. D can look some way to S without its looking to S as though it were C (for any C). . . . As I am using the phrase, D’s looking some way to S presupposes a sentient being (S) equipped with an appropriate visual apparatus by virtue of which, to employ an expression of the psychologists, D occupies a portion of S’s visual field. It presupposes or entails nothing about whether S notices D, whether he takes, or is inclined to take, D to be something in particular, or whether he exploits his visual experience in any way whatsoever.” But even though D can look some way to S without its looking to S as though it were C, Dretske should not exclude the possibility that D can look some way to S together with its also looking to S as though it were C. In the cases of conscious human adult perception discussed here this is often the case. 49 When I called Dretske on this, he answered as follows: “When I said that the coffee pot's looking some way to us was not something we are aware of I meant "aware of" in the way we are aware of objects and events (coffee pots, sunsets, shadows, trees, etc.)--things that, when we are (visually) aware of them, look some way to us. I don't think something's looking some way to us looks some way to us. We are, of course, aware of a thing's looking some way to us in the factive sense of "aware"; we are aware (of the fact) that it looks some way to us. But we can be aware that something looks some way to us (this fact) without being aware of the state of affairs--something's looking some way to us--that constitutes this fact just as I can be aware of (the fact) that I see a tree without being aware of (the state of affairs) my seeing the tree. I am aware (of the fact) that I experience things without being aware of my experience of things (this state of affairs).” Even if the last two sentences were true, what I am claiming is not that when I’m aware of seeing a coffee put I must also be aware of its looking some way to me; I’m only claiming that when I’m aware of seeing a coffee pot I can also be aware of its looking some way to me (even though its looking some way to me doesn’t itself look any way to me), and nothing Dretske says addresses this claim. But can I really be aware of a fact without being aware of the state of affairs constituting this fact? Only if the fact, for example, is psychological and the state of affairs is physical involving neural activity. But in that case I would deny that such a physical state of affairs could constitute (rather than just partially explain) the psychological fact. Moreover, if I am aware of the fact that I experience things, I can in a jiffy be made aware of my experience of things (the state of affairs) simply by, e.g., your asking me, Are you aware of your experience of things?

51

being directly aware of something not be like being aware of something? I think that Dretske

must have misspoken himself here.

But even before this last sentence there’s something howling for attention: How long after the

coffee pot ceases to exist can the state of affairs of the coffee pot’s looking some way to me

last? Presumably (Dretske would answer, and reasonably so), for as long as light rays from the

coffee pot are still streaming into my eyes and getting transduced into neural impulses that

accompany, or are identical with, my seeing the coffee pot. But if so, then, it seems to me,

Dretske is well on his way to losing his case. For consider:

In the normal case (of human interaction with coffee pots), at any time t at which I see the

coffee pot, the following statements are true:

(1) the time it takes for light to travel from the coffee pot to my eyes is well below the

threshold of human perceivability;

(2) the coffee pot exists;

(3) I see the coffee pot;

(4) it is possible that I am aware of seeing the coffee pot;

(5) it is possible that the coffee pot looks some way to me;

(6) it is possible that I am aware of its looking some way to me;

(7) if the foregoing three possible states of affairs were actual, then all the

states of affairs from (3) to (6) would be phenomenally simultaneous.

Now in this, the normal, case, it is (again) just false to say that what “the scientific facts show

us” is “that the coffee pot can continue to look some way to us after it has ceased to exist.” The

unvarnished truth is that the coffee pot can continue to look some way to us only for as long as it

exists and not a humanly discernible moment longer.

What can continue to look some way to us after it has ceased to exist is, if Dretske is right,

just some celestial body.50 But is Dretske right? Let’s suppose he is. Let’s also suppose that at

some point during the interval when light rays from the celestial body are still streaming into my

eyes and getting transduced into neural impulses that accompany, or are identical with, my

seeing of it, this celestial body is destroyed. Since Dretske claims that I’m still seeing the

celestial body, this can only be the body as it was many years ago; in other words, what I see,

according to him, is a long-past state of the no-longer-existing heavenly body. But can

something that now no longer exists look some way to me now? Can my two-storey, six-

bedroom-room house that burned to the ground last year, or even yesterday, look some way to

me today? What would it even mean to say, not that the cinders or ashes left behind, but that the

50 Or, at least, something very much further away from, and bigger than, us than visible coffee pots usually are; not even the most distant mountain would fit the bill here.

52

house itself looks some way to me now? Admittedly, there is no light reflected from the house

that now reaches my eyes, but that only explains why the house isn’t looking some way to me

now; it doesn’t explain what it would mean to say the house is looking some way to me now, if

it were looking some way to me now. Of course, it isn’t, and it can’t be. I can only have

thoughts and memories about how it was looking to me when it still existed; I can also imagine

what it would look to me now if it still existed; and I can imagine how it would look to me if I

were looking at it now; but it, the no-longer-existent thing, cannot now actually look any way to

me at all (in the way that anything that does now exist is looking to me now).

So why does Dretske think that the no-longer-existent star is looking some way to me now?

Obviously, it’s because (1) there’s still light from it being received by my eyes, as a result of

which (2) it appears to me as if I’m seeing that star. These two claims are something both

Dretske and I can agree on. But Dretske thinks (3) I am seeing the star (even though it no longer

exists) and therefore (4) the star is also looking some way to me, while I deny both (3) and (4).

Instead of (3) I claim that (5) I am seeing a sense-datum of the star (what Berkeley would call a

visual star) and instead of (4) I claim that (6) the sense-datum is looking like a star to me.

So which one of us is right?

At this point I would challenge Dretske thus. What can he (or indeed anyone else) mean in

saying the star is looking some way to us now? In our e-mail correspondence I had posed a

related question to him as follows:

“How can the coffee pot continue to look some way to us after it no longer exists? Isn't

it true that if the coffee pot looks some way to us, there must be some quality (say,

brownness) it looks to us to have? And isn’t it the case that its looking some way to us

just is its having some quality it looks to us to have? But if so, how can it look to us to

have a quality if it no longer exists, if it isn't there to have that quality or even just to

look to have it?”

To this Dretske responded:

“I admit it sounds a little funny. But just a little. I am willing to concede that x's looking

some way to us is just for there to be some quality x looks to have. And what the

scientific facts show us is that something can look to have a quality when it no longer

exists. But you slip from this into the more suspicious description “its having some

quality it looks to us to have.” I admit that it sounds much more strange to speak of

something actually having a quality (it looks to us to have) when it doesn't exist, but I

would resist the inference from:

1. x looks to us to have quality Q

53

to

2. x has a quality (viz., Q) it looks to us to have.

That looks like you are moving outside the scope of “looks” something that only belongs

inside the scope of this operator.”

To which I replied as follows:

“Since you admit that “it sounds much more strange to speak of something actually

having a quality (it looks to us to have) when it doesn't exist” and therefore resist this

inference, may I take it that when you say that “what the scientific facts show us is that

something can look to have a quality when it no longer exists,” you mean to imply

that the no-longer-existent coffee pot merely looks to have, but doesn't actually have,

that quality? Am I reading you correctly? If so, then at least you avoid what I would call

the contradiction of claiming that something can actually continue to have a sensible

quality after it has ceased to exist.”

Dretske answered that indeed this is what he meant. But in that case we must be careful to

observe that in the contexts at hand “looks” can either mean “seems” or “visually appears.”

When Dretske agrees that the no-longer-existent coffee pot merely looks to have, but doesn't

actually have, the quality of brownness, I think he has in mind “looks” as “seems.” But when he

claims that everything we see looks some way to us he I think has in mind “looking” as

“visually appearing.” Therefore I would strongly press the following objection: How can any

no-longer-existing thing so much as even merely look (=seem) to have a property, when that

property is a sensible (visually, tactually, aurally, olfactorily, gustatorily, kinaesthetically

apparent) one such as is necessarily present in that thing’s looking (=visually appearing) some

way to us, which is what Dretske says the star (really, not just seemingly) does when we see it?

It’s just evident, I would claim, that any star seen by human beings actually does have

qualities, one of which is looking (=visually appearing) to be a silvery luminescence, a pinprick

of light. Hence, if Dretske says we can see a no-longer-existent star, then he must say it has that

quality, too; so he cannot after all avoid (though he wants to) the conclusion that no-longer-

existent objects have sensible qualities. Consequently, since Dretske does say we can see a star

that no longer exists, he must also say that that non-existent star can still have sensible qualities

which we now see, even if it is only the property of looking to be a pinprick of light.

And now for the second of those two crucial paragraphs I mentioned several pages back:

I conclude, then, that if someone wants to define a sense of ‘direct awareness’

(‘immediate acquaintance’, ‘sensing’) which has built into it the idea that the elements of

which we are directly aware must exist at the time we are directly aware of them, then, in

those cases which are ordinarily described by saying that we see a physical object, D,

54

either (i) there is nothing of which we are directly aware or (ii) the meaning of the term

‘aware’ has been so shifted that the phrase ‘S is directly aware of A’ is but an alternative

way of describing the state of affairs which I have expressed by saying that D looks

some way to S. And there is nothing in all of this which should suggest the conclusion

that when we see a coffee pot, this visual achievement is mediated by our direct

awareness of something other than a coffee pot.

It seems to me that the first alternative Dretske mentions, namely, that “(i) there is nothing

of which we are directly aware,” is simply unavailable here. Certainly, in most such cases,

whenever we’re talking or thinking about coffee pots, or practically any other physical object in

front of our eyes, they must exist, and we are aware of them, when we see them. This is just to

repeat what was already said. Thus Dretske, I think, is just wrong on this point.

It also strikes me that in virtue of leaving open alternative (ii) Dretske is really

acknowledging, back-handedly, the truth of the sense-datum theory. For what is S’s having a

sense-datum of D if not yet another way of D’s looking some way to S? In effect, the latter

seems to be, in crucial ways, the functional equivalent, in Dretske’s account, of the former, in

the sense-datum theorist’s account. Just as Dretske wants D’s looking some way to S to be able

to last even after D no longer exists, so the sense-datum theorist wants S’s having a sense-

datum of D to be able to occur, in accordance with the TGA, even after D ceases to exist. And

just as D’s looking some way to S allows Dretske to provide for the possibility that S can see a

star that no longer exists, so S’s having a sense-datum of D allows the sense-datum theorist to

provide for the possibility that it appears to S as though he were seeing a star even when he is

not. Of course, the sense-datum theorist won’t think Dretske can really provide for the

possibility that S can see a star that no longer exists; what D’s looking some way to S allows

Dretske to provide for (in the sense-datum theorist’s eyes) is only its looking to S as though he

were seeing a star. From the latter Dretske then draws the conclusion that S, in effect, sees the

star directly, whereas the sense-datum theorist concludes that S directly sees a sense-datum and

thereby either (a) directly or indirectly sees the star, or (b) has evidence that there was a star to

be seen in the past.

This difference connects with the fact that in the sense-datum theorist’s account, S’s having

a sense-datum of D has a mediating function that D’s looking some way to S doesn’t have, in

Dretske’s account. For Dretske, the perception of all objects is direct (=immediate); thus he

finds no real application for the concept of indirect (= mediate) perception. The first time

Dretske explicitly mentions mediation—if only to dismiss it—is precisely in the last-quoted

sentence. But he can dismiss it only because (to repeat) visual awareness of the coffee pot is not

a good example for illustrating the difference between indirect and direct awareness. The sense-

55

datum theorist, of course, claims that both types of awareness occur in the coffee pot case,

although proving this requires something like the Berkeleian version of the TGA.

Now, at last, something else falls into place. Earlier in this section I quoted Dretske as

saying that the proponents of the view that the object of which we are directly aware must

(logically must) exist at the time we are directly aware of it “have gone on to draw the

conclusion that we are directly aware of something when we see a coffee pot. Where does this

conclusion come from?”

Well, it should be clear now that it comes from the fact that the coffee pot, if we see it and

are aware of it, must (logically must) exist at the time we are directly aware of it. This

conclusion follows from both the existence-requirement and the (by now overly familiar) fact

that a humanly visible coffee pot cannot go out of existence while some humans are still seeing

it. It should also be clear that if we change “coffee pot” to “sun” and go back to the Berkeleian

TGA of Chapter One, then an analogous conclusion comes from the fact that if it looks to

us now (at any time falling within the eight-minute interval between the time the sun went out

of existence and the time the last light rays from it reached our eyes) exactly as if the sun were

still shining there, then we are visually aware of something; but given that the sun no longer is

shining there, what we must be aware of is the sun’s sense-datum. All this is perfectly

reasonable; and what is surprising is that Dretske seems to be baffled by this (as shown by his

rhetorically asking, Where does this conclusion come from?).

Another philosopher who, like Dretske, thinks that a no-longer-existent thing or state could

be perceived and who wrote on this topic at roughly the same time, was Anthony Quinton. Thus

before completing this segment of my discussion of Dretske I would like to consider some

relevant passages of Quinton’s.

2.6. Quinton on the TGA

Quinton writes that “(i)f an astronomer says ‘there is an explosion at the north-west corner of

the sun at the moment’ what he says is false, or at best an accidentally true prediction about an

explosion he will be in a position to see in eight minutes' time but is not the explosion he takes

himself to perceive”.51 Let us assume the astronomer is speaking sincerely and Quinton is

speaking the truth. So the explosion the astronomer now takes himself to see does not exist, i.e.,

there just is no such explosion occurring there now. But why does this astronomer believe what

he says? Why does he believe that there is indeed an explosion occurring at the north-west

51 Quinton, The Nature of Things, 200.

56

corner of the sun now, when no such event is actually occurring there? Why does he take

himself to perceive an explosion there when there is no explosion to be perceived, no explosion

occurring there at all? He is not (I assume) hallucinating nor is he enjoying a clairvoyant vision.

Hence it must be because it seems to him as though there were an explosion occurring there

now. But why is that so? Why does it seem to him as though there were an explosion occurring

there now? Given that he's neither hallucinating nor being clairvoyant and that there's no real

explosion to be seen, mustn't it be because he is now aware of a proxy explosion—an explosion-

like sense-datum? How else could one explain what he takes himself to see? Quinton himself

neither raises nor answers this question.52 This, I think, leaves the sense-datum theorist with a

distinct advantage—simply because he does raise and answer this question.

A little later Quinton writes:

“(i)n seeing the sun I see a roughly spherical, luminous object lying in the heavens in a

certain direction from where I am standing. I also think I perceive, but do not, that this

thing has a certain temporal property, that of being contemporaneous with my act of

seeing. Though I am right to believe that there is such a thing out there this is not

justified by that perception. Only inference can justify what I correctly believe on the

basis of a misperception. But I do still see the sun, even if it is the sun of eight minutes

ago.”53

What is the misperception here? Well, it seems to be this: that the “roughly spherical,

luminous object lying in the heavens in a certain direction from where I am standing … has a

certain temporal property, that of being contemporaneous with my act of seeing.” But why do I

think (falsely, in Quinton’s opinion) that I perceive that it has this property? For it, the sun of

eight minutes ago, obviously doesn’t have it: while my seeing occurs now (at time t), the sun of

eight minutes ago (call it the sun-at-t-8) ceased to exist when that minute ended. So the sun-

stage that I am now seeing no longer exists: what I’m now seeing is a sun-stage of the past.

Therefore, the misperception is just this: seeing what I now see as existing now. I see the past

but mistakenly think it is the present. But what explains this mistake? Why do I make it? What

is it that causes me to think wrongly that what is actually past is really present? These questions,

again, are ones that Quinton does not raise or attempt to answer. But they seem to me to be

good, legitimate questions deserving of an answer.

52In discussing Quinton‘s treatment of an analogous situation (one in which a white wall that is seen appears yellow), David Smith says that “Quinton (together with legions of other philosophers who have followed him in this) has no answer at all” to the question of “what it is about the nature of a person’s perceptual experience that inclines” him to believe he’s seeing a yellow wall (Smith, The Problem of Perception, 39). 53 Quinton, The Nature of Things, 202.

57

Let us take this more slowly again just to make sure that the answer to these questions isn’t

something that’s already implicit in Quinton’s text. If seeing what he now sees as existing now

is a misperception, then is there something that Quinton is right about? He thinks he is right

about there being “a roughly spherical, luminous object lying in the heavens in a certain

direction from where” he is “standing.” Nevertheless, this correct belief of his is, in his own

words, “not justified by” what he perceives. Well, what does he perceive? He thinks he

perceives that there is this spherical, luminous object out there in the heavens, and he

erroneously thinks that he perceives—that is to say, he actually misperceives—that this object is

contemporaneous with his perceiving. It isn’t. What is contemporaneous with his perceiving is

the sun-at-t; but he can't be seeing this sun-stage as yet, because at t, when he is supposed to be

seeing a spherical, luminous object in the sky, this sun-stage has not yet emitted any light to his

eyes. It will do this, at best, only eight minutes after t.

So what does Quinton see? He sees a sun-stage, the sun-at-t-8. And what is this correct belief

of his? What is he right about? He thinks he is “right to believe that there is such a thing out

there.” But which thing is that? The sun-at-t-8 was, but no longer is, out there. What is out there

is the sun-at-t, but Quinton can't be seeing this sun-stage (as yet). What he does see, he thinks, is

the sun-at-t-8, and he thinks that he is right to believe that there still is the sun out there at t; but

this belief, he says, is not justified by his perception of the sun-at-t-8. So what does justify this

correct belief of his that there is a sun out there at t?

Recall what he says: “Only inference can justify what I correctly believe on the basis of a

misperception.” He doesn’t spell out this inference so we will try to do it for him. What we have

to work with, Quinton suggests, is three elements: a misperception, an inferential premise, and a

correct belief. The misperception (the basis for the inference) is that the thing he perceives (the

sun-at-t-8) is actually contemporaneous with his perception (at t) of it. The inferential premise is

presumably that it is highly unlikely that the sun went out of existence at some moment during

the last eight minutes. The correct belief is that at t the sun is still out there. So this inference is a

probable (=probably correct) inference from the past to the present.

This is as explicit a reconstruction of Quinton’s passage as I can manage. But still it yields no

answers to our questions about the misperception: When I see the past but mistakenly think it to

be the present, what explains this mistake? Why do I make it? What is it that causes me to think

wrongly that what is actually past is really present? The sense-datum theorist, of course, claims

that seeing what I now see as existing now involves, in one respect, no misperception and no

mistake at all; hence these questions don’t even arise for him. But they do for Quinton and he

fails to answer them. This again leaves the sense-datum theorist in a better position.

58

Now the respect in which there is no misperception for the sense-datum theorist is the

following: When he claims that seeing what I now see as existing now involves no mistake he

has in mind direct (immediate) perception only: in now seeing a spherical, luminous object that

exists now I am directly seeing a solar sense-datum which is not a sun-stage though it is related

to one, either as caused by it or as part of it. And the respect in which there may be a

misperception for the sense-datum theorist is that he may think that in addition to now seeing a

solar sense-datum he is also now seeing the star itself.

But suppose now the logically possible but improbable happened; suppose the sun

(mysteriously) did go out of existence five minutes ago. Then Quinton’s belief would be false; it

would not be the case that, at t, there was a “roughly spherical, luminous object” out there. So in

that case what would he say? Not knowing that the improbable happened, he would no doubt

continue saying that, at t, he saw the sun (despite the fact that, unbeknownst to him, it no longer

existed) and that it was the sun-at-t-8. Thus he would, again, be subject to a misperception (the

same misperception that the roughly spherical, luminous object he thinks he sees is

contemporaneous with his act of seeing); but, in addition, he would also be making a terribly

wrong inference (that there still is that roughly spherical, luminous object shining out there at t.)

Yet would he be right about anything? Yes, he would be right in thinking that the sun-at-t-8

existed. But what would this correct belief of his be based on? He presumably would say it is

based on (what he takes to be) the fact that he is directly seeing it; the sense-datum theorist

would say it is based on a highly probable inference to the past from the fact that he’s now

seeing a sense-datum of the star.

Fortunately, the improbable event just described never happens; but something similar often

transpires in the case of vastly more distant objects, such as the stars. So that when I, especially

at night, look out and look above me, some of the things that I see are still there when I see

them; but others probably are not. The light on my neighbor's porch is still on when I see it; but

the pinpricks of light in the sky, the stars—I don’t know: most of them are still “on” (I suppose),

but a few are long dead and gone (or so scientists assure us). The problem is that I (or the

scientists) have no way of knowing, just by looking, which of the stars I see are still extant and

which are dead: they all shine the same!

Moreover, they all seem to be equally in front of my eyes now—together with everything

else that I see. This consequence of the Dretske-Quinton view seems very funny, for two

reasons.

It seems funny, first, because we are told to accept the view that what is in front of our eyes

and is seen by us now may either exist now or not exist now, and that we might not be able to

tell, just by looking, which is the case. (This view is so strange that it could only be accepted if

59

there are exceedingly strong arguments for it, stronger than those that Dretske or Quinton have

provided.)

And it seems funny, second, because it doesn’t sit well with what Dretske already has taught

us about there being just one sense of the verb “to see” and its being governed by the existence

condition (see 2.1.). According to this doctrine, it’s just not possible “with a straight face and in

all sincerity, to truly say that” [at t] “one saw Harold at the scene of the crime” [at t] “without

even believing he was there” [at t], “and indeed without his having been there” [at t].54 But now

we seem to be told that it is possible “with a straight face and in all sincerity, to truly say that

one saw”, at t, the star in the sky, “without even believing” it was there at t, “and indeed

without” its “having been there” at t.

Why? Why does saying that at t I saw Harold at the scene of the crime entail that he was

there at t, while saying that at t I saw the star in the sky does not entail that it was there at t?

What’s the big difference between Harold’s being seen at the scene of the crime and the star’s

being seen in the sky?

It might seem that, for Dretske, the answer is obvious. But discussing it in Chapter Four will

lead us to wonder whether the existence-requirement for perception doesn’t straightaway imply

the simultaneity-requirement, after all.

2.7. Armstrong on the TGA

Writing just ahead of Dretske and Quinton there was another important (indeed, path-breaking)

Materialist philosopher who also denied the simultaneity-requirement but valiantly tried to face

the questions they either handled inadequately or totally ignored. David Armstrong, like

Dretske, discusses the TGA extensively and feels its force; in addition, as a great Berkeley

scholar he is fully sensitive to the pressure to acknowledge the existence of “ideas.” Though he

doesn’t succumb to it, he comes close (though not close enough) to admitting sense-data by

being prepared to talk of sense-impressions. We describe “our visual sense-impressions,” he

says, when “we speak about what we seem to see, in the phenomenological sense of the phrase,

but we restrict ourselves to the immediately seen.”55 And seeing immediately is “nothing but the

acquiring of immediate knowledge of, or inclination to believe in, facts about the physical

world, by means of the senses.”56

54 The bracketed expressions are not in Dretske’s text, but I assume they are consistent with what he has in mind there. 55 Armstrong, Bodily Sensations, 6. 56 Armstrong, Perception and the Physical World, 148.

60

In some ways Armstrong’s approach to the TGA is superior to both Dretske’s and Quinton’s.

It is superior to Dretske’s in that, like Berkeley, Armstrong is ready to differentiate between

immediate and mediate perception, and even more importantly, he isn’t baffled, the way Dretske

is, by the pressure to admit that in the case of the no-longer-existent heavenly body there is

something we seem to see immediately, something other than the heavenly body. And it is much

superior to Quinton’s in that it not only tries to answer the familiar, obvious questions but also

raises, and grapples with, some new ones.

It’s well worth quoting at length what Armstrong says on the TGA in his Perception and the

Physical World:

It is logically possible that perception of an event should occur at exactly the same

moment as the event itself occurs. Descartes actually thought that this did happen. He

thought that a rod-like movement originated in the object, which produced a movement

in the brain absolutely simultaneously. At the same instant that this movement in the

brain occurred, the object was perceived. But we know now that in every case a time

must elapse (sometimes a very short one) before a particular state of affairs is perceived.

In some cases the time-gap is appreciable, when we hear distant thunder it is a few

seconds, when we see the sun it is eight minutes, when we see the stars it is many years.

But in every case there is some time-gap.

But now one may begin to wonder how we can perceive now what happened then.

How can we cross the time-gap, and get to the past? The past is over and gone, and how

can we see what is over and gone? Must not perception take as its object what is

happening now? Under the influence of this line of thought one may be driven over once

again to the view that the only thing that can be immediately perceived is something that

occurs now, for example, our present sense-impressions. And then we have once again

accepted some form of Representative theory, with all its difficulties.57

Not wishing to be forced into Representative Realism (let alone Idealism, which, as a

Materialist, he seems to consider to be completely beyond the pale), Armstrong must reject the

TGA. He says that “a complete emancipation from this argument is not very easy, so we shall

proceed by stages”; and in the first stage he mainly distinguishes between physical and

phenomenological simultaneity, much as we have done in section 2.3. The conclusion of this

stage is that the world being as it is, “we are justified in saying that, in most ordinary

perception, there is no time-gap to worry about.” This is exactly our contention. The difficult

cases “are comparatively small in number (seeing the sun or a star, hearing distant thunder).” It

57 Armstrong, Perception and the Physical World, 143. Further separately unannotated quotations are from pp. 143-152 of this work.

61

is just with respect to these visual cases that the TGA comes into play; and here Armstrong

argues as follows:

“If we confine ourselves to what seems to be immediately perceived, that is, if we

confine ourselves to our sense-impressions, we must admit that our impression is always

of something that exists in the present. If an astronomer sees a star vanish, there is no

doubt that he perceives the past. And he may perceive it as past. But his sense-

impressions are impressions of something happening now: a light seems to go out in the

sky now. If our analysis of perception is correct, he will have an inclination to believe

falsely that something is happening up above him in the sky now. His immediate

perception involves illusion about time, even if about nothing else, and it is only his

mediate perception that is veridical. So it seems that we cannot have immediate

perception of the past as past. Our immediate perception of past happenings must

involve illusion, illusion about the time of the happening immediately perceived.

In the case of the star, it may be questioned whether our immediate perception really

involves any temporal illusion. It may be suggested that what we immediately perceive

is not the star, but a present happening, causally connected with the extinction of the star

many years ago. The star sends a message to us, as it were, and we immediately perceive

the message, not the star.

The foregoing two paragraphs are something that a sense-datum philosopher could assent to

without reservation. But in the next paragraph Armstrong starts to wonder what it is that we

immediately see in this case; and this is where the problems start:

What could the immediate objects of sight be? They could not be our sense-impressions,

for we cannot perceive sense-impressions. They could not be the light-waves from the

star, for we do not see light-waves. The only possible immediate object of sight is the

star itself. But then we must admit that our perception involves temporal illusion,

because the star's extinction appears to be occurring now.

It is exactly at the second sentence that the sense-datum philosopher will part company with

Armstrong. He will say that star-like sense-impressions (a.k.a. sense-data) is precisely what we

are visually aware of (even if we do not ordinarily say we perceive them); and he will ask, What

is the argument for saying we cannot be aware of them? Armstrong doesn’t give any in this

section; he might have done so elsewhere in Perception and the Physical World, to which,

unfortunately, I couldn’t gain access in its entirety. However, in another book written

immediately after this one Armstrong warns us that “we must not be led by the substantive

62

‘sense-impression’ into the unthinking assumption that sense-impressions are a sort of object.”58

To this the sense-datum theorist will surely respond, How could we fail to be?, and add that far

from being unthinking, this assumption is a very natural one to make. I assume that Armstrong

might have some of the usual objections to treating sense-data as objects, of which there are

many; I think they are sufficiently well disposed of in the works of Robinson, Foster, and

O’Shaughnessy.59

The next sentence is surely right; but it prompts the obvious question, Why couldn’t the

immediate object of sight be light from the star? Surely we can see light (rather than light-

waves), can’t we? This is just what Maurice Mandelbaum suggested in his book on sense-

perception;60 what Len Carrier asserts in his own treatment of the TGA;61 and what John Dewey

claimed in his 1916 lecture on “Naïve Realism vs. Presentative Realism.”62 Brian

O’Shaughnessy, even more insistently, argued that “we always see light” and “in a certain sense

only light” and that “we must see light whenever we see physical objects.”63 We will engage

these points in detail in 3.5.3., 4.6., and 4.7.

The third and fourth sentences suggest the fascinating possibility (not explicitly considered

by Armstrong) that the star isn’t extinct after all, that it still exists, that it isn’t “straightforwardly

past” (as we’ve already seen David Lewis putting it; we’ll again discuss this possibility in 3.3.

and Chapter Six). What Armstrong settles on instead is the view that we can and do see past

things and events, as he explains in the next paragraph:

So it seems that immediate perception is sometimes a perception of past happenings.

But, at the same time, the past cannot be immediately [my italics] perceived as past. All

perception of the past as past is mediate [my italics] perception. There is no perception

of a time-gap in immediate perception.”

In other words (to be entirely clear about this), we now immediately perceive the star as it was

but no longer is, we perceive a past state of the now-extinct star, but we do not immediately

perceive it as past: this we perceive only mediately. Immediately we perceive only what was;

but we falsely think we perceive what is now; we are subject to an illusion. Then Armstrong

58 Armstrong, Bodily Sensations, 6. 59 The sense-datum theorist, of course, would also contest Armstrong’s claim that perception is nothing but the acquiring of knowledge of, or inclination to believe in, facts about the physical world, by means of the senses. This view, he would say, makes sense-perception too much like cogitation (thinking) and assimilates sense experience too much to intellectual activity. More importantly, it thereby seems altogether to leave out the object perceived (whether immediately or mediately) and to leave unexplained how perceptual contact with it is made. 60 M. Mandelbaum, Philosophy, Science, and Sense Perception (Baltimore: The John Hopkins Press, 1964): 177-179. 61 L. S. Carrier, “The Time-Gap Argument,” Australasian Journal of Philosophy, Vol. 47, No.3, Dec. 1969: 263-272 and “Time-Gap Myopia,” Australasian Journal of Philosophy, Vol. 50, No. 1, May, 1972: 55-57. 62 John Dewey. "Naive Realism vs Presentative Realism" Chapter 9 in Essays in Experimental Logic. Chicago: University of Chicago (1916): 257. 63 Brian O’Shaughnessy, Consciousness and the World, 441.

63

revealingly admits the following: “Ironically, however, this absence of a time-gap in immediate

perception gives rise to two serious problems.” The first goes as follows:

We must admit that the notion of perceptual illusion is co-ordinate with the notion of

veridical perception. When we speak of immediate perceptual illusion, it should at least

make sense to speak of immediately perceiving the true state of affairs. When I press my

eye-ball, the candle-flame looks double. But it is possible for the candle-flame to look as

it is. Now suppose that when somebody perceives a star vanish, he simultaneously

perceives the flight of a bird. It looks to him as if a bright object vanishes from the sky at

the same instant as a thing with a birdlike shape moves across the sky. But, in fact, the

first event occurred long before the second event, and so immediate perceptual illusion is

involved. But if the concept of immediate perceptual illusion is co-ordinate with the

concept of immediate veridical perception, it seems that it should at least make sense to

speak of immediately perceiving the true state of affairs. We should be able to perceive

immediately that the first event occurred long before the second event. But the difficulty

is that we cannot understand what it would be like to perceive this (my italics). I can

understand having the immediate perception that the bright thing in the sky is very much

further away from my body than the object with the bird-like shape. (Of course,

immediate perception only gives me 'very much further away'.) But what would it be like

to perceive immediately that the extinction of the light occurred very much before the

motion of the bird-like object? This is far from clear (my italics).

Armstrong is making two very good points here: one, that if it makes sense to speak of

immediately seeing things the way they are not, it should also make sense to speak of

immediately seeing the very same things the way they are; and two, that it is hard to understand,

and far from clear, what it would be like to see, immediately, one event occurring very much

(centuries, days, or even minutes) before another.

It is indeed admirable that Armstrong admits this difficulty (as far as I know, he is the first

philosopher to have done so). But though he raises it, I can’t see how and where he solves it, in

this text or in any other. Therefore I can’t see him as disposing of the sense-datum account.

As noted above, Armstrong also identifies a second “serious” problem raised by the absence

of a time-gap in immediate perception:

We have put forward the view that immediate perception is nothing but the acquiring of

immediate knowledge of, or inclination to believe in, facts about the physical world, by

means of the senses. Now, if this is so, it seems that immediate perception of the past

should be logically possible. Why should I not acquire immediate information about the

past, knowing it to be the past, by using my senses? If we cannot give sense to the notion

64

of immediate perception of the past as past, it seems that our account of perception may

be in jeopardy. Yet it is not clear what immediate perception of the past as past would be

like.

Again, I agree and again, I don’t see where he provides a solution to this problem. The closest

(and it’s not very close at all) that he comes to suggesting one is at the very end of his

discussion of the TGA, where he says that:

So to talk about looking into the sky, and acquiring, by the use of the senses, immediate

information that something had happened in the past, would be a case for which we have

no linguistic rules ready. We could not call it memory, for memory entails previous

perception. We could not call it mediate perception, for that would be contrary to the

description of the case as the immediate acquiring of information. We could not call it

immediate perception, for we want to say that what is immediately perceived at least

appears to exist in the present. We can certainly imagine that, by use of our senses, we

should immediately acquire knowledge about the past, known as being past. We can

imagine this knowledge being of the same sort as our perceptual knowledge, and just as

detailed. But we should be reluctant to call it perception, although we would have no

other single term to label it by. There is, of course, an excellent reason for this: this sort

of phenomenon does not in fact occur. It is a mere logical possibility, for which our

language has not provided.

What is this “mere logical possibility” that Armstrong is talking about, “this sort of

phenomenon” that “does not in fact occur”? Presumably it is our immediately acquiring by our

senses knowledge about the past, known as being past. Although he admits “it is not clear what

immediate perception of the past as past would be like”, he does claim we “can certainly

imagine” this, yet—and that’s my complaint—he doesn’t explain how we can imagine it.

Nevertheless, he at least prods the reader to attempt to construct, in his imagination, a possible

scenario in which this sort of phenomenon does occur and in which we might immediately

acquire by our senses knowledge about the past, known as being past. Here’s my try at such a

construction:

Let’s suppose (contrary-to-fact) that when a star explodes it turns orange so that during and

after the explosion the light emitted (by what remains of the star) is orange. If this were a

scientific fact and I knew it, then the following would be the case: if I perceived an orange light

in the night sky at the same time as I perceived the flight of a bird, I would immediately know

that a star has exploded, even though “the first event occurred long before the second event.” Of

course, I would immediately know this only in a sense of “immediate” that does not exclude

experience and learning (which are necessary conditions for acquiring scientific knowledge).

65

But can this really be “immediate” knowledge? This is a type of knowledge that Berkeley

extensively treats of in his Essay towards a New Theory of Vision and calls “mediate,” as does

Russell in his own related discussions. What is “immediate” on anybody’s view is my

knowledge that my sense-impression of an orange light was simultaneous with the bird’s flight.

So it is doubtful whether my imagined example meets Armstrong’s needs; whether, that is, the

star’s explosion “in orange” really is immediately perceived as past in just the way my sense-

impression of an orange light is immediately perceived as simultaneous with the flight of the

bird. Consequently, I don’t find support for Armstrong’s claim that we “can certainly imagine

that, by use of our senses, we should immediately acquire knowledge about the past, known as

being past.”

Nonetheless, there is one more clear plus for Armstrong’s account in that he, additionally,

tries to explain why the simultaneity-requirement seems plausible. In the paragraph coming just

before the one I quoted last, he writes:

We have seen that in ordinary cases of immediate perception, in the paradigm cases

where the objects we perceive are at no great spatial distance, the events perceived occur

at the same time as the perception of them. An event that occurred in the past would

have been perceived in the past, and so could only be remembered now. I suggest that

this fact tends to be written into the concept of immediate perception. We tend to make it

a logical necessity that what I now immediately perceive must exist now. There is a

contingent fact involved, a fact about the speed of reaction of our sensory apparatus to

physical events in our environment; but a conceptual necessity about immediate

perception is erected on the foundation of this contingent fact. The discovery that we see

the stars as they were many years ago comes as something of a shock to our conceptual

system. But it does not force any extensive conceptual revision on us, for, when we look

at the stars with this new knowledge, it still looks as if we are immediately perceiving

present events. We need only say that what I immediately perceive now must at least

seem to exist now.

Let’s suppose that Armstrong is right and that it is not logically (conceptually) necessary that

the events perceived occur at the same time as the perception of them. What I claim

nevertheless is that they always do. In other words, it is not just the case that normally the

events perceived occur at the same time as the perception of them; it is also the case that in

absolutely every case of perception we do perceive something that really exists or occurs at the

time we perceive it. And I claim this because I can’t imagine an exception to this regularity, just

as I can’t conceive of something’s happening without a cause. I can’t conceive of an effect’s

happening without a cause because that would violate a logical (linguistic, semantic) rule; and I

66

can’t conceive of an event’s happening without a cause because that would violate a

metaphysical rule. In the same way I can’t conceive of perceiving something “this” second

without there being something there that is perceived “this” second. This would violate a

metaphysical rule, too. (We can call both logical and metaphysical rules conceptual). Therefore,

if that thing’s not physical, it must be non-physical; and there’s no way around this. I hope

finally to settle this issue in Chapter Four.

2.8. Conclusion about Dretske, Quinton, and Armstrong

Having surveyed three attempts to meet the TGA by denying the simultaneity-requirement, we

are ready to state a few conclusions. One is certainly that if one claims that one directly sees a

star now, then one must admit one sees it as it was (when it emitted the light in virtue of which

one sees it) and not as it now is. This point, already obvious, will be reinforced in subsequent

chapters. Now the TGA is an argument designed to show, with the help of the simultaneity-

principle, that if we do see physical objects, whether directly or indirectly, at all, then we do so

in virtue of directly seeing sense-data. And this argument has not been refuted by either Dretske,

or Quinton, or Armstrong.

Dretske articulates the argument in terms of an unsuitable example and thereby fails to bring

out the full force of it. When in looking at a coffee pot he asks what he can be seeing directly

other than the coffee pot itself, to prove to him that he also sees sense-data is very difficult

(though not impossible) because he does see the coffee pot directly, and additionally postulating

the direct perception of sense-data thus seems thoroughly uncalled-for. But claiming, as he does,

that the coffee pot can go out of existence before the seeing of it means ignoring the difference

between physical and phenomenal simultaneity, with only the latter being required for human

perception. And when he starts talking about the coffee pot’s looking some way to him and

asserts that this is simultaneous with his perception of it, he is introducing something

suspiciously close to sense-data. Moreover, he seems to be forced to admit that a no-longer

existent star still has sensible properties at the time it is seen. Thus Dretske’s case against the

simultaneity-requirement is wobbly.

Quinton fails to raise and to answer some obviously pertinent questions, such as: Why does

someone take herself to perceive a certain event somewhere when there is no such event to be

perceived, no such event occurring there at all? When I see the past but mistakenly think it to be

the present, what explains this mistake? Why do I make it? What is it that leads me to think

wrongly that what is actually past is really present? Failing so much as even to consider these

67

questions makes his case against the TGA and the simultaneity-requirement still more seriously

incomplete.

Armstrong makes a conscious effort to answer these questions without invoking sense-data,

and this might amount to a satisfactory treatment of the TGA while denying the simultaneity-

requirement, if he had also convincingly answered his own further questions: What it would be

like to see, immediately, one event occurring very much (centuries, days, or even minutes)

before another? Why should I not acquire immediate information about the past, knowing it to

be the past, by using my senses? The fact that Armstrong tries to answer questions that Dretske

and Quinton either shortchange or ignore and that he comes up with some pertinent questions of

his own, makes Armstrong’s account easily the best of the three, although Dretske’s is (no less,

and perhaps even more, than Armstrong’s) an excellent introduction to the problem.

68

Chapter Three: Denying the Simultaneity-Requirement (II)

3.1. Can Our Eyes Range into the Past (Only)?

In the philosophical world today denying the simultaneity-requirement is by far the more

popular position. But often this denial seems to be inadequately argued for; sometimes it gets

even less support than Quinton, say, gives it. On the other hand, there’s a certain almost

irresistible pull towards denying it, a pull of which one might have been conscious already as

soon as one had read one’s Dretske and especially one’s Armstrong. It might become stronger

still after one considers some of the philosophers yet to be discussed, in this chapter, before

finally realizing that both philosophical reason and common sense require us to resist this

ensnaring pull.

One of the first post-war philosophers to deny the simultaneity-requirement was A. J. Ayer in

1956. In The Problem of Knowledge, after stating the TGA, he went on to criticize it as follows:

This argument draws its strength from the fact that one tends to think of seeing as

connected only with the present. It is assumed that, unlike our memories or our

imaginations, our eyes cannot range into the past: whatever it is that we see must exist

here and now if it exists at all. But this assumption is not unassailable. Why should it not

be admitted that our eyes can range into the past, if all that is meant by this is that the

time at which we see things may be later than the time when they are in the states in

which we see them? And having admitted this, why then should we not also admit that it

is possible to see things which no longer exist? Such ideas might never have occurred to

us were it not for the discoveries of physics; but once these physical facts are recognized,

it does not seem too hard to adapt our way of speaking to them. We have to balance the

oddity of saying that we can see what is past against the oddity of saying that we do not

see physical objects; and to give our eyes access to the past may well seem the more

reasonable course.64

However, this is not the choice we are faced with: in seeing what is past we, if we saw

anything, would still be seeing physical objects; but if we do deny that we can see what is past

we are not thereby forced to say, for all that Ayer has shown, that we do not see physical objects

at all, only perhaps that we do not see them directly. Thus this reason that Ayer gives for

denying the simultaneity-requirement seems thin.

64 Ayer, The Problem of Knowledge, 94-95.

69

Moreover, in talking about adapting “our way of speaking” to the facts he seems to suggest

the view that on a certain (deep?) level there is no “fact of the matter” with respect to the

simultaneity-requirement, only “ways of speaking.” But he suggests—and there seems to be—

no reason to believe this.

Incidentally, while in 1956 Ayer seemed willing to “admit that it is possible to see things

which no longer exist,” in the late 1980s, not long before his death, he showed no sign of this

willingness when he wrote (in another context) that “it is at least very doubtful whether it makes

sense to talk of being a current observer of past events.”65

Another post-war proponent of denying the simultaneity-requirement was Roderick

Chisholm, who said:

“. . . we tend to assume, until we are taught otherwise, that any event or state of affairs

we perceive must exist or occur simultaneously with our perception of it. . . . We tend to

assume, more generally, that S can perceive a at t only if a exists at t.”

Chisholm proposed to refute this assumption as follows:

But to assume that S can perceive a at t only if a exists at t is no more reasonable than to

assume that S can receive or reflect light from a at t only if a exists at t. The perception

of a star that is now extinct should be no more paradoxical than the action of such a star

on a photographic plate or its reflection in the water.”66

It is perfectly true that a star can act on a photographic plate or be reflected in water when it, the

star, no longer exists. But is the action of a star on me while I perceive it really much like its

action on a photographic plate or its reflection in water? Is a perceiver really much like a

photographic plate or like water? Is perceiving really much like receiving a physical

impression? In short, is being on the receiving end of a physical action really all there is to

perceiving? The extent to which these questions call for a negative answer is the extent to

which Chisholm has not refuted the simultaneity-assumption. And that extent is immense if, as

seems obvious (and as Chisholm should be the first to admit), perceivers are beings capable of

consciousness and perceiving very often involves the perceiver’s awareness of what he/she

perceives. Therefore, Chisholm’s analogy appears weak.

Incidentally, I might mention that Chisholm made no secret of his great admiration for

Thomas Reid—he not only regarded him as “one of the great philosophers of all time” but also

claimed that “Reid’s philosophical theory of perception is the correct one.”67 Now it’s quite

surprising to hear the latter claim from someone who should not fail to be aware that Reid

65 Ayer, “Reply to T. L .S. Sprigge,” The Philosophy of A. J. Ayer, ed. L. E. Hahn (La Salle, Ill.: Open Court, 1992): 602. 66 Chisholm, Perceiving, 153. 67 Chisholm, “Keith Lehrer and Thomas Reid,” Philosophical Studies 60: 33.

70

accepted the simultaneity principle, and did so most emphatically.68 Of course, Reid gave no

argument for it—but that’s only because he took it for granted as a first principle needing no

argument. Reid did not thereby deny the principle might conceivably be disputed and even

proven false; but it’s fair to say that Chisholm’s remarks quoted above don’t come close to

disproving it.

Perhaps we might interpret Chisholm more liberally and generously as claiming that

perceiving is like receiving a message. We already encountered this claim when discussing

Dretske (see 2.4.) and Armstrong (see 2.7). In Dretske’s lips its import was unclear; but we

thought the sense-datum theorist could very well endorse the suggestion, made by Armstrong,

“that what we immediately perceive is not the star, but a present happening, causally connected

with the extinction of the star many years ago. The star sends a message to us, as it were, and

we immediately perceive the message, not the star.”69

Armstrong, as we’ve said, does not accept this suggestion literally; for he thinks that it only

seems to us as if something is happening in the sky at present, that what we really see is a past

happening, and, anyway, that we can’t perceive sense-impressions. The sense-datum theorist

can unabashedly embrace Armstrong’s suggestion and claim that the star’s message is the

sense-datum. But Chisholm, perhaps no less than Armstrong, wishes to avoid sense-data.

Nevertheless, his reasoning on this point, as expounded in several of his classic publications,

has, to my mind, been effectively countered by Andrew Chrucky.70 Be that as it may,

Chisholm’s reason for denying the simultaneity-requirement seems unconvincing.

On the other hand, Ingvar Johansson, writing in the 1990s, agrees that the simultaneity-

assumption must be categorically denied, but claims that Chisholm doesn’t grasp the full

implications of this denial. They are in Johansson’s view quite radical:

If direct realism is true in relation to things around us, then we perceive backwards in

time, but it is then an extremely short time interval, almost infinitesimal, which we

bridge. However, if direct realism is true even for star perception, then we can perceive

backwards over a huge time interval. Mostly, we do in one and the same perception

perceive things at different distances from us. This means that veridical perceptions of

the world are extended backwards in time. This is what we have to teach ourselves if we

want to be direct realists, but Chisholm doesn't even hint at this. As far as I know, there

is only one philosopher who has stated this feature of perception clearly, and that is

68 For example, in the first of his Essays on the Intellectual Powers of Man, Reid states: “Thirdly, the immediate object of perception must be something present, not something in the past. We can remember past events but we can’t perceive them.“ 69 Armstrong, Perception and the Physical World, 147-148. 70 www.ditext.com/chrucky/sensa.html

71

Samuel Alexander at the beginning of this century, but, for some reason, he mentioned it

only in a footnote.71

The relevant footnote is to a 1912 paper; it has Alexander speaking of experience as

“compresence within the world of the experiencer and the experienced” and claiming:

Perhaps I should say at once that compresence does not mean simultaneity in time. I am

compresent with a past event which I apprehend. And indeed the events I perceive

always are past, by however small an interval. Compresent means simply belonging to

the same universe.72

Thus, according to Johansson and Alexander, we are always without exception seeing things

that are more or less in the past. This was a rather popular position among scientifically-minded

philosophers before and long after the First War,73 and still continues to be.

For example, more recently, the same thesis has been advanced by Robert L. Solso. He says

that “[o]ne implication of the finite speed of light is that we always see the past: the light

striking our eyes brings us information from some finite time ago when it was reflected from or

generated by some object at a distance from us. While the terrestrial past is very short, the

cosmic past is very long indeed. Look at the stars tonight and think how long it took the light to

reach your retina, which is effectively how far into the past you are viewing.”74

The (by now) obvious reply to this is the following:

(1) practically everything we see closer to us than the heavenly bodies exists in the present

time after all, because what matters for our perception is human- phenomenal, not physical-

scientific, simultaneity;

(2) even if it were the case that we can see the past, this would apply at best to some

physical things and events only; for no matter what we see, we always see mental sense-data,

too, and these exist at the time we see them;

71 Ingvar Johansson, “Perception as the Bridge between Nature and Life-World.” In: hem.passagen.se/ijohansson/ 72 Quoted by Johansson (in the above-mentioned text) from S. Alexander 'The method of metaphysics; and the categories', Mind N.S. 21:1-20, p. 3 n. 2 (1912). 73 In 1912, a group of American philosophers (E. B. Holt, Walter T. Marvin, William Pepperell Montague, Ralph Barton Perry, Walter B. Pitkin, and E. G. Spaulding) published a cooperative work, The New Realism, in which they made the same point (in the words of W. P. Montague) as follows: “The events we perceive as present are always past, for in order to perceive anything it must send energy of some kind to our sense organs, and by the time the energy reaches us the phase of existence which gave rise to it has passed away.” E. B. Holt, from whom Chisholm apparently took his cue, put it this way: “An object is not only frequently but invariably seen at a moment of time later than that when it had the position and other circumstances which it still has in our vision of it. The illustration hallowed by the tenderest association for the idealist seems to be the case of seeing the sun or other heavenly bodies some millions of years behind time, or indeed millions of years after it has ceased to exist. But now what advantage over us has the photographic plate, as plainly physical and as little mental or illusory as we all grant that to be?” 74 Solso immediately goes on as follows: “To give some scale of the time involved, light from the sun takes about eight minutes to arrive, while light arriving tonight from the Andromeda nebula was already one million years old before humans appeared on Earth ...” See his Cognition and the visual arts, p. 8.

72

(3) the information vision affords us is not, literally speaking, from the past but sometimes,

at best indirectly, about the past.

The first claim is secure; the second and third we will finish arguing for in the next chapter.

We have already discussed in detail Dretske’s, Quinton’s, and Armstrong’s ways of denying

the simultaneity-requirement and found them more or less wanting. Most other similarly

oriented approaches follow in their footsteps. Thus, Georges Dicker, for example, claims that

the simultaneity thesis is shown false by the “relevant scientific facts.”75 I’ve already argued

(against Dretske, in 2.3. and 2.4.) that science shows no such thing. But another author writing

at about the time of Dretske, Quinton, and Armstrong (and standing strongly under the latter’s

influence) is George Pitcher, who deserves special treatment—mainly because of the way he

calls attention to, and evaluates the significance of, the fact that sound, too, travels at a finite

(but slower) velocity.

3.2. The TGA and the Velocity of Sound: Pitcher et al.

Dubbing the TGA the “time-lag argument” (though this makes little difference) and wishing to

uphold Non-Idealist Direct Realism, Pitcher says that a Direct Realist

does not have much to fear from the time-lag argument. He can simply insist that the

finite speed of light does not entail that we do not directly see things and states of affairs

in the “external world,” but only that we must see them as they were some time ago. We

see real physical things, properties, and events, all right, but we see them late, that is all.

According to a [D]irect [R]ealist, it is a mere prejudice of common sense—and one on

which the time-lag argument trades—that the events, and the states of objects, that we

see must be simultaneous with our (act of) seeing them.

Here’s Pitcher’s argument on behalf of the Direct Realist:

If light waves travelled as slowly as sound waves, he would say, common sense would

never have made that erroneous assumption. When one is watching a baseball game

from some distance away, the sound of the bat hitting the ball reaches us a few seconds

after it is actually made, but that fact does not, and should not, lead one to suppose that

he is (directly) aware only of an auditory sense-datum, and not of the crack of the bat

actually hitting the ball. He is (directly) aware of the real physical sound made by the bat

hitting the ball, but he simply hears it a few seconds late. It is a scientific discovery that

the same sort of thing must be said, mutatis mutandis, about our seeing of things and

75 G. Dicker, Perceptual Knowledge, 48.

73

events. But that discovery need not have the slightest tendency to make a [D]irect

[R]ealist abandon his theory in favor of a sense-datum theory.76

Let’s examine this passage more closely. Pitcher argues that the fact that the sound of the bat

hitting the ball reaches us a few seconds after it is made should not lead us to suppose that we

are directly “aware only [his emphasis] of an auditory sense-datum, and not of the crack of the

bat actually hitting the ball.” But what is the force of this “only”? Is Pitcher saying that we are

directly aware of this auditory sense-datum, but we’re also aware of something else: “the crack

of the bat actually hitting the ball”? Or is he saying that the auditory sense-datum is (identical

with) the crack of the bat against the ball? Or, thirdly, is he saying we’re not aware of an

auditory sense-datum at all, just of that crack of the bat hitting the ball?

I think he definitely prefers the third alternative (otherwise he’d hardly have put this passage

in a chapter entitled: “Sense-data and How to Avoid Them”). In that case, however, I would

argue (1) that we are directly aware of this auditory sense-datum; (2) that it is identical with the

crack of the bat, i.e., the cracking sound caused by the bat’s collision with the ball; (3) that we

are also, as Pitcher says, “aware of the real physical sound made by the bat hitting the ball,” and

(4) that this latter awareness incorporates our interpretation of the auditory sense-datum as a

physical sound. Thus by denying that we’re directly aware of the auditory sense-datum, Pitcher

would be leaving something important out of the story.

But perhaps Pitcher wouldn’t be leaving anything out, perhaps I’d be putting something in,

something that doesn’t belong in that story? For why on earth, someone might object, do we

need the auditory sense-datum when we’ve got the audible crack of the bat? Well, my reason

for saying so is this: let’s remember that we also have the bat’s collision with the ball as well as

multiple events of visually and audibly apprehending that collision. The collision is precisely

datable (in human perceptual terms): for every baseball game watcher who directly sees it, no

matter where they’re located in the arena, the visual awareness of the collision occurs at exactly

the same time (modulo human perception) that it itself occurs. But the auditory awareness of

that same collision is a different matter: the batter, the pitcher, and the players and viewers

close-by all hear the collision simultaneously with its occurrence and with its visual

apprehension; but those further away in the arena hear it fractions of a second later in increasing

lengths of time-lag directly proportional to their distance from the collision.

Thus there does seem to be a good case for saying there are multiple events each of which is

individually simultaneous with a certain baseball game watcher’s perception of that same event.

76 Pitcher, A Theory of Perception, 48.

74

That’s their hearing of the sound. So Pitcher, like the other Direct Realists discussed previously,

seems not to be able to escape sense-data, after all.

Moreover, it appears that the simultaneity-thesis is as well if not better supported by our

experience with the speed of sound than it is by our experience with the speed of light.

This is precisely (and perhaps surprisingly) what Armstrong, too, appears to believe. I’ve

previously quoted the following remark of his:

It may be suggested that what we immediately perceive is not the star, but a present

happening, causally connected with the extinction of the star many years ago. The star

sends a message to us, as it were, and we immediately perceive the message, not the star.

Armstrong, of course, rejected this suggestion for solving the visual time-gap problem, but

immediately following the above quotation he went on to concede:

Now this suggestion may be correct in the case of sound. There seems to be some force

in thinking of sound as actually spreading out from its source, like a balloon rapidly

inflating. (And here I am not speaking of the sound-waves.) So when two people ‘hear

the same sound’ it may be argued with some plausibility that they immediately hear two

different things, because they are in different positions. These considerations may

dissolve the problem of the time-gap when we hear a ‘distant’ sound. But the suggestion

seems inapplicable to the case of the star.77

I’ve already explained how it is applicable to the star (for visual sense-data are like sounds in

this respect), but the point to be stressed now, against Pitcher, is that it is applicable to the case

of hearing the baseball crack as well.78 To paraphrase Pitcher, “the events, and the states of

objects, that we hear” are undoubtedly “simultaneous with our (act of) hearing them,” no less

than the event and the state of the bat-ball collision we see is simultaneous with our (act of)

seeing it.

Len Carrier has also argued that “those who dismiss” the simultaneity-requirement “do so

for inadequate reasons”:

If conceptual truths are to be based ultimately upon factual ones, the latter ought to be

general enough to apply to all perception. This does not seem to be the case. Though the

speed of sound is far less than that of light, we do not consider it correct to say that we

hear physical objects that no longer exist. And though it takes time for nerve impulses to

77 Armstrong, Perception and the Physical World, 146-147. Incidentally, his assertions about sound (in particular, his explicit denial that here he is talking about sound-waves) are difficult to square with his taking sound to be physical. 78 With the understanding, of course, that the sense-datum theorist is not thereby prevented from claiming that we hear the physical collision, too.

75

reach the central nervous system, we do not speak of touching or tasting objects that no

longer exist.79

Quite agreeing with the spirit of the above quotation’s third sentence, I would only amend its

main clause to state (pedantically but accurately) that “we do not consider it correct to say that

we now immediately hear physical events that just occurred but we do consider it correct to say

we now immediately hear the sounds these events just made.” (And, of course, in immediately

(directly) hearing the sound we mediately (indirectly) gain information about the physical

collision.)

Finally, that the slower velocity of sound is of no help to the Direct Realist had already been

pointed out by Lovejoy in 1930. To his claim that

[t]he doctrine of the finite velocity of light meant that the sense from which most of our

information about the world beyond our epidermal surfaces is derived never discloses

anything which (in Francis Bacon’s phrase) “really exists” in that world, at the instant at

which it indubitably exists in perception”

Lovejoy added the following footnote:

The retardation of auditory sensation must so soon and so constantly have forced itself

upon the notice of primitive man that an implicit epistemological dualism with respect

to sound may be supposed to have prevailed from an early period in the history of the

race. It was, however, a vague dualism because a sound does not so clearly present

itself as occupying a definite place, or as an adjective of an object or event in such a

place.80

The location and nature of sound (I think by “adjective” Lovejoy meant “attribute”) still

engenders philosophical controversy (see, e.g., the articles on Sounds and Auditory Perception

in the Stanford Encyclopedia of Philosophy). The sound that an epistemological dualist in

Lovejoy’s sense (i.e., an Indirect Realist or Idealist) needs is a sensation “produced in the

observer when the sound waves strike the ear,” as D. L. C. Maclachlan has it.81 Or,

equivalently, it is a quale, a sound-quale, “whose esse is audiri,” as John Foster puts it, adding

that this “position comes to assume the familiar form of the traditional sense-datum theory.”82

As Foster explains,

“[q]ualia are realized in sensation. . . . We must take the sensory realization of qualia as

fundamental, and see their occurrence as objects of awareness as something to be

explained in terms of it. We must say, not that qualia are realized by being sensed, but

79 Carrier, “The Time-Gap Argument”, 269. 80 Lovejoy, 24. 81 D. L. C. Maclachlan, Philosophy of Perception, 26 82 Foster, The Case for Idealism, 104.

76

that they are sensed by being realized. . . . Such qualia . . . are the very sensation-types

(or sensory universals) of which . . . sensations are but the tokens (or instances. And

when realized, these qualia are displayed, as objects of awareness, simply because the

sensations which instantiate them, being episodes of consciousness, are self-revealing:

the sensations display their own intrinsic character and, thereby, display the sense-qualia

they entoken. . . . we must say that any sensation, in whatever sense-realm, is just a token

or instance of a certain sense-quale, and that it constitutes an awareness of that quale

simply by being, as an episode of consciousness, self-revealing. Thus we must say that

colour-patterns are visually realized because they are the very sensation-types (sensory

universals) of which visual sensations are the tokens (instances), and that sound-qualia

are auditorily realized because they are the very sensation-types (sensory universals) of

which auditory sensations are the tokens (instances).83

Another philosopher who affirms the existence of sense-data and of sounds as auditory

sensations is Brian O’Shaughnessy. He first imagines this possible exchange:

‘Where was the sound of the Krakatoa explosion an hour later?’—‘In Australia, India.

Japan’.

and then goes on as follows:

That one and same sound—assuming a single individuatable roar occurred—might have

reached all those places (while all was silent at Krakatoa); exactly as the first note

emitted by Heifetz’s violin at some concert, and heard by a thousand listeners, will have

reached the farthest recesses of the auditorium 1/5 of a second after coming into being;

which implies that ‘it’ inhabited all those places by that time and had left the violin for

good. . . . In sum, while the sound originates at a distance and we can hear that it is

coming from a direction and even place, and while there is no auditory experience of

hearing that the sound is where we are, the sound that we hear is nonetheless where we

are.84

The sound is where we are, I would argue, because it is in our minds; and our minds may, for

some purposes, be regarded as where we and our bodies are.85 Thus I would agree with

O’Shaughnessy in spirit (if not in important detail).

I conclude that facts about sounds do not militate against the TGA for sense-data and its

crucial premise, the simultaneity-requirement, in particular.

83 ibidem, 105. 84 O’Shaughnessy, Consciousness and the World, 445. 85 See the section on “Embodiment” in John Foster, The Immaterial Self, 261-266.

77

3.3. Perception, photography, television, and memory (I)

One of the most recent—and original—attempts to deny the simultaneity-requirement is due to

John Bengson.86 His aim is to challenge both the view that “we perceive only the present” and

the view that “in order to perceive, one’s sensory organs must be operating at the time of

perceiving. In other words, one must use one’s sense organs to perceive.” I think that to deny

this second view seems positively bizarre for sense-perception; nevertheless Bengson intends to

make it seem plausible both (1) that by virtue of remembering something we can perceive “with

[our] eyes shut” and (2) that the viewing of photography and of television broadcasts can be

tantamount to perceiving the past.

But first he takes up a more straightforward case. Bengson writes:

Consider the so-called “North Star,” Polaris. Polaris is 650 light-years away, so its light

takes roughly 650 years to reach us. So, when we look at Polaris now (in the twenty-first

century), we are seeing it as it was in the fourteenth century. And if Polaris had suddenly

exploded in 2003, we would not know it until the year 2653. This is because we cannot

see Polaris as it is presently; we can only see it as it was, in the past. As a result, this is

an instance of perception of the past.

So far this is just a restatement of the denial, not an argument for it. Bengson argues for

Polaris’s perception being an instance of perception of the past by appealing to the causal theory

of perception. According to his version of this theory, an agent A perceives a perceptual object o

if (1) A has an experience as of o and (2) A’s experience is appropriately and counterfactually

dependent on o. Thus, if o is Polaris, A has an experience as of Polaris, and A’s experience is

appropriately and counterfactually dependent on Polaris, then it follows that A’s experience is of

something in the past—A sees Polaris as it was in the fourteenth century. “In such a case, we can

legitimately be said to be in perceptual contact with what our present experience is of—namely,

something in the past.”

It is interesting to note that Bengson clearly recognizes that it was David Lewis, a champion

of the causal theory of perception, who, of all people, “challenged the idea that the stars we

perceive in the night sky are in the past.” Bengson correctly renders Lewis’s ideas and also, to

my mind, rightly criticizes them. Having cited his words that “...the stars, as I now see them, are

not straightforwardly past; for lightlike connection has as good a claim as simultaneity-in-my-

rest-frame to be the legitimate heir to our defunct concept of absolute simultaneity,”87 Bengson

86 See his paper, “How to Perceive the Past with your Eyes Shut,” 2006. All quotations from Bengson below are from this intriguing paper. 87 See 2.6.

78

goes on to state that “Lewis appears to believe that if is seen now, then because it is within an

observer’s light cone, is simultaneous with what is happening now— is “not

straightforwardly past”, but present. If this is so, then an explosion of Polaris that occurred

almost seven centuries ago is present so long as we see it now”.

But this view, Bengson justly observes, has an unpalatable consequence. For “presumably,

such an explosion does not occur in the present. Indeed, it does an injustice to our concept of the

present to apply it to that which occurred almost seven centuries ago: pace Lewis, we clearly

consider such an event to be past.”

In consequence, we have the following highly intriguing situation: both philosophers are

keen on claiming that we really do see the stellar explosion but Lewis wants to keep the

simultaneity-principle and thus is forced to deny that the explosion occurred way back in the

past; whereas Bengson, recognizing that the stellar explosion did occur way back in the past, is

thereby forced to insist (happily!) that we can see the past—even to the point of our being “in

perceptual contact” with something “that occurred almost seven centuries ago.”

I think that both philosophers would be much better off if they admitted sense-data: not only

could Lewis then keep the simultaneity-principle, but he wouldn’t also be forced to deny that the

explosion occurred in the past; and Bengson wouldn’t be forced to claim that we actually enjoy

“perceptual contact” when, in gazing at the stars, we take ourselves to be seeing an explosion

and infer that it “occurred almost seven centuries ago.” My question here would be: perceptual

contact with what?

We will come back to this decisive point later but first let us follow where Bengson daringly

leads us—into territory where the distinctions between (1) seeing a scene, (2) seeing a

photograph of it, (3) seeing a television broadcast of it, and (4) remembering (having seen) it

seem to blur. He engages “in a bit of science fiction” and supposes that Martians who’ve

“finally” arrived on earth

possess a more sophisticated perceptual system: they can “freeze” a given visual image

(perhaps by arresting the patterns of light currently on the retina), allowing them to enjoy

a given perception while the world around them continues to change. When they freeze a

given image, do the Martians continue to perceive what it is an image of? It would seem

so. After all, they are enjoying an experience that, all things being equal, is appropriately

causally and counterfactually dependent on what it is of. So, if the Martians did in fact

perceive the scene initially, because their frozen perceptions continue to satisfy the

causal and counterfactual conditions that mark out genuine perception, it would seem to

follow that they still perceive the scene seconds or minutes later when it is frozen in their

79

perceptual field. Thus, such “snapshot perception” enables the Martians to perceive

(indirectly but genuinely) a given scene long after it has occurred.88

Bengson then says that “snapshot perception might not be wholly science fictional” and in

support quotes Solso:

When we [human beings] look at an object . . . we do not see it all at once, as common

wisdom suggests, but go through series of scans in which the eye momentarily stops on

one feature . . . then darts on to another part . . . and then on to another . . . and so on.

Each feature is seen in a brief glimpse, then our eyes focus on another point for

additional processing. The span of material seen at each fixation point is sharply limited

due to the narrow field of foveal vision. Since this scanning/stop maneuver takes place

over very short time periods, the subjective experience is that we are seeing [an object]

all at once, when, in fact, our visual perception of it is built up from a series of discrete

“snapshots.”89

Thus “we, like the Martians, enjoy a form of snapshot perception,” says Bengson. Next he

claims that Kendall Walton “has convincingly argued that viewing photographs involves such

snapshot perception. In most cases, photographs are appropriately causally and counterfactually

dependent on what they are of,” although some “over—and under—developed or artistically

altered photographs” might not be. The important point for Bengson is Walton’s claim “that

unlike the content of a painting or sketch, which depends on what the artist believes that he or

she sees, the content of a photograph “is determined by what is really there before [the

photographer], regardless of what he [or she] thinks”. . . . Because they have such “natural

dependence” (i.e., counterfactual dependence not mediated by an agent’s intentional states) and

thus preserve real similarity relations between objects, photographs, Walton argues, are

transparent.” It is in virtue of their “transparency” that photographs allow us to see “through”

to the reality they display.

88 Much earlier than Bengson it was Charles B. Daniels who (in his 1970 paper replying to Carrier’s important 1969 paper) envisioned just such a science-fiction scenario, only involving not Martians, but inhabitants of much more distant planets. In his 1972 reply to Daniels, Carrier wrote: “He imagines an advanced civilization in a distant planetary system being able to gather light in such a way that they could see us going about our activities long after we had died. But this is a situation that I cannot imagine if we are to speak strictly and avoid the looseness that results from similarity in visual content. For instance, we speak loosely when we say that we see people ‘on televison’. Strictly, what we see is a picture of light and shade, the causal origins of which includes a physical object. That we make this distinction in common speech is good reason for saying that the members of Daniels’ imaginary civilization would not see us at all, but only patterns of light.” I can only applaud wholeheartedly this passage of Carrier’s. 89 Solso, Cognition and the Visual Arts, p. 26. The quotation I give is more complete than Bengson’s. Solso, as we’ve seen, also claims that in all cases of visual perception we see the past only. With respect to that claim, Bengson, perhaps surprisingly, states “that it is unclear to me whether Solso’s desired conclusion—namely, that we always perceive the past—is correct.” Thus Bengson, while arguing that we can see the past, seems to want to leave open the possibility that we can see the present as well.

80

Furthermore, there are, says Bengson,

relevant similarities between photography and other devices which clearly do not

preclude but rather enable (indirect) perception. For instance, we would not deny that

someone with eyeglasses perceives what her visual experiences are of. Even though the

lenses act as mediating transducers, if appropriate causal and counterfactual dependence

is maintained, then she still sees (albeit indirectly). Because mirrors also maintain such

dependence, we should not deny that when looking at oneself in a mirror, one is in fact

looking at oneself. A periscope produces a similar result: so long as appropriate causal

and counterfactual dependence is maintained, its mirrors act as mediating transducers

which allow one to see (again, albeit indirectly). Since photographs also maintain

appropriate causal and counterfactual dependence, they, like eyeglasses, mirrors, and

periscopes, appear to act as mediating transducers that allow one to see (once again,

albeit indirectly). . . .

Interestingly, that looking through photographs is on a par with clear cases of

perception is consistent with how we ordinarily speak about viewing photographs. Just

as we are likely to assert that we see ourselves when looking in the mirror, although we

do so indirectly, we are likely to assert that we see a photographed object, albeit

indirectly. When viewing a photograph of the Taj Mahal, for instance, it is only natural

to (non-metaphorically) say things like “I see the Taj Mahal.” The same is true of film,

television, and other so-called “moving pictures.” Consider watching a game on the

television: we frequently and unhesitatingly say that we see the game, the players, the

fans, the ball, the last-second-shot, and so on. In fact, we often think we see such things

better via the television than can those (e.g., the referees) who are actually there, seeing

them directly!

No doubt Bengson is right that some of us do “frequently and unhesitatingly” say such things.

One problem, however, is that saying them might not “withstand critical scrutiny.” This is what

Foster argues in The Nature of Perception.90 He does this in the course of an attempt to refute

Indirect Realism after having given what he takes to be conclusive reasons for the sense-datum

theory and accordingly assuming that theory to be true. Hence Bengson would most likely reject

Foster’s arguments.

But there are others that do not depend on the truth of the sense-datum theory. One is

suggested by Bengson himself: after presenting a few more examples of unusual types of

(alleged) perception, including “some episodic memories” supposedly as “instances of another

90 See p. 205-218.

81

form of perception, one that allows us to (visually) perceive the past, even with our eyes shut,”

he admits that if one finds all this “highly counterintuitive” and “plainly unacceptable,” then

“one may take the argument of this paper to be a reductio ad absurdum of the causal theory of

perception.”

I think this theory is indeed flawed, but Bengson hasn’t presented the best version of it. He’s

left out a condition that makes the theory half-way plausible, one that is explicitly insisted upon

by its advocate David Lewis. This condition is not only a clause in a proper definition of the

causal theory of perception but also a basic requirement upon perception itself. That’s the view,

in Lewis’s words, that “I see” if and only if “[b]efore my eyes various things are present and

various things are going on.” Having thus endorsed the existence-requirement (of 2.1.), Lewis

goes on to say, famously, that it “is not far wrong to say simply that someone sees if and only if

the scene before his eyes causes matching visual experience. So far as I know, there are no

counterexamples to this in our ordinary life.”

But to my ears this supplement to the existence-requirement sounds too much as if there were

supposed to be two corresponding things in play here—the scene before one’s eyes (with, say,

five trees) and the matching visual experience (of five trees)—when, in fact, on any credible

theory of perception there should seem to be just one five-treed complex involved here (namely,

that contained in the scene before one’s eyes). Lewis proceeds to explain that (1) the visual

experience, which he implies “goes on in the brain (or perhaps the soul)”, and (2) the scene

before one’s eyes “cannot match in the way that a scale model matches its prototype, or

anything like that. Rather, visual experience has informational content about the scene before

the eyes, and it matches the scene to the extent that its content is correct. . . . The content of the

experience is, roughly, the content of the belief it tends to produce.”91

I think that this conception of visual experience is wrong, starting with the droll idea that

visual experience is something going on in the brain (I will discuss this in 5.4.) and ending with

its suggestion (perhaps among the earliest in the literature) of the Representational Theory of

Perceptual Experience (which I will criticize in Chapter Five). However, my main objection to

the Causal Theory of Perception is the same as Ayer’s when he says that

it seems to me obvious that no strict description of the content of a sensory experience

can carry any implication concerning its cause. I also adhere to my view that the

insertion of a causal clause into our analysis of perceptual statements, whatever good

reason there may be for it, is not a primitive procedure, since it is logically subsequent to

91 D. Lewis, “Veridical Hallucination and Prosthetic Vision.” In: Dancy, J. (ed.) Perceptual Knowledge (OUP, 1988): 79-80.

82

the transformation of percepts [sense-data] into visuo-tactual continuants [physical

objects]. It is true that, once I believe that a physical object is there in front of me, I am

likely to accept the subjunctive conditional that, unless it really were there, I should not

in these circumstances even be having the experience on which my perception of it is

founded. But this by no means entails that the acceptance of any such conditional is a

constituent of the belief that the physical object is there. On the contrary, unless the

physical object were already posited, it would not be available to figure in the causal

hypothesis. Neither can it be posited simply as the cause of the experience, for this

would be consistent with its being anything whatsoever. . . . [The] reason to believe in

the existence of physical objects . . . must lie in the character of our sense experiences . .

. it cannot simply consist in the attribution of causes to them.92

Consequently, Ayer claims, there is “more to our perceiving objects than their causing us to

have certain perceptual experiences,”—even, I would presume, “matching” ones—and this

sounds entirely right.

But let us return to some of Bengson’s examples. In endeavoring to assimilate memory to

perception, he argues as follows:

For instance, we often say things like “I can still see the look on his face.” Such

statements reflect a peculiar fact about memory: we often remember past scenes in such

a way as to make them “present” to us once again . . . Episodic memories . . . involve

being in a state that is qualitatively identical or very similar to a specific past experience.

For instance, when remembering one’s childhood, one may have an experience as of the

house in which one grew up or the yard where one often played. In such cases, past

scenes become present to us once again. . . . Like photographs, genuine episodic

memories involve experiences that are appropriately causally and counterfactually

dependent on what they are of. In effect, like photographs, genuine episodic memories

maintain natural dependence and preserve real similarity relations between objects.

Given this, some episodic memories appear to have much in common with eyeglasses,

mirrors, and periscopes. It would seem that like these prosthetic devices, genuine

episodic memories make a contribution to the enterprise of perceiving. In short, there

seems to be reason to believe that genuine episodic memories, like photographs, are

transparent.

Bengson then explains how

92 A. J. Ayer, “Replies.” In: G. F. Macdonald (ed.) Perception and Identity (Ithaca, N.Y.: Cornell University Press, 1979): 293.

83

it is reasonable to hold that some episodic memories are continuous with so-called

“ordinary” perception. For one, there is the possibility of snapshot perception. Second,

there is the way that we sometimes talk about our remembrances. In addition, as we have

just seen, episodic memories are in ontologically relevant ways on par with “ordinary”

perceptions. That is, in some episodic memories,

(α) A has an experience (qua memory) as of o, and

(β) A’s experience (qua memory) is appropriately causally and

counterfactually dependent on o.

That genuine episodic memories satisfy the causal theory of perception is at least prima

facie reason to think that, in such cases, we can legitimately be said to be in perceptual

contact with what our present memories are of.

Bengson’s suggestion that some episodic memories are forms of perception has already been

challenged (effectively, I think) by Lewis when he presented the following:

Example 2: The Memory. I hallucinate not at random; visual memory influences the

process; thus I seem to see again a scene from long ago; this past scene causes visual

experience which matches it. I do not see. No problem: the past scene is not part of the

scene before my eyes.93

In other words, if I seem to see vividly some past scene in the way of hallucinating it, I cannot

really be seeing it, because that scene is not in front of my eyes now. For the same reason I

cannot see that past scene if I’m only vividly recollecting it in the way of an episodic memory.

What’s more, vivid memory of a scene has a different “feel,” a different phenomenology

from actually seeing that scene—this is perhaps the main reason for distinguishing memory

from perception.

But actually seeing something is also different from seeing it in a photograph. In a footnote

to his discussion of seeing the Taj Mahal, Bengson says

Consider the question “Did you see the Taj Mahal?” In certain contexts, one can

legitimately answer by saying, “Yes, I saw the Taj Mahal”, if, for instance, one was just

shown a photograph of the Taj Mahal. But in other contexts, one can legitimately answer

by saying, “No, I have never seen the Taj Mahal”, if one has never seen the Taj Mahal in

person (regardless of whether one has or has not seen photographs of it). Incidentally,

one with a propensity for linguistic analysis might say that on the present view,

statements made while looking through photographs, such as “A sees o”, are to be

93 Lewis, ibidem, 83.

84

analyzed as “A sees o in virtue of seeing the photograph.” An adverbial analysis is also

available: “A sees o photographically.” Either way, A sees o.

But I would object that it is one thing to see a photograph of the Taj Mahal and quite another to

see the Taj Mahal itself. What we have here are two different things seen, not two different

ways of seeing one and the same thing. Of course, seeing the Taj Mahal and seeing a

photograph of it have something in common: both give us information about the Taj Mahal. But

the amount of such information is usually greater—both actually and still more so potentially—

when gained by seeing the Taj Mahal itself than when gained by seeing its picture. One is in a

much better position to acquire visual information by inspecting the Taj Mahal that one sees in

reality than one is by inspecting its photograph. One can gain even more information by

walking around the really seen Taj Mahal; but one can’t walk around the photographed Taj

Mahal at all; whereas walking around the photograph (which is possible) gives us no new

information about the building whatsoever.

Seeing something in the flesh (bodily, in person, in reality) puts the observer in a position to

gain not only more visual information about that thing, but also to touch it, to feel it, to smell it,

to taste it, to hear it—something seeing a photograph of it is incapable of doing.

3.4. Perception, photography, television, and memory (II)

Still, there’s a way of understanding realistic photography that minimizes the difference

between looking at it and visual perception of the world. Kendall L. Walton has articulated such

an understanding masterfully. It’s high time to “deconstruct” this alluring but faulty conception,

sentence by sentence. (I will put his sentences in bold, and my rejoinder in regular typeface.) In

the abstract to his famous paper,94 Walton writes:

“Photographs are transparent; in looking at a photograph of something one sees the

thing itself.”

No, in looking at a photograph of something one sees just a photograph; one thereby sees a

picture, an image, of that thing, but not directly that thing itself; the thing itself one sees, at best,

indirectly, which in some crucial respects isn’t nearly as good as seeing the thing itself directly.

You can’t, for example, hold up the photograph in one hand and with your other shake hands

with the “indirectly seen” woman pictured in it unless she herself is standing right next to you;

in which case you’d be shaking hands with the (directly seen) real woman, not with her image.

And being able to shake hands or otherwise interact with the persons (things) you see is

94 Kendall L. Walton, “Transparent Pictures: On the Nature of Photographic Realism”, Nous, Vol. 18, No. 1 (March 1984): 67.

85

absolutely essential to direct seeing, which is part of the reason why in looking at things

photographically you’re not really seeing them. Having said this, I am fully aware that more

needs to be said about the distinction between seeing things directly and seeing them indirectly;

this I will do in the next section.

“Photography is not just a new means of producing pictures.”

True; but it is principally that.

“It is also an aid to vision, as are eyeglasses, mirrors, telescopes, and microscopes.”

True again. But there are big differences between these mere aids to vision, on the one hand,

and both photography (as a means of producing pictures) and photographs (the pictures

produced thereby), on the other. Whereas the main use of the above-mentioned aids to vision is

merely instrumental, photographs and photography can have important aesthetic functions as

well. They (photographs and photography) are also useful in criminal and historical

investigations of all kinds: visually comparing photographs of people with the people

themselves by looking at both photographs and people can help, e.g., policemen and historians

to identify accurately both photographs and people. All this presupposes that mere aids to vision

(on the one hand) and photographs (on the other) belong to essentially different kinds of human

artifact.

“Mirrors enable us to see around corners. Telescopes and microscopes make distant

and small objects visible.”

But when you look at things in mirrors or through eyeglasses, telescopes, or microscopes,

even if you directly see only their images, you also automatically gain knowledge about where

you’d have to go and what you’d have to do to see them without these aids, i.e., directly.

Looking just at photographs of things never by itself tells you any of that.

“With the help of photography we can see into the past as well.”

Not really. We can see into the past only in the way of seeing pictures of it; but we can’t see

the past itself; the past we can only remember. Of the preceding three clauses, the first is true

and we’ve already admitted it. The second and third are in the course of being proven.

“We must resist the tendency to water down this claim, to take it as a colorful and

exaggerated way of saying that in viewing a photograph one has the impression of seeing

the thing photographed, or that the photograph one sees is some sort of substitute or

surrogate for the object. Watering it down in any of these ways endangers both its interest

and its truth. We really do, literally, see our deceased ancestors when we see photographs

of them.”

No, we don’t, not really and not literally. In this life, we really and literally don’t see our

dead ancestors at all. We last saw them, really and literally, some time before they passed

86

away, while they were still alive. (And my last three sentences would be true even if there were

no afterlife, even if this life were all there is.)

Now, to be sure, in viewing a photograph one does not have the impression of seeing the

thing photographed—“impression” is the wrong word here. But neither does one actually see

the thing (that was) photographed; what one may have is the illusion of seeing the thing in the

photograph. Notice how Walton subtly slips from talking about “seeing the thing

photographed” when denying it (the thing photographed) is an impression to talking about “the

photograph one sees” when denying it (the photograph) is a substitute or surrogate for the

object. Of course, the photograph is no substitute or surrogate; it’s more of a memento of the

object: a reminder rather than a replacement. Nothing is a substitute for a dead person,

especially if you loved her or him.

Finally, why must we “resist the tendency to water down” Walton’s claim if in so doing we

merely “take it as a colorful and exaggerated”—and figurative—way of saying that enjoying a

piece of realistic photography is like veridical perception. They are alike in that both present the

world as it is, although perception does it incomparably more powerfully and resourcefully.

Looking at photography is to real perception of the world somewhat as a computer is to a

human being. There are likenesses here without the former becoming a form of the other, as

Walton claims.

“Slippery slope considerations give the claim an initial plausibility. If we see through

eyeglasses, mirrors, and telescopes, don’t we also see via closed circuit television monitors

and live television broadcasts? If so, on what grounds would it be reasonable to deny that

we see athletic events when we watch delayed broadcasts of them, or that we see through

photographs and photographic films?”

We do see objects through eyeglasses, in mirrors, and through telescopes. But in the first

case we see them directly, and in the second and third, indirectly, by seeing their images. When

I see something through my glasses, I usually know just in virtue of seeing it how much and in

what direction I have to move in order to touch that thing. But—and this is crucial—I usually

don’t know this, and not in the same way, when I just see a live television broadcast of that

thing. Even if I see the thing or event on closed-circuit television, the image might not tell me

how to get from the monitor to the thing or event; usually I must have additional information to

do that. Much less do I see that thing or event if I only watch a delayed broadcast of it, or see a

photograph or photographic film of it.

On these points I am supported by O’Shaughnessy. He supposes that things can be so

arranged that we see both a TV image of a scene from a spectacle and next to it the scene itself;

thus, he says, “we could look at the scene, and then at the TV image, and whereas in one case

87

we really would be looking at the scene, in the other case we would be looking at an array of

coloured patches before us as presenting to view a particular simultaneous spectacle; and the

one simply is not the other. Despite the identity of content and appearance and time, and despite

the significant overlap of cognitive utility in the case of these two phenomena, only one of these

visual experiences is an example of “seeing the scene in the flesh,” of actually “setting one’s

eyes upon” that scene.”95

O’Shaughnessy then criticizes the analogy between seeing things in a mirror and seeing

them on TV. He admits that

one is initially inclined to say that the mirror shows the scene in the very same sense . . .

as the TV image. . . . and say that in one case we see “the real thing” and in the other

case “merely an image of the real thing.” But if we do not see “the real thing” when we

see the mirror image, what that is not “the real thing” do we see when we stare at

whatever it is that we see when we look into a mirror? It is empty verbiage to say that

“we see an image,” bearing in mind that in this situation we are unable to produce an

analogue of the coloured areas on the TV screen.96

But why (I ask) is it just empty verbiage to call that which we see in a mirror an “image”?

You can place yourself vis-à-vis the mirror and a certain thing in such a way that you both see

“the real thing” (suppose it’s black and three-cornered) and its image in the mirror. Then you

see two black and three-cornered things, one being “the real thing,” the other its mirror-image.

Why is that “empty verbiage”?

Nevertheless, the basic point (a correct one) to take from O’Shaughnessy is that one should

be very careful with slippery-slope considerations here.

Alva Noë 97 also discusses the difference between “normally” seeing a thing and seeing it on

TV:

When you watch a live sporting event on television, you are able to track what’s

happening, but you do so in a perspectivally non-veridical way. Perhaps you adopt the

standpoint of one or more cameras. Crucially, you don’t correctly or veridically

experience the event’s spatial relation to yourself. It would be dogmatic to deny that you

see the sporting event, that you see it ‘through’ or by means of the television cameras.

It may be “dogmatic” to deny this, but since dogmas may be true or false the real question is

whether this dogma is true, and so far Noë hasn’t shown that it is not. He goes on to say

95 O’Shaughnessy, “Sense Data,” 180. 96 Ibidem. 97 In his “Causation and Perception: The Puzzle Unravelled,” Analysis 63,2, April, 2003.

88

But it would be just as dogmatic to insist that there is no difference whatsoever between

normal perception, in the here and now, and televisual perception. The difference is . . .

easy to explain: when you witness events in person, your experiences track not only how

things are, but also how things are in relation to you.

I fully agree that there is this difference; that’s why I assert the first dogma and deny I see the

sporting event through or by means of the television cameras. I would say I do not see the

sporting event at all, whether directly or indirectly; I only have the illusion of seeing it.98 What I

do see are live TV images of it, which give me information about the sporting event either as

good as, or better than, or worse than the information I would get from seeing the sporting event

itself.

In a footnote added to the quoted passage, Noë writes:

I take it that it matters that you are watching the game on TV ‘live,’ i.e. in real time. The

security guard, for example, really watches (and sees) the crowd on closed-circuit

monitors. Exactly what the role of time is in perceptual experience is tricky. Can you see

the stars in the heavens, even though they may no longer exist?

Thus Noë is withholding his judgment on the TGA. However, he is denying the dogma and

asserting that we can directly see sporting events on TV if we see them “live.” The latter is what

I deny, on Fosterian grounds (explained further in 3.5.4.)

A few criticisms similar to mine of Walton’s main thesis (that photographs are transparent)

have also been voiced earlier by Gregory Currie and Noël Carroll. “With ordinary seeing we get

information about the spatial and temporal relations between the object seen and ourselves. . . .

Call this kind of information “egocentric information”. . . . Photographs, on the other hand, do

not convey egocentric information,” claims the former,99 although agreeing that “photographs

can serve, along with information from other sources, in an inference to egocentric

information.”100 According to the latter, as quoted or paraphrased by Walton, “I can ‘orient my

body’ spatially to what I see, either with the naked eye or through a telescope or microscope.

But when I see a photograph I cannot orient my body to the photographed objects. The space of

the objects is ‘disconnected phenomenologically from the space I live in’.101

Currie and Carroll are right, I think, and Walton’s response seems unpersuasive. He admits

that an “account of what it is to see should explain how seeing enables organisms to acquire

98 See Foster, The Nature of Perception, 218. 99 Currie, Gregory, Image and Mind: Film, Photography, and Cognitive Science (Cambridge: Cambridge University Press, 1995), 66. 100 ibidem. 101 Noel Carroll, "Towards an Ontology of the Moving Image," Film and Philosophy, ed. Cynthia Freeland and Tom Wartenberg (New York: Routledge, 1995), 71.

89

information about their surroundings.” But then he claims that “there is no reason to assume”

that such an account “must limit seeing to cases in which that is done.”

However, it seems obvious and indisputable that cases in which organisms are not enabled to

acquire information about their surroundings just aren’t cases of seeing, period. Being enabled

to get information about the percipient’s environment constitutes the very essence of visual

perception, at least when consciously undergone by humans above the age of infancy.

In “regarding viewers as actually seeing things when they see photographs of them” Walton

claims that he “was not especially concerned to be faithful to the ordinary sense of the word

“see” (if there is such a thing).”

Well, there certainly is; it’s been around in languages and cultures for thousands of years;

and that’s what we’re concerned with here. Nevertheless, I’m aware that some philosophers

think that in view of such considerations as Walton offers perception has “turned out to be

rather different from how common sense took it to be.”102

But these philosophers have nowhere near conclusively shown that they’re right and that the

ordinary sense or concept of perception needs to be (as opposed to: may be) replaced, changed,

or even just extended. I admit, though, that Walton and Bengson have raised new and intriguing

questions.

Still, on the issue immediately relevant here, whether we can see into the past, the verdict is

still out, as far as the evidence so far presented is concerned. By the time we have finished the

next chapter, this situation, I hope, will have changed.

3.5. Direct vs. Indirect Perception

We’ve already and repeatedly spoken of perceiving “directly” or “indirectly”; I think it is very

important not to conflate different ways of making the distinction as well as different senses in

which perception may be direct (immediate) or indirect (mediate).

3.5.1. Cornman, Russell, Berkeley

The first post-war analytic philosopher to insist seriously on this point seems to have been

James W. Cornman. In 1971 he was able to claim, probably correctly, that “I know of no one

who has worked very carefully at the task” of explaining what is meant by the expression

“directly perceive an object,” even though Berkeley and Russell are “[t]wo philosophers whose

102 See Robert Hopkins’s review of Foster, The Nature of Perception in The Philosophical Quarterly (Vol. 51, No. 205, Oct. 2001): 555.

90

philosophy depends on a clear understanding of direct perception.”103 According to Cornman,

direct perception for Berkeley implies an exclusively “factual relationship” between perceiver

and object perceived and “has no epistemological component,” whereas Russell “adds an

epistemological feature to Berkeley’s purely factual interpretation.” I will now proceed to

explicate their conceptions (and others’) in my own way and then draw some conclusions.

According to Russell, “we are directly aware” of something when we are aware of it “without

the intermediary of any process of inference or any knowledge of truths,” in which case we have

“knowledge by acquaintance” of it or simply are just “acquainted with it”. As Russell says,

“[t]hus in the presence of my table I am acquainted with the sense-data that make up the

appearance of my table.”104 On the other hand, “[m]y knowledge of the table as a physical

object . . . is not direct knowledge. Such as it is, it is obtained through acquaintance with the

sense-data that make up the appearance of the table.” This is the first component of his

distinction between direct (immediate) and indirect (mediate) awareness.

The second component is his claim that “it is possible, without absurdity, to doubt whether

there is a table at all, whereas it is not possible to doubt the sense-data.” Thus, according to

Russell, direct awareness (or perception; in this context it doesn’t matter which word is used)

not only isn’t knowledge of a truth or the result of an inference, but its object cannot be doubted.

Third, he calls his knowledge of the table “knowledge by description” and says that “the

table is 'the physical object which causes such-and-such sense-data'. This describes the table by

means of the sense-data. In order to know anything at all about the table, we must know truths

connecting it with things with which we have acquaintance: we must know that 'such-and-such

sense-data are caused by a physical object'. There is no state of mind in which we are directly

aware of the table; all our knowledge of the table is really knowledge of truths, and the actual

thing which is the table is not, strictly speaking, known to us at all. We know a description and

we know that there is just one object to which this description applies, though the object itself is

not directly known to us. In such a case, we say that our knowledge of the object is knowledge

by description.”

From any of these statements of Russell’s, does it immediately follow that the thing

perceived must exist at the time it is perceived? This is perhaps debatable. Russell, as far as I

can tell, always assumed the simultaneity-requirement (to my mind, utterly plausibly and

correctly), but in no writing of his that I’m familiar with does he explicitly state it or, for that

matter, argue for it. Thus we might be tempted to ask, Isn’t it perfectly consistent (with what

103 Cornman, Materialism and Sensations, 222. The other quotations from this book are from p. 224. 104 Russell, The Problems of Philosophy (New York: Oxford University Press, 1959), 89. The other quotations from this book are from the same page.

91

I’ve quoted from Russell) to say that at time t I am aware of, and acquainted with, a sense-datum

of the star and hence that this sense-datum indubitably exists, but that the time it exists is earlier

than t?

Well, it seems that the answer is negative, after all. Here, unexpectedly perhaps, it is

Cornman who not only sets us straight but explicates Russell’s words as straightforwardly as

possible when he states the following: “Clearly, if to perceive something directly is, as Russell

claims, to be acquainted with it, to have it present and presented to the perceiver,” then the

thesis that all objects exist at the time they are directly perceived “is not only true but analytic.”

Cornman then proceeds to formulate a TGA which he thinks is very plausible; we will discuss it

in Chapter Four.

A simpler and more frequently encountered way than Russell’s of making the distinction

between direct (immediate) and indirect (mediate) perception takes its cue from Berkeley when

in the guise of Philonous he implies that things are “perceived immediately” when they are

perceived “without the intervention of others.”105 But, says Philonous,

. . . I grant we may in one acceptation be said to perceive things mediately by sense: that

is, when from a frequently perceived connexion, the immediate perception of ideas by

one sense suggests to the mind others perhaps belonging to another sense, which are

wont to be connected with them. For instance, when I hear a coach drive along the

streets, immediately I perceive only the sound; but from the experience I have had that

such a sound is connected with a coach, I am said to hear the coach. It is nevertheless

evident, that in truth and strictness, nothing can be heard but sound: and the coach is not

then properly perceived by sense, but suggested from experience. So likewise when we

are said to see a red-hot bar of iron; the solidity and heat of the iron are not the objects of

sight, but suggested to the imagination by the colour and figure, which are properly

perceived by that sense.106

Incidentally, Cornman seems to be in error when he says that “direct perception for Berkeley

implies an exclusively “factual relationship” between perceiver and object perceived and “has

no epistemological component.” The factual relationship is primary in the sense that it is what

defines direct perception, but in the same Three Dialogues Berkeley is certain that what is

directly (immediately) perceived is also indubitable. This is where epistemology comes in:

Wood, stones, fire, water, flesh, iron, and the like things, which I name and discourse of,

are things that I know. And I should not have known them, but that I perceived them by

my senses; and things perceived by the senses are immediately perceived; and things

105 Berkeley, Works, 3D, 174 106 ibidem, 204.

92

immediately perceived are ideas; and ideas cannot exist without the mind; their existence

therefore consists in being perceived; when therefore they are actually perceived, there

can be no doubt of their existence. Away then with all that scepticism, all those

ridiculous philosophical doubts. What a jest is it for a philosopher to question the

existence of sensible things, till he hath proved to him from the veracity of God; or to

pretend our knowledge of this falls short of intuition or demonstration? I might as well

doubt of my own being, as of the being of those things I actually see and feel.107

In his earlier Essay towards a New Theory of Vision, Berkeley had repeatedly made pretty

much the same distinction between immediate and mediate perception: “It is evident that when

the mind perceives any idea, not immediately and of itself, it must be by some other idea. . .”

which “is itself immediately perceived in the act of vision.”108 And “I know evidently that

distance is not perceived of itself. That by consequence it must be perceived by means of some

other idea which is immediately perceived, and varies with the different degrees of distance.”109

Again, “we must bear in mind that there are two sorts of objects apprehended by the eye, the one

primarily and immediately, the other secondarily and by intervention of the former.”110 In the

later Theory of Vision Vindicated and Explained, Berkeley even more emphatically generalized

the point for all the senses:

Ideas which are observed to be connected with other ideas come to be considered as

signs, by means whereof things not actually perceived by sense are signified or

suggested to the imagination, whose objects they are, and which alone perceives them.

And as sounds suggest other things, so characters suggest these sounds; and, in general,

all signs suggest the things signified, there being no idea which may not offer to the

mind another idea which hath been frequently joined with it.111

A bit later in the same work Berkeley applied the same distinction to sight and touch:

What we immediately and properly perceive by sight is its primary object, light and

colours. What is suggested or perceived by mediation, therefore, are tangible ideas which

may be considered as secondary and improper objects of sight.112

In Alciphron, Berkeley once more clearly reiterated this distinction by putting it in the mouth of

Euphranor, his spokesman:

107 ibidem, 230. 108 Berkeley, NTV, Sections 9 and 11. 109 ibidem, Section 18. 110 ibidem, Section 50. 111 Berkeley, TVV, Section 39. 112 ibidem, Section 42.

93

To me it seems that a man may know whether he perceives or no; and if he perceives it,

whether it be immediately or mediately; and if mediately, whether by means of

something like or unlike, necessarily or arbitrarily connected with it.113

Some of these points have in the 20th century (after the appearance of Dretske’s path-

breaking book) been elaborated upon by Moreland Perkins, Brian O’Shaughnessy, and John

Foster, among many others. Perkins, as far as I know, has not discussed the simultaneity

principle; Foster has (in private), but he has not endorsed it; O’Shaughnessy has (though

perhaps not as unequivocally as one might wish); and Berkeley, as I will argue in Chapter Four,

is committed to it, but not exactly because it follows from his definition of “immediately

perceive.” Thus I stand by my earlier claim that the simultaneity-requirement is not what

philosophers interested in a definition of direct perception start out with but is something that

reflection on direct perception may, and usually does, lead to.

3.5.2. Perkins

Perkins starts his whole book114 with a brief consideration of Berkeley:

In Berkeley’s language, the question from which this book arises is this one: Is what we

immediately perceive by the senses something that depends for its existence upon our

perceiving it and therefore something internal to our minds? My answer is, yes.

However, I prefer to speak of immediate “awareness.” . . . What we are directly aware of

are sensuous qualities whose instantiation within us partly constitutes our sensory,

conscious states. . . . Unlike Berkeley, whose power to believe in the incredible is one of

the wonders of the philosophical world, I am, like Locke, a merely commonsensical

Newtonian who presumes that there are mind-independent objects which we succeed in

perceiving by sense. These objects, therefore, I hold that we indirectly perceive; also,

that within perception the sensuous qualities belonging to us of which we are directly

aware represent for us objective properties of physical objects before our sense organs.115

No doubt because he finds it “incredible,” Perkins does not consider the idealist possibility at

all—the option that “the sensuous qualities belonging to us of which we are directly aware” do

not represent but actually constitute “for us objective properties of physical objects before our

sense organs,” thereby making them mind-dependent.

113 Berkeley, Works, 3:152. 114Sensing the World is sadly underestimated. David Smith correctly appraises it as a “fine work” containing an “exhaustive and unanswerable polemic against reductive accounts of sensory experience” (Smith, p. 279, n. 49). 115 Perkins, Sensing the World, 1.

94

What I wish to emphasize in the present context is the fine way in which Perkins makes

most of the required distinctions; I will again cite at length from his unfortunately little-known

text:

Sometimes we become conscious of one thing through becoming conscious of another.

A man walks into his house and sees on a table an open handbag; through becoming thus

visually conscious of the handbag he may become mentally conscious of the presence of

his wife somewhere in the house. He does not perceive that the handbag is there by

seeing or hearing his wife; he becomes conscious of his wife’s presence through noticing

the handbag. Although we do not ordinarily do so, we could describe his situation by

saying that after entering the house the man is visually conscious of the handbag, that he

becomes (mentally) conscious of his wife’s presence in being more directly conscious of

the handbag, and hence that he is only indirectly conscious of his wife.

Instances of sensory awareness of one item mediating a less direct awareness of

another are not limited to occasions, like the present one, of achieving a merely mental

awareness of the second item. Suppose that from his living room, where the handbag is

visible on the table, the same man now hears his wife’s voice coming from the kitchen.

He does not discover that a sound has occurred through hearing (or seeing) his wife; he

becomes conscious of his wife’s presence in the kitchen through becoming conscious of

the sound she makes in speaking from there. He is, one might say, more directly

conscious of the sound his wife has produced than of his wife herself. Of his wife in

person he is not yet, we may say, directly conscious. However, whereas in seeing the

handbag and thereby becoming aware of his wife’s presence somewhere in the house he

did not succeed in seeing his wife, so his awareness of her presence in the house could

not yet be said to be sensory (because not visual), and so was merely mental, now,

hearing her speak from the kitchen, he not only hears the sound she makes, he also, as

we always say, hears her: his indirect awareness of his wife we conceive to be also

auditory. But auditory awareness is sensory awareness; so when he hears her voice his

indirect awareness of his wife is sensory awareness. . . .

If we have begun to accept this mode of describing instances of awareness as more or

less direct, it will be natural by now to say that if this same man walks into the kitchen

and sees his wife, then he has, at last, become quite as directly conscious of his wife as

he had before been, first, of the open handbag and then, a moment later, of the sound of

her voice. It does not now seem to be the case that it is through being conscious of

something else that he can tell that his wife is there before his eyes, just as it does not

seem that it was through being conscious of something else that he earlier noticed the

95

handbag on the table or, a moment later, heard a voice. Without further critical reflection

we might naturally say that this man is directly conscious first visually of the handbag,

then aurally of the human, vocal sound, and finally again visually of the woman

herself—of whom he had, until at last seeing her in the kitchen, at first been only

indirectly conscious.116

All this, suggests Perkins, is what we might say without further critical reflection. But if we do

engage in such reflection we should come around to agreeing with those philosophers who

have held that what we unreflectively believe about direct, sensory awareness of things

before our sense organs is mistaken. They have held that we are directly conscious

through our senses of no objects or events situated before our sense organs, but always

only of items that are not before these organs—items, furthermore, which depend for

their very existence upon our being conscious of them. According to this view, of

everything before our sense organs of which we achieve sensory awareness we are

always only indirectly aware: we discover by sensory awareness of it that each such

object or event is out there before our sense organs only in virtue of being directly aware

of something that is not out there—an interior, sensuous item that depends for its

existence upon our direct awareness of it. This philosophical thesis concerning our

sensory consciousness of the physical world I shall call “indirect realism.”117

This is the thesis that Perkins in his book tries to show “is the truth about sensory awareness.”

The false thesis, by contrast, is the one that we “unreflectively” accept, namely, the thesis of

“direct realism,” which holds

that, of however many sorts of things before our eyes and other sense organs we may be,

in attending to them, only indirectly conscious, for each sense there exists before our

sense organs at least one sort of thing of which we can and commonly do become

directly conscious: of these physical objects and events—and of some of their features—

we achieve awareness without mediation through a more direct awareness of anything

else.

These and other remarks of Perkins’s constitute an elaborate expansion of Berkeley’s (quoted

above). In addition, paying particular attention to Perkins’s contrast between merely mental

awareness (roughly corresponding to Berkeley’s suggestion to the imagination and perception

by the imagination) and sensory awareness (which Berkeley usually calls immediate

perception), we can find in Perkins’s text plenty of ammunition against Walton’s and

Bengson’s claims that we can both “see the past” and “see with our eyes shut.”

116 ibidem, 11-12. 117 ibidem, 13. The next quote from Perkins is on the same page.

96

In this connection it is interesting to note the following divergence between some of the

“modernists” (Berkeley and Perkins included) and some of the “postmodernists” (Walton,

Bengson, et al). The two groups strive in opposite directions—the former seeking as it were to

set boundaries within and around normal vision; the latter endeavoring to loosen them up and

expand it (normal vision) considerably.

Thus, on the one hand, Berkeley in particular insists that “it must be acknowledged that we

never see and feel one and the same object” because “[t]hat which is seen is one thing, and that

which is felt is another.”118 He also insists that “[o]ur experience in vision is got by the naked

eye” and “[w]e apprehend or judge from this same experience, when we look through glasses”;

yet “[w]e may not . . . in all cases, conclude from the one to the other, because that certain

circumstances, either excluded or added, by the help of glasses, may sometimes alter our

judgments, particularly as they depend on praenotions.”119 On the other hand, the

“postmodernists” aim to minimize or erase altogether the boundaries between perception and

memory, and between seeing things in the flesh and seeing them in photos and/or on television.

3.5.3. O’Shaughnessy

Another philosopher for whom the concept of mediation is crucially important is Brian

O’Shaughnessy. He says that almost always

the perception of objects is through the perception of certain mediating X’s which are

non-identical with the object, and the same is true of the X’s themselves. With the single

exception of proprioception, mediation is needed for the entire gamut of publicly

perceptible items in physical space (proprioception evading the rule through being an

inter-systemic phenomenon). Epistemologically almost everything in physical nature lies

at a remove. Mediation is a near-universal necessity, and is exemplified in the perception

of material objects, light, movement, colour, sound, and much else.120

Since O’Shaughnessy’s nearly 700-page-long text is very dense, I will next avail myself of

David Smith’s convenient summary.121 According to O’Shaughnessy’s “multiply

representationalist” account, says Smith,

[w]henever a material object is perceived, it is in virtue of our perceiving several

‘mediators’. Something m is a mediator in relation to object x if it is non-identical to x

118 Berkeley, NTV, Section 49. 119 Berkeley, TVV, Section 69. 120 B. O’Shaughnessy, Consciousness and the World, 592-93. 121 A. D. Smith, “O’Shaughnessy’s Consciousness,” The Philosophical Quarterly, Vol 51, No. 205, Oct. 2001: 532-539. The quotations below are all from pp. 537-539.

97

and x is perceived in virtue of m’s being perceived. Mediators fall into two categories.

Those in one category, though non-identical with a given x, are also non-distinct from it.

For example, we see a physical object at any given moment only in virtue of seeing a

limited portion of its surface. Most would agree with this, if only because there is little

temptation to suppose that we see surfaces instead of material objects, and because

surfaces do present a visual appearance. O’Shaughnessy’s theory becomes much more

contentious when he argues that any material object is also seen only in virtue of our

seeing mediators that are entirely distinct from those objects. There are two sorts of

distinct mediators in the theory. The first is light. Whenever we see a material object

otherwise than as a silhouette, we see (in the full sense that we are visually aware of) the

light which that object is reflecting onto our retinas: “We always see light. Indeed, in a

certain sense only light. . . . Light epistemologically precedes the objects it brings to

view.”122

O’Shaughnessy also recognizes an even more proximal type of visual mediator,

sense-data. We ultimately see whatever we see in virtue of seeing sense-data. In part he

justifies this claim by the familiar appeal to causal considerations. It is possible for

visual experiences to be enjoyed in the absence of a suitable real-world object, and

indeed in the absence of any light impinging on our retinas. (O’Shaughnessy is

especially fond of citing after-images in this connection.) Should this occur, we would be

aware of something even more ‘inner’ than our retinas, namely, sense-data. All we get in

standard cases of perceiving objects in the world is something more: light, causal

relations to some material object, and so on. Such additional factors cannot obliterate the

form of awareness which was present in the former situation. Therefore sense-data are

present as immediate objects of awareness in all perceptual situations. There are,

however, distinctive features to O’Shaughnessy’s account. For one thing, he stresses

more than some do the way in which sense-data are quite independent of whether we are

aware of them or not, having purely physical causes in the optical system. Sense-data are

brute presences in consciousness, which we can overlook or attend to, just as we can the

physical objects whose presences they are.

With respect to a sense-datum’s “having purely physical causes in the optical system,”

Roderick Firth in his 1949-50 paper “Sense-data and the Percept Theory” complained that

“[s]ome philosophers have complicated the matter by actually defining a sense-datum as that

constituent of a perception which is caused by the physical stimulus.” As an example he cites

122 O’Shaughnessy, Consciousness and the World, 441, 442-443.

98

Russell, who on one occasion, in his Philosophy, wrote that a sense-datum is “the core, in a

perception, which is solely due to the stimulus and the sense-organ, not to past experience.”

Firth agrees that it “is empirically demonstrable . . . that the nature of particular states of

perceptual consciousness is determined partly by the direct physiological effects of the

stimulus, and partly by the past experience and present interests of the perceiver.” But then he

cautions that “to conclude from these facts alone that there must be at least two constituents in

every state of perceptual consciousness, one of them (the sense-datum) corresponding in some

simple fashion to the direct physiological effects of the stimulus, would be to commit the

physiological fallacy and obscure the actual character of perceptual consciousness.”123

Although it isn’t very clear what this alleged fallacy amounts to (and, indeed, whether it is a

fallacy) and thus not clear whether O’Shaughnessy actually commits it (or comes close to

committing it), his conception of sense-data seems somewhat to resemble Russell’s (as

represented above), and it does have some unusual features. As Smith points out, for

O’Shaughnessy,

[t]he existence of a sense-datum does not by itself constitute the having of a visual

experience. For that to occur, the sense-datum must be attended to and taken in some

way. What we take the object to be is the internal object, or the content, of the

experience. This, unlike the sheer awareness of sense-data, is an intentional affair. . . .

For one thing, intentional content is not determined by purely physical causes alone, as

sense-data are, but also by the recognitional and conceptual capacities of the subject. For

another, intentional content can have no ‘gaps’ in it: we cannot overlook part of what we

take an object to be, as we can overlook a sheer sensuous presence in our visual field.

Indeed, it makes no sense to speak of noticing the internal objects of experience: noticing

must already be there for such objects so much as to arise. O’Shaughnessy, indeed, uses

this point as an argument against the currently popular attempts to give a purely

‘cognitive’ or ‘intentional’ account of perception. . . .

Another important and distinctive aspect of O’Shaughnessy’s theory is his account of

what he terms the transitivity of attention. Because sense-data stand in reliable projective

relations to the arrays of light at our retinas (i.e., we normally see a red square sense-

datum to the right of a round yellow one because a red square beam of light impinges on

the retina to the left of a round yellow one), and because such light stands in similar

projective relations to (parts of the surfaces of) material objects, in seeing sense-data in

ordinary situations we ipso facto see the causally responsible light and material object.

123 R. Firth, “Sense-data and the Percept Theory, in: J. Troyer, ed., In Defense of Radical Empiricism, 29-30.

99

Smith goes on to make a number of interpretive and critical comments, one of which is the

following:

[T]hroughout the development of his account of perception, many readers will feel

surprise that O’Shaughnessy never addresses the epistemological problems which

generations of philosophers have believed to attach to representationalism. There simply

appears to be no problem here, as far as O’Shaughnessy is concerned.

It is difficult not to sympathize with this assessment. At the same time it is occasionally less

than clear whether and to what extent O’Shaughnessy really is a Representationalist: some

crucial passages of his, though partly Representationalist, tend in an Idealist direction. To be

sure, in a few other passages he distances himself from Idealism quite explicitly, though he

never deems it worthy of any serious discussion or refutation. Still, let us direct our attention to

the following:

In any case a theory of sense-data can be defended which comes close to the traditional

theory . . .It takes the following form. We shall assume that we are giving an account of

a normal enough example of seeing: more exactly, of monocular seeing (for simplicity).

Then in the first place the visual experience or perception is understood to be an event in

the attention: more precisely, an event of noticing. Secondly, the object of noticing is

something that is sited in the visual field. The third and most contentious part of the

theory is sense-data-ist in character. It claims that the immediate object of the above

attentive event is a visual sensation: more specifically, the colour and two-dimensional

properties of a two-dimensional array of visual sensations. Meanwhile, the visual

sensation itself is taken to be a a psychological individual of type sensation, which is

located in the visual field, and endowed with colour-brightess and an extent that is

characterizable through the use of two variables (such as right/left and up/down). The

theory concludes with the contentious and representationalist claim that the seeing of this

sensuous phenomenon is identical with the seeing of whatever public physical

perceptible it manages to make visible. Thus, one perceptual or attentive event is said to

fall under double descriptions: one involving reference to the sensuous object, the other

to the physical object.124

Though the event of seeing is one, the objects seen are two: this is, as far as it goes, a

Representationalist claim. But more typically a Representationalist would assert there to be a

duality of events, too; and he would mark this by using two different verbs: the sensuous object,

say, is sensed, the physical object is perceived. Sometimes he doesn’t use different verbs but

124 O’Shaughnessy, Consciousness and the World, 466.

100

still leaves no doubt that there are not only two objects, but two events as well. In both cases

this would be full-fledged Representationalism.

O’Shaughnessy at any rate shies away from it; he sticks to the one event-two objects story

and a bit later claims the following:

Now this account is representationalist in that it is in (or through) experiencing the

colour and contour of the immediately perceived sensation that we experience the

identical colour and contour of the distinct and mediately perceived public perceptible.

The claim is that the immediate material object of the visual experience is the visual

sensation—a senseless psychological primitive with colour and expanse—and nothing

else. It is not the public physical object, for it is the sensation; nor the experience, for no

experience can be its own immediate material object; and it is not the content of the

experience, for the content is not distinct from the experience and is not an individual. It

is a senseless psychological primitive that is endowed with only some of the experienced

properties of the public physical perceptible, namely its colour and two-dimensional

layout.125

So the visual sensation is not the physical object yet the seeing of this sensation is identical with

the seeing of the physical object. This account is (as I’ve admitted) partly Representationalist,

yet it tends toward Idealism. If the public physical perceptible had been a cherry and if

gustatory, tangible, olfactory sensations had been brought in in addition to the visual ones, then

these sensations or “senseless126 psychological primitives” taken together would have been

endowed with all of the experienced properties of the public physical perceptible, and

O’Shaughnessy would then surely have been able to echo the following words of Berkeley:

Take away the sensations of softness, moisture, redness, tartness, and you take away the

cherry. Since it is not a being distinct from sensations; a cherry, I say, is nothing but a

congeries of sensible impressions, or ideas perceived by various senses: which ideas are

united into one thing (or have one name given them) by the mind; because they are

observed to attend each other. Thus when the palate is affected with such a particular

taste, the sight is affected with a red colour, the touch with roundness, softness, etc.

Hence, when I see, and feel, and taste, in sundry certain manners, I am sure the cherry

exists, or is real; its reality being in my opinion nothing abstracted from those

sensations.127

125 ibidem, 467. 126 By “senseless” O’Shaughnessy most likely means “generated completely independently of intellect” (see ibidem, 477), but I would more naturally take “senseless” (somewhat like Berkeley’s “stupid”) as meaning “not itself sensing” (though, of course, the senseless is capable of being sensed.) 127 Berkeley, Works, 3D, 249.

101

Berkeley is saying that the cherry, being a congeries of sensible impressions, although not

identical with any or all of them, is also not distinct from these impressions. This seems to be

thoroughly Idealist just because in experiencing the sensation I am experiencing the physical

object insofar as this sensation partly constitutes that object.

3.5.4. Foster

John Foster is explicitly concerned to argue for what he calls Phenomenalistic Idealism. In his

latest and last book on this topic, he restates three key notions necessary for expounding and

defending his case. One notion is that of constitution, which he defines as follows:

A fact F is constituted by a fact F’, or by a set of facts S, if and only if the obtaining of F

is logically necessitated by the obtaining of F’ (the obtaining of the members of S), and,

with respect to that source of necessitaton:

(1) the obtaining of F is logically due to the obtaining of F’ (the

obtaining of the members of S);

(2) the obtaining of F involves nothing over and above the obtaining

of F’ (the obtaining of the members of S).128

It is Foster’s aim of prove that “the physical world is something whose very existence is

constituted by facts about human sensory experience, or by some richer complex of non-

physical facts in which such experiential facts centrally feature.”129 This is the thesis of

phenomenalistic idealism.

In order to distinguish the three rival views of perception mentioned at the beginning of

Chapter 1.1., Foster introduces two other key notions, one of which is that of perceptual

mediation and the other that of -terminal perception. He explains these as follows:

First, given a subject S and two items x and y that he simultaneously perceives, I shall

say that S’s perceiving of x (the fact of his perceiving x) is mediated by his perceiving of

y (the fact of his perceiving y) if and only if (1) S’s perceiving of x is constituted by the

combination of his perceiving of y and certain additional facts, and (2), apart from any

concern they may have with S’s perceiving of y, these additional facts do not involve

anything about S’s perceptual condition at the relevant time. Where a subject’s

perceiving of one item is mediated, in this way, by his perceiving of another, we can

128 Foster, A World For Us, 6. 129 Foster, ibidem, 40.

102

speak of him as perceiving the second item more immediately than he perceives the

first.130

Second, I shall say that a subject S -terminally perceives an item x at a time t if and

only if x is a physical item and S perceives x at t and there is no other physical item y

such that S’s perceiving of x at t is mediated by his perceiving of y at t. So, the

perceiving of an item qualifies as -terminal just in case the item is physical and there is

no other physical item that is, in the context of that perceiving, perceived more

immediately. (This leaves open the possibility of there being a non-physical item that is

perceived more immediately.) I shall take it for granted that physical-item perception is

not infinitely regressive, and that whenever a physical item is perceived, there is some

physical item which is, in respect of that perception, -terminally perceived.131

Now we can begin to bring out the difference between the three rival views Foster is concerned

to discuss as follows:

Suppose, at a certain time t, Ralph sees an apple on the table in front of him. His seeing

of the apple is mediated, in the sense defined, by his seeing of a certain portion of its

surface (it is constituted by the combination of his seeing of this portion and the fact this

latter item is a portion of the apple’s surface), and his seeing of this persisting surface

portion is, in an exactly analogous way, mediated by his seeing of a certain momentary

stage of it . . . This momentary stage of the apple’s surface portion is what, in the sense

defined, Ralph -terminally sees at t—what, relative to the domain of physical

candidates, he most immediately sees. . . . Let us call this -terminal object O. The

[Direct Realist] will say that, given that O is the -terminal object of perception,

Ralph’s seeing of O is a fundamental aspect of his psychological condition at t: his

fundamental psychological state, though wholly a matter of what is occurring within his

mind at that time, inherently involves his standing in this awareness relation to this

external physical item. In contrast, the [Representative Realist] will say that, instead of

being psychologically fundamental, the fact of Ralph’s visual contact with O breaks

down into two components. One component will cover all the relevant aspects of

Ralph’s psychological condition at t—all that obtains or occurs in his mind at that time

that in any way logically contributes to the obtaining of the relevant perceptual fact—and

the [Representative Realist] will insist that these aspects, on their own, do not secure

130 ibidem, 7. 131 ibidem.

103

perceptual contact with O or suffice for physical-item perception at all. The other

component will cover the remaining facts that are relevant to the securing of visual

contact with O—facts that do not involve anything further about the subject’s

psychological condition at t. In practice, [Representative Realists] will take the first

component to consist in Ralph’s having a certain kind of visual experience at t, and will

take the second component to consist in, or centrally involve, facts concerning the

qualitative relationship of this experience to O and the nature of the causal process from

O to the occurrence of the experience. [Direct Realists] too will take the seeing of O to

involve the occurrence of a visual experience and will accept that this experience occurs

at the end of a causal process starting from O. But they will take this experience to be

one which, by its intrinsic psychological nature, puts the subject into perceptual contact

with O.132

So far this concerns just Direct Realism and Representative Realism. Foster, of course, thinks

these two Realist views can be refuted. Very briefly put, the first “fails because it does not allow

an adequate understanding of the relationship between perceptual contact and phenomenal

content”; whereas the second “fails because once we accept that the subject’s contact with the

relevant physical item decomposes in the envisaged way, there is no way of understanding how

that contact can qualify as genuinely perceptual.”133

Now both of these views take it for granted “that the physical world is something external—

something with an existence distinct from, and logically independent of, facts about human

mentality.” Both assume “that our awareness has to reach to things beyond the boundaries of the

mind if it is to make contact with things in the world.” The alternative Foster recommends is to

embrace a brand of Idealism on which “there would be no difficulty in supposing that our

sensory experiences bring us into perceptual contact with physical items. For we would no

longer have to think of these items as belonging to a reality that lies beyond the realm of

experience. Awareness would not need to reach beyond the boundaries of the mind in order to

make ontological contact with the factors by which the existence of the physical world is

constitutively sustained.”134

In his earlier book Foster explains a bit more fully how Idealism of the Phenomenalistic kind

enables us to understand the way “our sensory experiences bring us into perceptual contact with

physical items.” Putting Pauline in place of Ralph and “visual sense-quale” in place of “visual

132 ibidem, 9-10. 133 Foster, A World For Us, 37. 134 Foster, A World For Us, 40.

104

experience” and assuming that Non-Idealist Direct Realism (what Foster calls Physical Realism)

has been refuted, when

Pauline -terminally sees a certain momentary stage of an apple’s surface . . . the only

item immediately before Pauline’s mind is a certain visual sense-quale, and, in the

Realist system, the only way in which the occurrence of this quale would be connected

with the relevant portion-stage would be by a complex causal process, running through

the subject’s eyes and brain. It is impossible to see how this causal connection could turn

the awareness of the quale into one which genuinely reaches to the external item. But, in

the Idealist’s system, the situation is quite different. There is still a sense in which the

contact with the portion-stage is mediated: the occurrence of the sense-quale does not in

itself qualify as awareness of the physical item. But, in the new situation, the mediating

and mediated objects are ontologically linked. For the occurrence of the quale is itself an

instance of the operation of the sensory constraints by which, on their own or in

combination with other factors, the existence of the portion-stage is constituted. In such a

case, it is no more difficult to see how the presentation of the quale succeeds in giving

access to the portion-stage of the apple’s surface than to see how the perceiving of this

stage gives access to the persisting portion, or the perceiving of this portion gives access

to the whole apple. Nor, indeed, is it difficult to see how, despite the mediation, the

access to the physical item can—in accordance with the traditional conception of

Idealism—be thought of as direct.135

There are two points to be noticed here. One is that it is just this absence of an ontological link

between mediating and mediated objects, this ontological separation between sense-qualia or

sense-data and physical items, that vitiates popular attempts to establish Representative Realism

(such as Frank Jackson’s, to be discussed in 5.3.). The other point is that there is a type of

mediation whose occurrence does not prevent the perception of the mediated item to be direct.

Such perception can be direct if the mediating and mediated items share an ontological nature;

for example, if the mediating item is a part or portion of the mediated item, or if the former is a

sense-datum and the latter is constructed out of sense-data. Hence, there can be direct perception

with mediation, or without it; but indirect perception is always with mediation.

3.5.5. Conclusion about Direct vs. Indirect Perception

135 Foster, The Nature of Perception, 256.

105

Perkins, O’Shaughnessy, and Foster would agree that (1) all mediating items (mediators) are

non-identical to the items they mediate and that (2) some but not all of these mediators are also

essentially distinct from the items mediated. They would further agree (3) that one standardly

sees a mediated item (say, a physical object) in virtue of, or by, or through, seeing a spatial or

temporal part of it (the mediating item). Finally, they would agree (4) that a mediating item

(mediator) can be more or less mediate; hence that a perception of it can be more or less

indirect.

But they might disagree on whether seeing a mediated item is always a case of indirect

perception. Thus Foster thinks that on his Idealist story seeing a physical object can still be

direct because the mediating sense-qualia are identical in nature with the object. This identity of

ontological nature would not obtain, in his view, if Representative (Indirect) Realism were true.

Since Perkins and O’Shaughnessy think that it is true, they also seem to think that any mediation

makes the perception of the mediated physical object indirect.

Cornman, we will recall, claimed that the distinction between direct and indirect perception

can be entirely “factual” or non-epistemic, as drawn by Berkeley; or it can combine both a

“factual” and an epistemic component, as expounded by Russell. But we have seen that,

contrary to Cornman, Berkeley’s distinction also has an epistemic component (see 3.5.1.). And

we haven’t yet mentioned a third way discussed by Cornman of drawing the distinction; namely,

Norman Malcolm’s, which is entirely epistemic. According to this way, also extensively

discussed by Dretske, “S sees D directly if and only if there are some properties of D about

which S cannot be mistaken.”136

It seems that Perkins’s, O’Shaughnessy’s, and Foster’s ways of distinguishing between direct

and indirect perception all hail more or less from Berkeley.

3.6. Conclusion about Seeing into the Past

Those philosophers who think that we can see into the past are divided between (1) those who

think that we can see present things and happenings in addition to seeing past things and

happenings, and (2) those who think that whenever we do see something, it is always in the past.

The first group includes, among many others, Dretske, Quinton, Armstrong, Ayer, Chisholm,

Pitcher, and Bengson; the second group includes, among others, Alexander, Johansson, Solso,

and the American New Realists. In both cases philosophers base their relevant conviction on

what science has allegedly shown. In both cases the belief that we can see into the past is also

136 Dretske, Seeing and Knowing, 62.

106

motivated by the desire to avoid postulating sense-data and countenancing idealism. We have

seen how this belief, though in some ways highly attractive, may be challenged; in the next

chapter we will argue that this challenge is insurmountable.

107

Chapter Four: Affirming the Simultaneity-Requirement

4.1. Dancy’s Objection: ‘Presence’ Is Ambiguous

Earlier (in 3.5.1.) I quoted James W. Cornman as saying that “[c]learly, if to perceive something

directly is, as Russell claims, to be acquainted with it, to have it present and presented to the

perceiver,” then the thesis that all objects exist at the time they are directly perceived “is not

only true but analytic.” Cornman claims this in the course of constructing a plausible TGA that

he thinks can be refuted only by invoking the adverbial theory of perception. We will return to

this issue in the next section. But first I want to take up a challenge to this thesis issued by

Jonathan Dancy.

He says that it is a

“mistake . . . to suppose it obvious than an object of which we are directly aware must

exist and have the qualities we attribute to it at the moment at which we are aware of it.

This is a thought about time; the awareness and its object must exist at the same time. In

[this] expression of that thought there is . . . a hint of the desire for infallibility. But

forgetting that as irrelevant, we can still ask whether the object of direct awareness must

not at least exist at the moment of awareness. Surely the direct object is presented to us;

it is directly present to us. But how can it be now present to us if it is now non-existent?

At best the non-existent must be thought of as absent (temporally and geographically),

and so not present.

There is an unfortunate ambiguity in the notion of ‘present’ here, and I suggest the

argument above trades on it. In one sense of ‘present’, the present is contrasted with the

absent (temporally or geographically). In another sense, the present is that which is

presented, that of which we are directly aware. So long as we do not trade on this

ambiguity, I see no argument to force us from one sense to the other. In the sense

defined, I suggest that an object such as a distant star can have ceased to exist by the

moment at which we are of it.137

But how can that which is not present (in the first sense), which is temporally or geographically

absent, how can that be present (in the second sense), i.e., be presented? If it’s not there, how

can it show up for us to have it be presented to us and for us to be aware of it? It just can, and I

know of no argument that it can’t, says Dancy; it just can’t, and no further argument is needed

for my position, counters Cornman.

137 Dancy, Introduction to Contemporary Epistemology, 146-147.

108

But, obviously, whether or not the latter is right, further and more probing argument is

always desirable. This I will presently try to give on Cornman’s (and my) behalf.

Next Dancy admits that his suggestion (that an object such as a distant star can have ceased

to exist by the moment at which we are aware of it) “may seem counter-intuitive. If you find it

so, you will be confirmed (I expect) in this when you notice that if it is right there will be

available a sense in which memory can be direct awareness of the past.”138

I do find it so; I do notice that if Dancy’s view is right, then “there will be available a sense

in which memory can be direct awareness of the past,” but the latter claim (that memory can be

direct awareness of the past) is precisely what I’m inclined to deny; and I’m inclined to deny it

for the reasons that A. O. Lovejoy gives in his The Revolt Against Dualism:

Intertemporal cognition, the knowing at one time of things which exist or events which

occur at another time, seems a patent example of a mode of knowledge which we are

under the necessity of regarding as potentially genuine and yet as mediate. When I

remember, for example, not only is there a present awareness distinct from the past

memory-object (that alone would imply only the duality of act and content), but the

present awareness manifestly has, and must have, a compresent content. But the past

event which we say the memory is of cannot be this compresent content.139

Lovejoy then discusses the view of those who deny that “when an event is past it ceases to

exist” and who claim, like C. D. Broad does, that “[o]nce an event has happened it exists

eternally” and that past events “are always ‘there’ waiting to be remembered; and there is no a

priori reason why they should not from time to time enter unto such a relation with certain

present events that they become objects of direct acquaintance.”140 According to Lovejoy,

[t]his view, however, implies an inconceivable divorce of the identity of an event from

its date. The things which may be said to subsist eternally are essences; and the reason

why they can so subsist is that, by definition, they have no dates. They do not “exist” at

all, in the sense in which dated and located things do so; and if “events” eternally existed

after they had “once happened” (and when they were no longer “happening”), they

would likewise exist before they happened; eternalness can hardly be an acquired

character. The present image and the past event may be separate embodiments of the

same essence; they are not identical particulars, because the particularity of each is

undefinable apart from its temporal situation and relations. The duality of the memory-

image and the bygone existence to which it refers seems to be inherent in what we mean

138 ibidem. 139 Lovejoy, 21. 140 Lovejoy, 22. The quotes from Broad, cited by Lovejoy, are in The Mind and its Place in Nature, 252.

109

by remembrance; if the two were one our intertemporal knowing would defeat its own

aim of apprehending the beyond, by annulling its beyondness. The very wistfulness of

memory implies such duality; the past, in being known, still inexorably keeps its

distance. Plainest of all is that a man’s experiencing of yesterday, the event of his then

having an experience, does not seem to him, in being remembered, to become today’s

experiencing. Common sense, however much inclined in its more self-confident

moments to believe in direct perception, has never, I suppose, believed in direct

memory; it has been well aware that what is present in retrospection is a duplicate which

somehow and in some degree discloses to us the character, without constituting the

existence, of its original.141

That this “common-sense” view of memory is natural and defensible is also implied in the

Stanford Encyclopedia’s article on Memory. This, among other things, states that

[a]lthough it takes many significantly different forms, this idea that a ‘trace’ acquired in

past experience somehow ‘represents’ that experience, or carries information about it, is

at the heart of ‘representative’ or ‘indirect’ realism in the philosophy of memory. This

has been the dominant view of memory in modern philosophy of mind, and it is assumed

in much work on memory in cognitive science. Research programmes for representative

realism thus seek to clarify the nature of representations in memory, and the various

processes in which they are involved. . . Some recent work in the cognitive sciences of

memory. . . is intended to respond to or incorporate the more powerful . . . criticisms [of

the entire representative realist framework] within revised forms of representative

realism.

Thus because I think it is highly respectable to reject Direct Realism, and to accept Indirect

Realism, with respect to recollective awareness of the past, I continue to take it as counter-

intuitive that any object can go out of existence before we become directly visually aware of it.

This is one, but perhaps not a conclusive, argument against Dancy’s position. Others will

emerge presently.

4.2. Cornman and the TGA

Some of the most sustained work on the TGA and its simultaneity-principle has been done by

James Cornman142 in two books, in which tackling these issues is part of a broader philosophical

141 ibidem. 142 He, too, has called the TGA “the most forceful argument for sensa derived from causal facts of perception” (Perception, Common Sense, and Science, p. 47). (“Sensa,” of course, is his term for what others call “sense-data.”)

110

program. One of these books, Materialism and Sensations, dating from 1971, attempts to

provide a materialist account of sensations by first rebutting Reductive Materialism and

Eliminative Materialism and then arguing for Adverbial Materialism, which in turn requires

rejection of sense-data and acceptance of an adverbial construal of sensations. The other book is

the 1975 volume Perception, Common Sense, and Science, which Dancy in 1985 called one of

“the best available on the philosophy of perception” and one to which his own “discussion is

much indebted”.143 In that second book Cornman examines perception of the external world and

advances a metaphysical theory he calls “compatible common-sense realism.” Throughout both

these works the author emphasizes that his conclusions, though painstakingly (if not always

perspicuously) argued-for, are only “tentative and provisional.”

He presents versions of the TGA in both books, including two versions that he regards as

plausible, to say the least. The first version, presented in the earlier book, argues for the

existence of sense-data. Though Cornman does not accept the conclusion, he praises the

argument as strong. The second version, presented in the later book, is directed towards proving

the falsity of naïve realism; Cornman regards it as entirely successful. I will discuss them in

turn; but first a few words about some of his many discussions of the simultaneity-principle.

In Materialism and Sensations he states two versions of it:

(a) It is impossible that something is perceived at a time it does not exist.

(b) All objects exist at the time they are perceived.

Cornman argues against these as follows. With respect to (a) he asks,

But why should anyone think that it is logically impossible? Whether or not it is

reasonable, it is at least logically possible that we see a star at the time it no longer

exists. There is no contradiction entailed here.144

But this is just a vigorous denial of (a), not a strong argument for rejecting it. And (b), he

asserts,

is unreasonable, because it is quite reasonable to claim that some stars are perceived

when they no longer exist.145

Once again, this is just a bare denial of the claim at issue. Fortunately, Cornman doesn’t leave it

at that but goes on to say that this “claim is bolstered by construing perceiving” adverbially; that

He goes on to say that “[p]hysiologists, such as Brain and Eccles, have described these complex causal processes as culminating in experiences of sense-data or percepts, but they seem to provide no reason for preferring this description to one that mentions adverbial sensings. . . . I should think that it is at least as reasonable to hold that adverbial sensing events are effects of neural events as to hold that these effects are phenomenal objects or properties.” 143 Dancy, 158. 144 Cornman, Materialism and Sensations, 220. 145 ibidem, 221.

111

is, by taking it that “the event of a person perceiving a star is the event of the star appearing in

some way to him as a result of the star being a proper stimulus of him. This would allow for

people sometimes perceiving stars when the stars do not exist.”146

In Perception, Common Sense, and Science he elaborates on this as follows:

Often, when someone claims that we directly see a physical object, such as a hand, he

takes this to imply that we are perceptually acquainted with the hand, that is, that the

hand is not only presented to us when we see it, but it is present with us when we see it.

This is clearly the view of the perception of physical objects that a naïve realist and,

initially at least, a Berkeleian would have, and it is the view of the experience of sensa

that an indirect realist and a Berkeleian have. However, . . . it is not required for direct

perception of physical objects. If my present perceptual experience of a distant star

consists in my being in a state of star-sensing now as a result of my eyes being suitably

affected by stimulus previously transmitted from the star, then I directly perceive the star

now, even if it does not exist now and I am not perceptually acquainted with it. And . . .

the causation of such a sensing event is at least as reasonable an effect of physical and

neurophysical events as is a sensum. Thus the causal facts involved in the time-gap . . .

provide no reason to accept the existence of sensa.147

Therefore, Cornman is saying, there are two ways of dealing with the time-gap. The one he

prefers denies the simultaneity-principle and, accordingly, denies sense-data but affirms

adverbial sensings and concludes that stars are perceived directly. The other affirms the

simultaneity-principle and concludes that stars are perceived indirectly because the objects of

direct perception are the existing sense-data.

Although admitting the plausibility of the second way, he wishes to reject its affirmative

conclusion about sense-data. Since this is a conclusion I wish to defend, let’s look at how

Cornman proposes to reject it; but for that we first need to lay out the argument in detail. This is

the one for sense-data found in Materialism and Sensations. He introduces it by saying that

“[n]ot everyone will agree that perceiving something is to be identified merely with having a

perceptual experience of a certain kind. To perceive an object, especially to see it or touch it, is

seemingly to be in some kind of contact with it, to have it presented to one. But this kind of

relationship to an object requires its existence.”148 Here is the argument (let me call it the MS

Argument):149

146 ibidem. 147 Cornman, Perception, Common Sense, and Science, 49-50. 148 Cornman, Materialism and Sensations, 221. 149 ibidem, 219-226. All further quotations from this book are from these pages.

112

(1) If something is directly perceived, then it is directly perceived during the time a

veridical perceptual experience of it occurs.

(2) Any veridical perceptual experience of an external physical object occurs after the

stimulus energy transmitted from the object first affects a sense organ.

Therefore,

(3) If an external physical object is directly perceived, then it is perceived after the

stimulus energy from it first affects a sense organ.

(4) If some external physical objects are directly perceived after the stimulus energy

from them affects a sense organ, then some external physical objects do not exist at the

time they are directly perceived.

(5) But all objects exist at the time they are directly perceived [the simultaneity-

principle].

Therefore,

(6) No external physical objects are directly perceived.

(7) If no external physical objects are directly perceived, then either no external physical

objects are perceived, or some are indirectly perceived and sense-data are directly

perceived.

(8) Some external physical objects are perceived.

Therefore,

(9) Sense-data are directly perceived.

Therefore,

(10) Sense-data exist.

Of course, someone using the TGA to argue seriously for sense-data would reformulate

premise (4), perhaps by replacing its consequent “some external physical objects do not exist at

the time they are directly perceived” with the consequent “some external physical objects do not

exist at the time it seems as though they are directly perceived.” But since Cornman’s intention

is the opposite, he accepts (4) as it stands and can afford to declare that in this argument “[n]o

premise is clearly dubious” and that “[t]he only way” to avoid its conclusion “is to claim that

there are cases of perception of physical objects in which there are no objects of acquaintance,

because no objects are given or presented in those perceptual experiences.”

For Cornman this means rejecting premise (7). “This may,” says he, “appear to be

egregiously implausible, but it may seem less outrageous if it is realized that this is merely a

claim about the factual character of perception and is compatible with something always being

given in a purely epistemological sense.” This remark seems obscure; I will come back to it.

113

Then Cornman goes on to say that “[t]here remains, however, the problem of how it can be

that some things are perceived when nothing is directly perceived.” His answer consists in a

highly involuted discussion, as a result of which premise (5) is now held to have been “shown to

be dubious by the fact that it is surely plausible to claim that certain stars are perceived at times

they do not exist, because perceiving them consists in star-sensing as a result of stimulus energy

from the stars and some of these stars do not exist when the star-sensings occur. And with no

intermediaries, they are directly perceived in a Berkeleian sense but not in an acquaintance

sense.” The latter requires physical objects to be “present and presented in perceptual

experiences,” so that “a direct perceptual realist who subscribes to the adverbial theory . . .

might even go on to argue that no objects of any kind are perceptually given in this sense.

Because perceptual experiences are adverbial and because of the time gap, no external physical

objects nor any other objects are presented.”

But since, although these objects are not perceptually given (being neither present nor

presented) they nevertheless are seen directly, what might be “given in a purely epistemological

sense” is the knowledge that (say) a star is or was there. Perhaps this is what Cornman had in

mind with the obscure remark cited above.

Even so, his reasons for rejecting (7) seem unconvincing. I find (7) to be obviously true.

Therefore, in view of the fact that both Cornman and I find none of the other premises “clearly

dubious” (to say the very least), I conclude that the MS Argument, with premise (4) as amended

above, is a fine TGA for the existence of sense-data.

In Perception, Common Sense, and Science he lays out a somewhat similar TGA (call it the

PCSS Argument) for a somewhat different conclusion:150

(1) If something is perceived by s at a time t, then t is one time at which s has a veridical

perceptual experience of it.

(2) No veridical perceptual experience of an external physical object by s occurs before

the stimulus energy transmitted from the object first affects a sense organ of s.

Therefore,

(3) If an external physical object is perceived by s at t, then t is not a time before the

stimulus energy from the object first affects a sense organ of s.

(4) Some external physical objects that are perceived by a person, s, have ceased existing

before stimulus from them first affects a sense organ of s.

Therefore,

150 Cornman, Perception, Common Sense, and Science, 243-247. All further quotations from this book are, if not indicated otherwise, from these pages.

114

(5) Some external physical objects are perceived by a person, s, at a time when the

objects do not exist.

(6) If naïve realism is correct, then if a person, s, perceives an external physical object at

t, then he is perceptually acquainted with it at t.

(7) If a person, s, is perceptually acquainted with an object at t, then it exists at t and, if s

should be normal and conditions optimal, s would experience it approximately as it is at

t.

Therefore,

(8) Naïve realism is incorrect.

The simultaneity-thesis, of course, is implied by premise (7). Of this argument Cornman says

that it is

is valid and it certainly seems sound. Surely a naïve realist will not refute it by attacking

premises (1), (6), or (7). They are not only clearly true, but are true by definition. His

only hope, then, is to attack one of the two empirical premises, (2) and (4). But (2) is an

obvious truth and, moreover, is surely accepted by a naïve realist. . . . He must,

consequently, concentrate on (4).

To refute premise (4), a naïve realist must counter the strong scientific reason for

accepting it. There is good reason to believe that exploding stars have been seen, and

that, given the scientific hypotheses about their distances from the earth and about the

speed of light in outer space, they exploded light years before their explosion was seen.

He must, then either refute the claim that they are at such great distances, or justify that

the speed of light in outer space is substantially greater than 186,000 miles per second.

Cornman is right, of course, in firmly rejecting both alternatives mentioned in the last sentence.

But is there really “good reason to believe that exploding stars have been seen” and that “they

exploded light years before their explosion was seen”? Yes, there is good reason to believe—

and natural to say—these things. But can’t this same good reason be accommodated by

believing and saying that sense-data of exploding stars have been seen and that they have been

seen light years after the stars exploded? I don’t see why it can’t except for a wish to avoid

sense-data. If one lacks this wish then one can rewrite (4) as

(4a) Some external physical objects whose sense-data are perceived by a person, s, have

ceased existing before stimulus from these external

objects first affects a sense organ of s.

Now (5) doesn’t follow (it’s false anyway), and this argument against naïve realism

collapses, but there are others that are much better (such as our Berkeleian TGA). Thus when in

his “conclusion about naïve realism” Cornman says that “[w]e have finally come to agree with

115

the many philosophers who have claimed that naïve realism should be rejected,” I can assent to

the last clause—but only partially and not in the sense Cornman intended.

Indeed we must be careful here. What is naïve realism, according to Cornman? He says that

it comprises two theses. The first is that

the world consists in large part of perceivable objects that are not perceivers, that exist

unperceived, and that are not affected either by being perceived or, generally, by changes

in the usual conditions in which they are perceived; many of these objects have and are

perceived to have . . . the observable, occurrent physical properties of size, shape,

weight, solidity, texture, motion (rest), location, sensuous color, and hotness

(coldness).151

This, the thesis that most physical objects exist and have many of their properties

independently of the mind, is not refuted by the PCSS Argument. And that it is not so refuted is,

indeed, why he actually endorses this version of the TGA, for Cornman, like most modern

philosophers, is emphatically not an Idealist.

But it is the thesis that Idealists (and therefore I) wish to refute; consequently they (and I)

cannot accept this argument.

Moreover, we cannot accept it because what this argument refutes is the second thesis,

namely, that “to perceive a physical object is to be . . . perceptually acquainted with it.”152 This

is precisely the thesis that Idealists wish, not to refute, but to preserve.

Cornman defines perceptual acquaintance “in terms of being perceptually presented to and

present with a perceiver,” and perceptual presentedness (presentation) and perceptual presence

are, in turn, defined by him as follows:153

p is perceptually presented to s at time t = df.

(1) s immediately perceives p at t, and

(2) if, at t, s should be a normal perceiver who immediately perceives

p in normal conditions of perception, the p would appear to s at

t as it is at t.

p is perceptually present with s at time t = df. p exist at t when s

immediately perceives p.

This is the sense of “perceptual acquaintance” in which the Direct Realist says physical

objects can be directly perceived and the Representative Realist (Indirect Realist) denies that

they are, claiming instead that physical objects, if perceived at all, are perceived indirectly. The

151 Cornman, Perception, Common Sense, and Science, 219. 152 ibidem, 224. 153 ibidem, 223,

116

Idealist Direct Realist also accepts that physical objects can be directly perceived in this sense,

although he calls attention to the fact that there may be two meanings of “directly” that the

Direct Realist and the Indirect Realist can make use of here.

4.3. Valberg on Presence

One philosopher who has well-articulated views about presence and related concepts is Jerry

Valberg. He says that “[b]y an ‘object of experience’ we shall mean something present in

experience.”154 But, he asks, What do we “mean by ‘present in experience’, ‘present to us’, and

so on”? Here’s the first part of Valberg’s answer:

Presence (in experience) connotes a kind of direct or immediate availability. An object

which is present is right there, available to us. This makes it tempting to view presence

as the reciprocal of Russell’s idea of acquaintance. That is, an object with which we are

(in Russell’s sense) acquainted is present in experience; and an object which is present in

experience is one with which we are acquainted.155

Valberg does not resist this temptation; he endorses this tie between acquaintance and presence,

since “both ideas imply immediate availability”; but there is one difference: whereas for Russell

objects of acquaintance “may be either ‘particulars’ or ‘universals’,” for Valberg “experiential

presence is limited to the category of particulars. What is present, directly available, is always a

particular.”156

That’s the first point about presence, says Valberg, and continues:

But since particulars traditionally “are temporal (but not necessarily spatial) objects . . .

of which it makes sense to ask how long they have existed or lasted, when they began to

exist, and so on . . . , the second point to make about presence in experience is than an

object present in experience is always a temporal object.157

A third point about presence is “the intimate connection between experiential and temporal

presence”:

We have a strong inclination to view objects that are present in experience as being, at

that very time, existent: if something is now present, it now exists. This inclination to

encompass the object within the temporal present does not extend to reference, or

thought in general. There is no problem about referring to, or thinking of, objects which

154 Valberg, The Puzzle of Experience, 4. 155 ibidem. 156 ibidem, 5. 157 ibidem.

117

no longer exist. But our inclination is to say that such things cannot be present in

experience, that they cannot now be objects of experience.158

Having said this, Valberg then adds the following all-important footnote:

The reason for saying (guardedly) that we are ‘inclined’ to view experiential and

temporal presence as connected in this way, rather than flatly asserting that they are thus

connected, is that the connection is sometimes used by philosophers to argue . . . that the

object of experience is always internal. I have in mind the famous ‘time-lag’ argument.

Some philosophers, when they see this coming, claim that the connection between

experiential and temporal presence is only apparent. The case of stellar explosions, they

will say, shows that things (events) can be present in experience, after they cease to

exist. There is no point getting into this issue, since, for the purpose of providing a

distinguishing mark of experiential presence (for the purpose of distinguishing objects of

experience from, say, objects of reference), the appearance of a connection suffices.159

But why, in Valberg’s view, does the (mere) appearance of a connection suffice? Why, for the

purpose of distinguishing objects of experience from objects of thought, isn’t it necessary that

the objects present in experience actually be (rather than just appear to be) at that very time

existent? We will pursue this question in due course.

For now just a few more words of clarification about “the ‘direct availability’ that objects

have in virtue of being present.” This involves two ideas: “that of ‘focusing on’, ‘fixing on’, or

‘picking out’ an object; and that of ‘demonstrative reference’ to an object.”

According to Valberg, “we cannot focus on something unless it is present in our experience:

the fact of an object’s presence is what makes the object available for us to focus on. Thus an

object which is present in experience is, in virtue of that fact, available (then and there) for us to

focus on and pick out.”160

And “when we make a demonstrative reference to something, we thereby” do “focus on

(pick out) the object to which we refer.” So “just as we can think of the fact of an object’s

presence as what makes the object available for us to focus on, or pick out, we can think of this

latter fact as what makes the object available for demonstrative reference (what makes the

object demonstratively available). A fact of presence, we might say, is what creates the

possibility of focusing, and what creates the possibility of focusing creates the possibility of

demonstrative reference.”161

158 ibidem. 159 ibidem, 19. 160 ibidem, 6. 161 ibidem, 7.

118

Thus, an object of experience is “something present in experience: something which is right

there, available for us to pick out or focus on, and refer to demonstratively.”162

Next I would like to bring out how much Valberg values the TGA. In the footnote referred to

several paragraphs above, he goes on to say that it “has much in common with the problematic

reasoning,” which is what Valberg calls the argument leading into his “puzzle of experience.”

Both the TGA and Valberg’s “problematic reasoning” depend on what he calls “the causal

picture of experience,” which is a set of “basic facts about the transmission and reflection of

light, the nature of the eye, and the process in the nervous system and brain which results or

culminates in visual experience.”163 (It is what we outlined, using Dretske’s words, at the

beginning of 2.2.). Valberg continues:

It is worth noting, however, that, even if the time-lag argument were correct, the

problematic reasoning is more fundamental. It depends on less; it does not depend on the

fact of a time-lag. The problematic reasoning would work equally well if light travelled

instantaneously.164

So what is Valberg’s “problematic reasoning”?

4.4. The Puzzle of Experience

Here is Valberg’s “problematic reasoning” in three discursive stages.165 The first stage is the

thought that “[t]he present activity in my brain . . . is the causal outcome of a chain of events

involving light rays being reflected from a certain object—the book on the table, say. In turn,

the activity in my brain is causally responsible for my experience being as it now is.”166

The second stage is the reflection that

if the activity in my brain could somehow be held constant, the earlier parts of the causal

chain might be eliminated without this having any effect on my experience. If the

activity in my brain were to continue as it is, my experience would continue as it is. It

would continue as it is, even if, say, something interfered with the light rays being

reflected to my eyes; or even if the object reflecting the light rays were miraculously

annihilated. God (it is handy to bring God in here) might arrange such a situation. God

might eliminate the object while maintaining the activity in my visual cortex. Is that not

a coherent state of affairs, something God might bring about? And, if the activity in my

162 ibidem. 163 ibidem, 9-10. 164 ibidem, 19. 165 He takes much longer to expound it than I do here. 166 Valberg, The Puzzle of Experience, 10.

119

brain were to continue on as it is, my experience would also continue on as it is. It would

continue to run along perfectly smoothly, despite the fact that the external object would

no longer be there.

This possibility, which is implicit in our everyday knowledge about the causal dependence

of experience on objects and processes in the world (implicit, that is, in the causal picture of

experience), might be summed up by saying that the external thing, the object in the world, is

‘potentially irrelevant’ to experience.”167

Now comes the third stage. It consists of four steps.

The first is to look at the book on my table and to focus “on whatever it is that is present

now in my experience while I am looking at the book. Certainly there is, right now, something

present in my experience.”168

The second step is to “remain focused on it for a little while (about five seconds should be

enough). I do this very closely, never letting my attention waver, always staying right with the

object on which I am focused.”169

The third step is to “reflect on this bit of my history in the light of the potential irrelevance

of the book. Consider the hypothesis that half-way through the last five seconds God . . .

eliminated the book but maintained the activity in my brain just as it was when the book was

there. When the book was there, it reflected light to my eyes; this process eventuated in the

activity in my visual cortex. The hypothesis is that, half-way through the last five seconds, God

(as it were) took over from the book, having eliminated it, and directly maintained the activity

in my brain.” But in reflecting on this hypothesis I realize “that it seems to be compatible with

how things have been in my experience during the last five seconds . . . with the fact that this

object, the object on which I am now focused, has been present in my experience for the last

five seconds,” even though, “[h]ad God intervened, the book would have ceased to exist two-

and-a-half seconds ago” but “this object would have remained (just as it has remained) present

in my experience. There would have been a rupture in the world; a certain external object, a

book, would have suddenly ceased to exist. This would have been evident to an onlooker, but

for me, within my experience, things would have flowed smoothly on, without a ripple or

flicker—just as they have done. That is to say, this object would have been (just as it has been)

present to me for the whole of the last five seconds.”170

The fourth step is to draw the obvious conclusion.

167 ibidem, 10-11. 168 ibidem, 14-15. 169 ibidem, 15. 170 ibidem, 15-16.

120

It follows (plainly) that this object on which I have actually been focused for the last five

seconds, is not the book. For my having been focused on this object is compatible with

the supposition of God’s intervention, and it is part of the supposition that, half-way

through the last five seconds, God eliminated this book. So the object on which I have

actually been focused, this object, is such that it might have survived the elimination of

the book. So it cannot be the book.”171

This, the probematic reasoning, “is not restricted to the object on which I am now focused

and the book in front of me on the table. It works for any object in the world and any object on

which I might focus. Any object in the world, any external thing (apart from my brain), is such

that God could eliminate it compatibly with keeping the activity in my brain the same. . . . With

respect to anything present in my experience . . . I can reduce to absurdity the assumption that it

is part of the world.”172

But if “the object present in my experience is never part of the world, never an object whose

existence is independent of the fact of its presence,” then it must be “an object whose existence

is not independent of its presence in my experience, an internal object,” a sense-datum. This,

then, is the conclusion of the problematic reasoning.173

“On Russell’s use of the term ‘sense-data’, which has become pretty standard,” Valberg

claims in another important footnote,

sense-data are by definition internal objects. Russell’s procedure is first to argue that the

objects of experience are internal, and then (in effect) to dub these internal objects

‘sense-data’ . . . But Russell’s usage is not universal. Moore, unlike Russell, seems to

define the term ‘sense-data’ . . . in such a way as to leave it open whether sense-data are

external or internal. In effect, ‘sense-datum’ for Moore means: that which is present in

experience (object of experience). The question for Moore, about which he can never

quite make up his mind, is whether sense-data (thus defined) are or are not external

things (or parts of external things) . . . Thus, if we express the conclusion of the

problematic reasoning by saying that the object of experience is always a ‘sense-datum’,

this would accord better with Russell’s use of the term than with Moore’s.”174

There is one troubling aspect of Valberg’s thought regarding internal objects and sense-data

that I now wish to call attention to. He says that

[a]n ‘internal object’ is an object whose existence is not independent of its presence in

experience. In this case, existence and presence collapse into one. The fact of existence

171 ibidem, 16-17. 172 ibidem, 17-18. 173 ibidem, 18. 174 ibidem, 19-20.

121

and the fact of presence are the same fact. Examples of internal objects are after-images

and hallucinatory objects (such as Macbeth’s dagger). Thus we do not regard such

objects as part of the world.175

But this last sentence (introduced by “Thus”) doesn’t seem to follow from the previous ones;

and it seems in any case to be unargued-for and false. An argument of sorts appears in the

immediately following passage in which Valberg makes the familiar distinction between

material objects, which “are external,” space-filling, and “part of the world,” and “immaterial

objects,” which likewise are “part of the world”:

The world will contain, as Descartes thought, two radically different categories of

objects and phenomena. Some will consist of matter, others of spirit. Whether Descartes

was right in this view is not relevant here. What is relevant is simply that internal objects

are not immaterial objects. (An after-image is not like a ghost.) Internal objects are

neither material nor immaterial. They do not consist of anything. They are not part of the

world.176

Here Valberg barely conceals his disapproval of the Cartesian partition of the world into

matter and spirit. I, like Foster, think there is nothing wrong with this partition; I also think that

ghosts are not the best examples of spiritual things, that ghosts almost certainly do not exist, and

that souls do (thus I would reject Gilbert Ryle’s characterization of the soul as a “ghost in a

machine”). The best examples of spiritual things are thoughts (thinkings), feelings, emotions,

mental images, after-images, sensations, and sense-data: these are all mental, almost certainly

immaterial, and definitely part of the world. That they are not part of the world is plausible only

on the assumption that the world is wholly material, but this assumption is daring (it must be

strenuously argued-for) and most likely wrong.

However, the problematic reasoning is only one half of what Valberg calls “the puzzle of

experience.” The other half comes about by simply “being open to our experience.” When you

open up to your experience, this is how the conflict, the antinomy, the puzzle of experience,

presents itself:

I start by going through the problematic reasoning. I conclude that this object, the object

present to me when I look at the book, cannot be the book. It cannot be the book

because, by the reasoning, it could survive the elimination of the book. So it, this object,

is an internal object, something which exists only in so far as it is present in my

experience. But wait, this object is a book. The object present to me when I look at the

book on the table is the book on the table. There is nothing else there. Now I realize that,

175 ibidem, 8. 176 ibidem.

122

as a contribution to philosophy, thoughts of this sort may appear a trifle quick and

simple-minded; yet it is precisely such thoughts that come over me when I reach the

conclusion of the problematic reasoning. And when they come over me, they totally

overcome the conclusion. The reasoning establishes that this is an internal object. But it

is not. It, this object, is a book.

What gives me the right to say this? Where is the argument? There is no argument.

The arguments are all on the other side. I do not conclude that the object present to me is

a book, but that is all I find—the book. That is to say, the book is all I find when I am

open to how things are in my experience. If I reflect on the fact that how things are in my

experience is the causal product of what is happening now in my brain, I seem driven to

the conclusion that the object present in my experience when I look at the book, this

object, is not a book. Yet if I am open to (how things are in) my experience, all I find is

the book.

Thus the antinomy is a function of the fact that my experience is a subject-matter on

which I can reflect in two different ways. There is the indirect way: reasoning, in terms

of the causal picture of experience, to a conclusion about how things are in my

experience. And the direct way: simply being open to how things are in my

experience.177

That’s Valberg’s puzzle of experience: it arises from first reasoning about our experience (in a

way reminiscent of but much more concise than the TGA) and then becoming open to our

experience. Having presented and articulated this puzzle, Valberg goes on to explain further

how it arises and then to look for possible solutions to the puzzle. A solution would be finding a

mistake in the problematic reasoning or in our attitude to the reasoning. But he finds no

mistake; and thus concludes that the puzzle of experience is unsolvable, that we are “stuck”

with it.

I will argue in my final chapter that there is one solution to the puzzle that Valberg has short-

changed a bit: namely, the theistic Phenomenalist (Idealist) one. But let us first look in greater

detail at Berkeley’s position regarding perceivings and objects perceived, and the time at which

perceivings occur.

4.5. Berkeley on Perceivings and Perceiveds

At Principles, Section 5 Berkeley says that “it is impossible for me to conceive in my thoughts

any sensible thing or object distinct from the sensation or perception of it.” This, plus the

177 ibidem, 21-22.

123

statement in the first edition that “in truth the object and the sensation are the same thing and

cannot, therefore, be abstracted from each other” and the PC585 entry “wherein I pray you does

the perception of white differ from white” all suggest a principle that S. A. Grave and Colin

Turbayne have called the Identity Principle; but A. C. Grayling thinks that this label is “wholly

misleading,” since Berkeley, according to Grayling, is not committed “to the view that ideas are

identical with perceivings of them.”178

If that were Berkeley’s view, the conclusion that if a thing is perceived at a certain time then

it must exist at that time would readily follow. But that’s not his view, acccording to Grayling:

Berkeley is not claiming

that a particular idea is to be identified with a particular perceiving of it, for the non-

identity of states of awareness and their contents is a thesis already well entrenched as

the view that, from the finitary viewpoint, ideas, at any rate those that constitute real

things, are not dependent for their existence on any particular finite perceiving of them.

What it means not to be able to ‘conceive apart’ any ‘sensible thing or object distinct

from the sensation or perception of it’ (P5) is therefore far from a claim that the idea and

the perception of it are numerically the same thing, but rather that an idea is always and

essentially an idea perceived, and perception is always and essentially the perception of

ideas.179

So the question is as follows: For Berkeley, are sensations and their objects merely non-

abstractable one from the other, as Grayling claims, or actually numerically identical, as

Turbayne claims?

I think that this specific objection to Turbayne’s reading misfires. When Grayling says that

“from the finitary viewpoint” ideas that constitute real things “are not dependent for their

existence on any particular finite perceiving of them,” this is true, of course, if ‘dependent’

means ‘causally dependent’ (and in this context it cannot mean anything else). Now, for

Berkeley, all ideas are particular and of particulars, and my idea of the visible sun can be like

yours and a lot of other finite observers’, yet all of these ideas are in some way ontologically

dependent on God (a point not only not denied, but positively emphasized, by Grayling). Which

way that is, scholars dispute about: does God Himself perceive ideas like those of finite

observers; or does He have a ‘volitional policy’ vis-à-vis finite observers, and if so, how does

that work? But none of this in the least implies “the non-identity of states of awareness

[particular perceivings] and their contents [particular ideas perceived],” as Grayling contends.

On the contrary, the non-dependence of any particular idea’s existence on any particular finite

178 Grayling, Berkeley, 169-170. 179 ibidem, 171.

124

perceiving of it seems to be perfectly compatible with any particular perceiving of it being

numerically identical with that particular idea’s being perceived; just as this identity claim is

compatible with Grayling’s own, weaker reading of Berkeley “that an idea is always and

essentially an idea perceived, and perception is always and essentially the perception of ideas”.

Still, I think Grayling is absolutely correct in thinking there is something wrong with the

view (whether Berkeley held it or not) “that a particular idea is to be identified with a particular

perceiving of it.” I will discuss this important point a bit later; first I briefly want to take up

another thought of Grayling’s, namely, that there are many examples of distinct items that are

related by mutual non-abstractability.

In his words, “[e]very process-product relation, for example, is of this kind; bread and the

baking of it, a picture and the painting of it, are cases in point.”180 Obviously, the loaf of bread

cannot exist before it has been baked nor can the baking of it come after it, the loaf, has ceased

to exist. But is the perception of a thing a process-product relation?

According to Belfrage, that’s exactly what it is in Berkeley’s Theory of Vision, where he

“analyzes the perceiving process, not the end-product, so-called ‘sensible things’, which play a

crucial part in the Principles.”181

In Grayling’s view, “the natural model for Berkeley’s account of perception” is neither the

process-product model nor even the act-object model

because finite minds are passive in perceiving real ideas, which are input from an

independent external source. Their passivity is embedded in activity, however, in the

sense that to receive an idea involves having, say, to turn one’s eye or sniff at a rose

(1D196). But although to this extent perception of real ideas involves activity, the

content of a particular state of awareness is not produced by an act of will, as in the case

of ideas of imagination, but is ‘somewhat consequent’ to eye-turnings and sniffings

(ibid.), and therefore to be distinguished from them (1D197).182

The best model, Grayling implies, is, therefore, the “state-content model”; and it is just in terms

of such a model that the Robinsonian argument (for the premise that if I see something now then

it must exist now; see 2.1) has been formulated.

If Turbayne’s interpretation of the ‘Identity Principle’ in Principles, section 5 is correct, i.e.,

if the thing immediately seen is identical with the seeing of it, then, of course, it just follows that

if I see something now it must exist now. The problem here is not only that if someone doesn’t

already accept the simultaneity-requirement he might be even less inclined to accept this as a

180 ibidem, 173. 181 Belfrage, The Scientific Background of George Berkeley’s Idealism, 217. 182 Grayling, Berkeley, 173.

125

reason for it, but also, and more importantly, that this is just dead wrong. A thing, whatever it is,

can no more be identical with the seeing of it than a horse can be identical with the kicking of it.

A headache, of course, is identical with the feeling of it; but this in no way warrants the claim

that this pain (this headache, this feeling of it) is identical with the head that aches. The

headache is only like the head in that both are mind-dependent.

Tom Stoneham argues forcefully in this connection that Berkeley is best understood as

adhering to the subject-object model of perception, to be distinguished from the act-object

model (which, according to Stoneham, is better called the ‘subject-act-object model’ to reveal

all its elements). In addition to these two, there are three more models relevant to discussions of

Berkeley: the adverbial model; the Representative Theory’s subject-object-object model, which

Berkeley tries to refute; and the intentional model, which he never explicitly discusses or even

acknowledges.

The subject-object model, which Stoneham also calls SMP (The Simplest Model of (Sense)

Perception), is introduced by him in the context of “the debate about whether Berkeley holds an

act-object model of perception or an adverbial model”; according to the latter the distinction

between act and object collapses. As Stoneham writes,

It is standardly assumed that if this distinction collapses, we are left with just the act, the

perceiving, and consequently the difference between seeing red and seeing green must be

the difference between seeing redly and seeing greenly. But a collapse can go either way:

the object could collapse into the act or the act into the object. SMP holds that we do not

have a duality of act and object, but merely an object (and a subject, of course, but the

identity and properties of the subject are irrelevant to the content of the perception, to

what is perceived). This alternative is obscured by the universal assumption that if S

perceives O, then there is a (mental) event which is a perceiving of O by S. SMP simply

denies this: when S perceives O, S and O exist, and they stand in a relation, namely

perceiving, but their standing in this relation is neither constituted nor enabled by any

concurrent event or occurrence in S.183

Their standing in this relation is a state. That state, though (obviously) not constituted or enabled

by any (other) event or occurrence in S, is trivially constituted by itself, by S’s perceiving O; and

it exists as much as S and O do. There being this state of perceiving makes SMP a case of the

state-content model (where content=object) preferred by Grayling (again, it might be more

perspicuous to call it the ‘subject-state-content model’).

183 Stoneham, Berkeley’s World, 54-55.

126

Thus SMP is committed to maintaining that when S perceives O there is a subject of

perceiving, a relation of perceiving, and an object or content perceived. Now every model of

perception except the adverbial one must agree that there is between S and O a relation of

perceiving. Thus it is clear that SMP is incompatible with Turbayne’s account of Berkeley,

according to which the perceiving and the perceived are the same (unless the perceptual relation

between S and O is not S’s perceiving of O; or that perceptual relation is identity; or there is no

perceptual relation between S and O at all. But all three options are absurd.) This is another

thing wrong with Turbayne’s reading, assuming that SMP really is Berkeley’s “default account

of perception,” as Stoneham claims (correctly, in my opinion).

On Stoneham’s account of Berkeley, if S’s perceiving of O is realized at time t, then

obviously S must exist t; but doesn’t the same go for O? I would say so, but some philosophers

(e.g., John Foster) claim it is logically possible for a perceptual relation to obtain now between a

currently existing S and a past item O. I will return to this issue shortly (in the next section).

If Grayling’s interpretation of Principles, section 5 in terms of non-abstractibility is correct

(something very likely given the failure of Turbayne’s), then we do not get from that alone the

conclusion that if I see something immediately now it must exist now. It could have existed

earlier: what I immediately see now is something that existed earlier and was then also

immediately perceived, of course; but not necessarily by me.

However, if perceiving is a process-product relationship, as Grayling has Berkeley allowing

“for ideas of imagination and for all God’s ideas”184 and Belfrage has Berkeley allowing for all

human ideas that constitute sensible objects, then of course the object seen cannot cease to be

before the seeing of it, but must come to be some time during the seeing of it (most likely,

toward, or at, the end of it). And very often, the object endures after its production. Thus at least

Belfrage’s Berkeley supports the premise we need.

4.6. Berkeley and the Simultaneity-Requirement

John Foster, as we saw, is one philosopher who says he lacks the intuition that perception and

perceived must at least be partly simultaneous. He also claims that insisting on this requirement

would mean begging the question against the Non-Idealist Direct Realist.

But why can’t we take a hard line here and say the simultaneity-requirement is so obvious

that it’s the Non-Idealist Direct Realist who’s begging the question against both the Indirect

Realist and the Idealist Direct Realist? If the speed of light had been infinite, or as long as its

184 Grayling, Berkeley, 172.

127

finitude had not dawned on people, nobody would have questioned the simultaneity requirement

on perceivings and perceived items in the first place. I think that this is how Berkeley would

have seen things. And here’s my argument.

In Dialogues 238, when Hylas brings up the oar with one end in the water and asks

Philonous how a man can be mistaken in thinking it crooked, Philonous gives a long reply, from

which it is evident that Berkeley did take the simultaneity-requirement to be achingly self-

evident, so much so that he (like Russell about two hundred years later) neglected to spell it out

explicitly. (This is in line with Robert J. Fogelin’s observation that “Berkeley’s grounds for

committing himself to idealism involve a direct appeal to intuitions concerning the nature of

things we encounter in experience”.185) I quote each of Philonous’s sentences and follow it with

my elaboration of what I take to be Berkeley’s relevant intuition.

Philonous: “He is not mistaken with regard to the ideas he actually perceives; but in the

inferences he makes from his present perceptions.”

What are they, these present perceptions? The perceivings he now has, or the ideas he now

perceives? For Berkeley these presuppose each other. In the realm of ideas, every perceiving is

of an idea; and every idea is actually perceived. The man’s present perceptions include both his

perceivings and the ideas he perceives, about which, Berkeley emphatically tells us, he cannot

be mistaken. His perceptions are not just his states of perceiving; they also contain the objects of

perceiving, what he actually perceives; and it isn’t—and most likely also can’t be—the case that

he is at present perceiving something and that thing is not present but past; the perceiving and

the perceived both occur at present, that is, now.

Philonous: “Thus in the case of the oar, what he immediately perceives by sight is certainly

crooked; and so far he is in the right.”

The perceived thing is, not was, crooked. It could have been, but needn’t have been, crooked

before; but now, at the time of perception, it certainly is crooked.

Philonous: “But if he thence conclude, that upon taking the oar out of the water he shall

perceive the same crookedness; or that it would affect his touch, as crooked things are wont to

do: in that he is mistaken.”

So if he takes the oar out of the water, in the immediate future, he will perceive not

crookedness but straightness; now, not having taken the oar out of the water yet, he sees

crookedness and would, if he touched the oar now, feel straightness; but upon taking the oar out

he would both feel and see straightness (or, more precisely, feel tangible straightness and see

visible straightness).

185 Fogelin, Berkeley and the Principles of Hman Knowledge, 42.

128

Philonous: “In like manner, if he shall conclude from what he perceives in one station, that,

in case he advances towards the moon or tower, he should still be affected with the like ideas, he

is mistaken.”

Each perceiving at a station has its own distinct perceived object at that station; a station is a

place, and at each place a moving man perceives an idea at least numerically different from the

previous one. The perceived object is at the same time as the perceiving but not at the same

place in physical space, since the man is moving toward the moon or tower. But, of course, in

the final metaphysical synthesis, that is, at Grayling’s level 3, the stations and the perceived

objects are at the same metaphorical places in God’s mind and really (actually or potentially) in

any number of finite minds.

Philonous: “But his mistake lies not in what he perceives immediately, and at present (it

being a manifest contradiction to suppose he should err in respect of that) but in the wrong

judgment he makes concerning the ideas he apprehends to be connected with those immediately

perceived: or, concerning the ideas that, from what he perceives at present, he imagines would

be perceived in other circumstances.”

“What he perceives at present”, mentioned twice—this refers both to the time of the

perception and to the time of the what, the content, of the perception: these times are the same.

And this is something he can’t be mistaken about (it would be “a manifest contradiction to

suppose” such an error).

Philonous: “The case is the same with regard to the Copernican system. We do not here

perceive any motion of the earth: but it were erroneous thence to conclude, that, in case we

were placed at as great a distance from that as we are now from the other planets, we should

not then perceive its motion.”

We do not here and now see the vanishing of the sun or the explosion of a star; but because

we know the Laws of Nature which, as Berkeley says at Principles 105, “extend our prospect

beyond what is present, and near to us, and enable us to make very probable conjectures,

touching things that may have happened at very great distances of time and place, as well as to

predict things to come”, we know that if we had been placed sufficiently close to the sun or that

star, we would at once have seen the vanishing or the explosion; but here and now, of course, we

only see the sun’s shining or the star’s twinkling.

Moreover, by the same Laws of Nature, we know that if it were possible (which it physically

isn’t) to travel now, instantaneously, closer to (say, within ten million km of) where the

vanished physical sun used to be, we wouldn’t see any visible sun anymore, not even the much

greater one we’d see if the sun were still there.

So, all in all, it does seem as if Berkeley accepts the simultaneity requirement.

129

4.7. O’Shaughnessy and the Simultaneity-Requirement

In the famous chapter called “Seeing the Light” O’Shaughnessy briefly discusses perception at

cosmic distances in two paragraphs to which I will come shortly. But first let me outline, and

briefly criticize, his general view of seeing light. He holds (1) that we invariably do see light

whenever we see any physical object and also many times when we don’t, (2) that our

perception of light is non-representational, and (3) that, by contrast, our seeing of material

objects is mediated by light. Thus he (4) distinguishes between the “light-representative theory

of light perception,” which he rejects, and the “light-representative theory of object perception,”

which he accepts.

Let me begin with the all-important first point. O’Shaughnessy starts by saying that “surely

we sometimes see and identify samples of light”:

For example, a torchlight shining across a dark and dusty room. This is something that

has a shape and position, that can be viewed from angles and exhibits foreshortening,

that can be individuated and singled out in opposition to particular objects like the

furniture. So is this not a paradigmatic example of seeing the light? In fact it is an a-

typical instance of the seeing of light. Indeed, it is because one naturally selects such

phenomena as paradigmatic that one begins to doubt that the light that makes sight

possible is itself on view. One knows that there are beams emanating from all the things

we see—but where are they? I think it is fair to say that those beams, criss-crossing the

room inches from our very eyes, are for the most part completely invisible. More, we see

nothing and a fortiori no light between our eyes and the objects that we see . . . And there

seems to be no such thing as the individuation within the visual field, and in opposition

to the visible material objects, of the light beams that are coming to our eyes from those

objects. Nor do we see any such beams as shaped and positioned and foreshortened

(etc.).

In a word, if we see light all the time then it must be markedly unlike the seeing of

objects—or of a torchlight across a dusty room. For the truth is that the seeing of this

torchlight is thanks to the seeing of objects rather than the reverse, that it is the seeing of

an object-collective of the nature of a crowd, viz. a cylindrical collection of dust

particles. Indeed, for the most part the beam itself is invisible, and those shining specks

are like so many Man Friday’s footsteps—evidence of the unseen! Then because these

cases of the seeing of light are in so many ways on a par with the seeing of material

130

objects, we may falsely assume that all cases are, and go on looking for something of the

nature of a Light Object.186

But there is no Light Object, says O’Shaughnessy; and if we think there is, it’s only because we

fail to realize that seeing a torchlight is seeing a material object, a collection of dust particles.

Nevertheless, we may ask, is that right?

First, in seeing “a torchlight across a dusty room,” I see specks of dust immersed in, and

enveloped by, a shining cylindrical light-beam that is separated (even if not sharply and

precisely, then anyway sufficiently clearly) from the surrounding darkness. In this case I see

both the particles and the beam. Secondly, suppose the room isn’t all that dusty. In that case I’ll

see even fewer dust particles suspended or floating in the beam—but, again, I will see that

beam. Thirdly, suppose the light in the room is very dim, so dim that it’s almost totally dark, and

it’s only because the room is familiar that I don’t stumble over, or bump into, any furniture.

Then I won’t see any dust particles at all; there’ll be no “shining specks” whatsoever—but I will

still see that beam. All these facts make O’Shaughnessy’s claim that “for the most part the beam

itself is invisible” difficult to accept—unless that “for the most part” is meant to allow for just

those three cases.

If so, it seems indeed to be a fact, as O’Shaughnessy goes on to say, that “all is empty

between eye and object” if by that is meant that we don’t see the rays “criss-crossing the room

inches from our very eyes.” Nor do we see what these rays are composed of, i.e., the photons.

But we do see beams or columns or shafts of light in the dark; while in daylight, the material

objects we see are all strongly or softly bathed in sunlight (and/or artificial light), which we also

invariably, unavoidably, and sometimes enjoyably see. Yet neither the all-pervading artificial

light in a room, nor the all-pervading sunlight (whether bright or dim) during the day, nor the

torchlight or other lights seen at night, are identical with, or like, material objects. They are all

sensuous presences dependent on consciousness: I would not hesitate to call them visual sense-

data, which, according to O’Shaughnessy, “engage the attention before anything else.”187

Indeed, what else can they be, if not sense-data?

Having argued that we do see light sometimes, O’Shaughnessy then argues, convincingly to

my mind, that “[a]s after-images demonstrate the reality of visual sensations, and thereby open

the floodgates—so that every point of every visual field is necessarily visual-sensation-

inhabited, so if we ever see light at any point in the visual field, we must always see light at

every non-black point of every veridical visual field. Now either we never, or sometimes, or

always see light. Then who will say we never see light? Catching sight of torchlight, of car

186 O’Shaughnessy, Consciousness and the World, 440. 187 ibidem, 465.

131

headlights, the sun, moonlight, a distant window in the forest at night, all rule out this

possibility.”188

On a page further down he again asks whether it not be “preposterous to suggest we never

see light? What is it that violet radiation has that ultra-violet radiation lacks—if not visibility?

For why the ‘ultra’? What is the blue ‘what’ that we see when we see the blue sky above, if not

the light raining down upon us? Do we not see light when we see rainbows? stars?

conflagrations?”189

Stars and other stellar bodies are our concern here, and it is regarding them that

O’Shaughnessy adduces the following two-paragraph argument, after having first supposed that

he is in a room with an orange lamp glowing in one corner and an orange-painted sphere

illuminated in the other:

We shall suppose a vast stellar orange body, with the angular dimensions in our sky of

the moon, situated a light year away. Assume this object to be almost always

unilluminated and invisible to us. Then let us suppose that for one second an intense

beam of white light falls upon its surface at 9 p.m. on 1 January 1999 (GMT). The

orange light reflected in that second reaches our eyes at 9 p.m. on 1 January 2000. For a

second we see ‘orange light’, ‘an orange light in the sky’, ‘a heavenly object’, ‘an orange

surface’, ‘the orange colour of a heavenly object`—not all that different in appearance

from the orange sphere glowing in the corner of the room, or the normally illuminated

orange-painted object in the other corner. Where is the heavenly object? A light-year

away. Where is the orange surface? Ditto. Where is the orange colour of that surface?

Ditto. Where is the orange light in the sky? Ditto (for this is the sense in which we say, ‘I

saw a light’ [i.e., something bright] ‘moving slowly down the mountain side’). Where is

the orange light? Where our eyes are. The temptation to postulate a light-representative

theory of light perception—such that we set eyes upon light that is light-years away

through setting eyes on light that is here—comes to the fore in the tension between the

last two questions and answers.

I conclude that when I look at the orange lamp glowing in the corner of the room, and

the properly illuminated sphere in the other corner, then (A) I see the two spheres, as and

at where they are, (B) I see their colour as and at where it is, (C) I see their surfaces, as

and at where they are, and all of this is both identical with and due to the fact (D) I see in

188 ibidem, 441. 189 ibidem, 443.

132

either case orange light that is where I am. This constitutes a light-representative theory

of the visual perception of objects.190

Before commenting on these two paragraphs (which we may call O’Shaughnessy’s TGA), I will

quote some of what he says a little later:

The light that we see, in seeing objects at a remove in space, is not merely located where

we are; to be precise, it is located upon our retinas.191 It is not situated on the surface of

the eye, nor a little in front of the eye, nor a mile or light-year in front of it, for the

radiation at these points has yet to encounter the visual system—and be seen. . . . Then

just as the sound that we hear is where our ears are situated, so too is the light that we see

upon our retinas. Perception at a distance, which is something that light effects for

material objects, fails to obtain in the case of each of these mediators. The light and

sound that we perceive are invariably close at hand. More exactly, they tend to be sited

on or just within our perimeters.

But what kind of experience could it possibly be to set eyes upon light that is situated

at the back of one’s eyes? The answer to this question is simple: it is no different from

any other visual experience. And the reason for this is equally simple: it is one and the

same as those familiar experiences. However, that they are the same experience is

concealed by the interpretational content of most visual experience. For just as we see

material objects at a distance in and through seeing light that is on our retina, so in

reverse fashion we see light that is situated on our retinas in the mode of seeing material

objects at a distance.192

But what is it to see light situated on our retinas in this mode? O’Shaughnessy proposes

answering this “by answering a different question: what would it be like to see light on the retina

not in the mode of seeming to see objects at a distance? What would it be like to see it in the

mode of seeing light?” He says:

190 ibidem, 448. 191 This is what Berkeley, too, claims—at least if Gary Thrane is right. In his 1977 paper “Berkeley’s “Proper Object of Vision” Thrane states on p. 243 that “there are three relatively well-known positions on this issue. In the nineteenth century it was usually held that (1) Berkeley's proper object of vision was the pattern of light on the retina and (2) that this was a reasonable thesis. A second, more recent position, is that Berkeley indeed thought (1) that we see the pattern of light on the retina but (2) that this is absurd (D. M. Armstrong). And, finally, it has been held (Colin Turbayne) that, (1) since it is absurd to hold that we see the pattern on the retina, (2) Berkeley could not have meant that the proper object of vision is the retinal pattern. I shall urge that Berkeley does indeed hold that the proper object of vision is the pattern of light on the retina. In part, my case rests on showing that this is not an absurd thesis. Further, I will show how Berkeley's characterization of the proper object of vision follows from the thesis that the proper object is the retinal pattern.” Much of what O’Shaughnessy says characterizes seeing light on our retinas coincides with what Thrane says is Berkeley’s account of retinal images as the proper objects of human vision. 192 O’Shaughnessy, Consciousness and the World, 448-449.

133

It would be an experience not all that dissimilar from the one we encounter when we

close our eyes in daylight. We would experience a continuous visual field, populated by

coloured patches, given in particular directions out from the head, containing visible

spatial structures of the kind of circles and lines. No more. What would be absent would

be experiences of depth, of spatial structures like hemisphericity, indeed of material

objects generally, and the near flawless typing of contents (‘car’, ‘tree’, ‘chair’ etc.).

Then if such visual experience were to be generated veridically by light, and in a manner

sufficiently resembling our present experience of sound, we would I suggest be seeing

the light that is located on our retinas—without seeing its material object sources. And

we would presumably see that light as light. . . .

And yet, even though in this imagined situation we experience the light on the retina

as light, we do not experience it as on the retina. . . . For it must be emphasized that, just

as we do not hear the sound which is situated at our ears as or to be at our ears, since our

experience of sound is purely directional experience, so likewise in the case of the

experiencing of light. We do not see the light that is on our retinas as on our retinas, let

alone see it to be there. To believe so would be to misconstrue the present theory as

involving the rather wild supposition that we see our own retinas in seeing material

objects. Seeing the light on our own retinas is not to be likened to seeing a patch of light

upon the carpet. While the latter necessitates the differentiation of a surface within a

space of three dimensions, the former does not. This is because it is purely directional in

character. Light is given in directions, but not in depths, out from the body. The third

dimension of light is simply invisible.193

Now let us return to O’Shaughnessy’s TGA, those two paragraphs quoted earlier. According to

O’Shaughnessy’s supposition, we see all the things he enumerates for just a second at 9 p.m. on

1 January 2000. Notice how O’Shaughnessy just mentions one time—the time—at which all

these things are seen by us and also explicitly states where they are when we see them, namely,

a light-year away; but he doesn’t explicitly state whether he thinks we see them as they are at 9

p.m. on 1 January 2000 or as they were at 9 p.m. on 1 January 1999. Of course, at 9 p.m. on 1

January 2000 the moon-sized stellar body still exists, even though it is unilluminated. Its orange

surface, too, still exists and it is still orange even though there’s no light falling on it. It is

reasonable to say that we see the body and its surface and the latter’s orange colour at 9 p.m. on

1 January 2000 as they were at 9 p.m. on 1 January 1999 (even if during the intervening year

they may have changed very little, relative to what human vision is able to discern).

193 ibidem, 449.

134

But what about the orange light in the sky? At 9 p.m. on 1 January 1999 there was no light in

the sky—because if, as O’Shaughnessy says, “this is the sense in which we say “I saw a light . .

. moving slowly down the mountain side’,” then ex hypothesi on 1 January 1999 there just

wasn’t any light in the sky that we or any other human beings could have seen. Here, I think,

O’Shaughnessy correctly assumes that there is one central, Berkeleian, sense of “light” such that

light is necessarily something that is seen. And, of course, the orange light that, according to

O’Shaughnessy, is where our eyes are (i.e., our eyes on 1 January 2000), was there on 1 January

2000, and not a year earlier. And it was there for just a second in that year—the same amount of

time the orange body was illuminated in the previous year. Thus my conclusion from this case is

that O’Shaughnessy does accept the simultaneity-principle.

That he accepts it may also seem evident from what he says in the Introduction of his book.

Yet on reflection this is not quite so evident. The passage I have in mind is the following:

I suggest there is no mystery as to what consciousness is. Unlike much in the mind, there

is reason for thinking that consciousness is not an indefinable, whether of the type of

simple qualia, or such as we encounter in some of the fundamental phenomena of the

mind like experience, belief, desire. Consciousness has a determinate character of

internal type, and there exist logically necessary and sufficient psychological conditions

for the presence of this phenomenon. . . .

Now a cornerstone in the analysis of consciousness is the fact that its presence entails

contemporaneous experience (a ‘stream of consciousness’). This is a transparently clear

analytic necessity.194

Thus, for O’Shaughnessy, consciousness at t clearly entails experience at t of something.

Abstractly, there are four possibilities of what that something might be: (1) a material object

existing before but not at t; (2) a material object existing at t; (3) a mental object existing before

but not at t; or (4) a mental object existing at t.

Now since mental objects are dependent on consciousness of them, then if I am awake with

my eyes open, possibility (4) is in this case certainly actual: I certainly am conscious of (what is)

a mental object (a sense-datum) at t. In virtue of that, possibility (2) may also be actual, and it

usually is: whenever I’m awake, for example, I see (some of) my physical surroundings at t.

But what about the abstract possibilities (1) and (3)? Are they logically excluded? It might

seem that they are: for after all, if the presence of consciousness entails contemporaneous

experience, how can I be conscious at t of a mental or material object existing before t? But, for

all that O’Shaughnessy has said in this passage, I can: what is excluded is non-contemporaneous

194 ibidem, 4-5.

135

experience; but in cases (1) and (3) the experience at t is contemporaneous with consciousness at

t: only the objects of the experience (and the consciousness) are not; they exist before t.

Nevertheless, that O’Shaughnessy means to exclude possibilities (1) and (3) is sufficiently

evident from his own treatment of the TGA discussed earlier. And that there is overwhelming

reason to accept the simultaneity-principle will be additionally argued for in the next sections,

particularly 4.14.

4.8. Carrier et al. on Seeing Light and the Simultaneity-Requirement (I)

One contemporary philosopher who, unlike O’Shaughnessy, wants to avoid sense-data but, like

O’Shaughnessy, embraces the simultaneity requirement (and does so more emphatically and

explicitly than O’Shaughnessy), is Len Carrier (whom we have previously described as a Non-

Idealist Direct Realist). He presents the TGA as consisting of three premises and a conclusion

as follows:

P1. In order for a physical object to be seen, it is causally necessary that the object has

emitted or reflected light.

P2. Light travels at a finite speed.

P3. In order for a physical object to be seen at a certain time T, it is necessary for this

object to exist at time T.

Conclusion. Whenever one sees anything he sees something that is non-physical.195

(Note that P3 is just a corollary of the simultaneity-requirement.) Carrier thinks the premises are

true but the conclusion doesn’t follow. Of course it doesn’t—anyone can see that. But it’s also

evident why that’s so: some premises are missing. One of them would be the premise (Carrier,

in his exposition, doesn’t give it a label) that by the time light emitted by a certain star reaches

an observer’s eyes, it has ceased to exist. “But if the star does not exist at time T,” Carrier goes

on to remark correctly, “then, according to P3, it is not the case that the star is seen at time T.

We cannot deny, however, that the observer does see something at T.”196 The latter, of course, is

another premise required for the conclusion indicated.

Nevertheless, the only valid conclusion to be drawn from the argument so far is that what the

observer “sees must be something different from the star.” It “has not yet” been “established that

what is seen at time T is non-physical. But what else can there be, in this case, that is physical?

There seem to be only two viable alternatives:

(a) some physical property of the star, or

195 Carrier, “The Time-Gap Argument”, 263. 196 ibidem, 263-264.

136

(b) the light that the star once emitted.”197

Carrier opts for (b), and gives some good arguments; but first he disposes of (a), saying that it

runs afoul of the following argument. Every property requires a bearer. Since that which

does not exist can possess no properties at all, it cannot possess any physical properties

that remain to be seen. A Platonist might reject this argument by denying that properties

require bearers; but this move affords the Direct Realist no solace. For even if physical

properties do exist apart from their exemplification, in this separate state they are only

subject to contemplation. If they are also to be seen, they must attach themselves to some

physical item. But this item is no longer available in the story about the exploded star.198

Carrier then mentions two objections to (b), his own preferred candidate for the physical object

seen in the extinct star case. They are as follows:

(i) Light is causally necessary for seeing but is not itself seen. In this respect it is

like the optic nerve, which is also causally necessary for seeing.

(ii) If light were seen, then one would have to see light waves or particles. But since

these are physically impossible to see, light cannot be seen.199

Later he shows that these objections are unfounded. But if you accept them, then neither (b) the

light that the star once emitted nor (a) some physical property of the star (discredited, as we’ve

just seen, by Carrier himself) is available for you an item that is seen in the disintegrated star

case. Thus, “(s)ince (a) and (b) exhaust the possibilities of seeing something physical, what the

observer sees in the case of the disintegrated star must be something non-physical.”200

4.9. The Continuity Argument

Once we are clear about what it is that we see in the no-longer-existent star case, we can give a

plausible Continuity Argument to cover all cases of seeing. Carrier formulates it thus:

(i) There is no qualitative visual difference between the case in which a star exists

and the case in which it does not. Since all the criteria for identification coincide,

if something non-physical is seen in the latter case, then something non-physical

must be seen in the former case, too.

(ii) If we always see something non-physical in the case of distant objects, such as

stars, then there is no non-arbitrary place to set the distance at which we fail to

see something non-physical. No one is tempted to say that we must give a

197 ibidem, 264. 198 ibidem. 199 ibidem. 200 ibidem.

137

different analysis of vision when we extend distances from five to fifty yards, or

even to a mile or to five miles. Nor is it conceivable that we must explain seeing

the moon differently from seeing the sun, or from seeing the stars. There is a

continuity between near and far objects, so that there is no justification for saying

that there is some near point at which we cease to see what is non-physical.

Hence, whenever anything is seen, something non-physical is seen.201

I don’t have anything at all to object to in this argument, except that I would substitute ‘mental’

for ‘non-physical’. And I would do this because I would like the TGA to prove the existence of

sense-data; and if Carrier’s version requires the conclusion “that in all cases of seeing we see

light, either emitted or reflected” (and it does, as we shall see in 4.12.), then I wish that light to

be a sense-datum, or for there to be sense-data of light, with the understanding that sense-data

are essentially mental items (that might turn out to be physical, too).

Michael Huemer, thirty-odd years later, constructs a continuity argument similar to Carrier’s.

He first supposes that 1000 light-years away from us there was a star destroyed 300 years ago

and then goes on to say:

For suppose that there is a second star, also 1000 light-years away, next to the first one,

but that this second star has not been destroyed and continues to exist now. Could we,

with any plausibility, claim that what we are seeing in these two cases is a radically

different kind of thing? If the first case is one of seeing a sense-datum of the star, isn’t

the second case also one of seeing a sense datum? After all, one’s experience has exactly

the same kind of etiology and the same introspectible character in the two cases. The

only difference is that in the one case, the star was destroyed 700 years after it emitted

the light that’s now causing your experience—but that has no effect whatever on you or

your experience. So if a sense datum exists in one case, why wouldn’t it exist in the

other?

If you buy this argument, you will also have to accept sense data for all the objects

around you. For in the case of any perception, there is always some time gap. When you

look at your own hand, there is a (very small) delay between the time it bounces off its

surface and when you have an experience of seeming to see your hand. So if the time

gap argument shows you are not really seeing a star, it also shows you are not really

seeing your hand.202

But then Huemer introduces and rejects the solution offered by Carrier. He does this as follows:

201 ibidem, 264-265. 202 Huemer, Skepticism and the Veil of Perception, 131-132.

138

Perhaps what you are really seeing is simply the light emitted by the star, rather than

either the star itself or a sense-datum. The light from the star continues to exist at the

time you have the visual experience, so there’s no problem, right?

Well, almost. There is still a delay (very short) between the time light rays strike

your retina and the time you have a visual experience, required for the signals to reach

your visual cortex and your brain to process the information. In fact, the particular

photons causing your visual experience do not exist at the time your experience occurs,

having been absorbed by your retina. Therefore, if the time gap argument shows that

you’re not really seeing the star, it must equally prove that you are not really seeing the

light either. Nor, for the same reason, could you be seeing retinal images. I assume no

one will be tempted to say you’re seeing your brain. So the only thing left for you to be

seeing appears to be a sense datum.203

This conclusion that Huemer seems to be left with is, from my point of view, ideal: the TGA

indeed shows that what we are seeing are sense-data; and that’s the conclusion I want. But

Huemer doesn’t want to accept this conclusion; therefore, his only option is to reject the

simultaneity-principle, and that’s what he does. He says that

(a)fter this elaborate argument, my response may seem disappointingly simple. What are

you aware of when looking at the star 1000 light-years away? You are aware of the star,

as it was 1000 years ago. I see no reason why one should not be able to perceive

something in the past. Obviously, the time at which your perceiving occurs cannot be

before your experience occurs. But why must the time of the perceived state of affairs be

identical with the time of the perceiving?204

If this isn’t clear yet, it will be by the end of this chapter, when we will have presented many

reasons why one cannot perceive something that exists in the past rather than present. Of course,

I am also claiming that when, in the no-longer-existent-star case, we are seeing sense-data, it is

the sense-data of light that we are seeing.

At any rate, to finish with the Continuity Argument, here’s C. D. Broad’s version from 1959:

The case where the body seen is near to the percipient's body cannot fairly be considered

in isolation from cases where it is very remote. We see bodies which are at all sorts of

distances, from close at hand to many millions of miles away. The external causal

conditions of the visual perceptions are, so far as we know, precisely similar in kind in

all these cases. Is it really credible that there is a certain range of distance, on one side of

which the immediate objects of visual sensations are parts of the surfaces of bodies

203 ibidem, 132 204 ibidem.

139

emitting light to the percipient's eye, and on the other side of which they are of a wholly

different nature?

. . . I think that all the arguments for the non-corporeality of visual sensibilia rest on

considerations of continuity. In view of the continuity in the external conditions of our

visual sensations, I find it very hard to believe that some of the visual sensibilia which

we sense are parts of the surfaces of the bodies which we see, and that others are not

parts of the surface of any body, if to be a “body” and to be “a part of the surface of a

body” be understood in [a] simple literal way . . . Now I also find it very hard to believe

that all the visual sensibilia which we sense are parts of the surfaces of the bodies which

we see, if "body" and "part of the surface of a body" are understood in that way.

Therefore I am strongly inclined to think that none of them are.205

And I might also remark here that Broad never explicitly discusses or questions the

simultaneity-requirement. He seems to regard it as self-evidently obvious.

4.10. Carrier et al. on Seeing Light and the Simultaneity-Requirement (II)

Let us return to the TGA as presented by Carrier. He wishes to preserve the simultaneity-

principle, which (we will remember) he formulates thus:206

P3. In order for a physical object to be seen at a certain time T, it is

necessary for this object to exist at time T.

But some philosophers who reject it propose to replace it with the following:

P4. In order for a physical object to be seen at time T, it is necessary and

sufficient that some state of this object be seen at T.

Carrier claims, however, that P.4 will not by itself “get around the hypothesis of the exploded

star.” He points out that

states, like properties, must be borne by individuals; and if the individual in question

does not exist, no states of that individual can exist, either. There can be no state that

survives the destruction of a thing of which it is a state and lingers about to be seen.

He then goes on to claim:

At this juncture, supporters of P4 would insist that what is seen in the case of the

disintegrated star is some past state of it. It must now be determined whether good

205 Broad, “Reply to my Critics”, 807-808. 206 Carrier, “The Time-Gap Argument”, 265.

140

sense can be made of this move and, if so, whether the price of making it might prove

too great.207

In order to make good sense, this move, says Carrier, requires us “to interpret the existential

quantifier itself, as well as all predicates, in a tenseless fashion, inserting temporal indicators

whenever necessary” so that “the star itself, regardless of its disruption, is always on hand to

support its physical states and properties. . . . Thus, even if the exploded star no longer exists, in

the tensed sense of ‘exists’, P4 requires our saying that the star really does exist after all,

speaking tenselessly.”208

But, says Carrier, requiring a tenseless way of speaking is just too high a price to pay for a

“solution to the problem of the time-gap.” Why? He thinks there are just three possibilities as to

the best way of describing reality. They are:

(i) A completely tenseless way of speaking best describes the world.

(ii) A completely tensed way of speaking best describes the world.

(iii) Both the tensed and tenseless ways of speaking are equally accurate, covering the

same territory like two maps with different projections.

He himself proposes “to remain neutral” on “which, if any, of these answers is correct”; but

what he finds

strange is that in order to make good sense of the P4 solution to the problem of the

time-gap, answer (i) must be chosen. Now it might be the case that (i) does give the

right answer but, if so, this should require independent argument, and not be forced

upon one by his treatment of the time-gap in perception. I therefore consider it a defect

in the P4 solution that it makes demands on the way we are to speak about the world

and consequently makes demands upon our choice of ontology. Those, for example,

who believe in temporal becoming, and believe that objects come into existence,

change, and then pass away, will be disappointed to find their ontology incompatible

with the P4 solution.209

Carrier then says that his preferred solution (namely, that in the case of the disintegrated star—

and, because of the Continuity Argument, in all other cases of seeing as well—what we see is

light) “depends upon no such ontological decision and will be congenial to any of these three

ways of speaking. It seems obvious that a solution that is ontologically neutral here has just that

much more chance of being correct.”210

207 ibidem. 208 ibidem, 266-267. 209 ibidem, 267. 210 ibidem.

141

Towards the end of 4.8 we mentioned two objections to the claim that we can and do see

light. In answer to the first, Carrier says that

light is external to our bodies, so that it is not analogous to the optic nerve. The former

is a causally necessary factor which is seen, whereas the latter is a causally necessary

factor which is not seen. Ordinary language supports this point. We speak naturally of

seeing ‘a beam of light’ or ‘a point of light’, or seeing ‘nothing but bright lights’ or

‘seeing the first light of day’. We have no such constructions that involve the optic

nerve, and justly so, since we can only see what is before our eyes.211

Now it might seem that Carrier is contradicting here what O’Shaughnessy claims about the

location of the light that we see: for the former, it is always external to our bodies; for the latter,

the light that we see is always on the retina. But the contradiction, though it exists, is superficial:

O’Shaughnessy claims that the light falling on our retinas is always interpreted as being external

to our bodies; whereas for Carrier the only light we directly see is the light external to our

bodies, before our eyes: the light on our retinas is light we do not see. In this respect it is like the

optic nerve, a causally necessary factor for seeing but not itself seen.

Carrier makes this even clearer in his response to Charles B. Daniels. There he says that

Daniels objects to my view that what we see in the case of an extinct star is only its light.

He says that when we see light we can usually point to the place where we see it to be.

But where is the light from the defunct star? The obvious answer is that this light is

everywhere that it manages to reach, and that includes the path on which it travels to

reach one’s eyes. Daniels then imagines a situation in which the only light ‘that lies on

the line between my retina and the place the star last occupied all happens to be located

inside my eyeball’ . . . This situation creates no puzzle for my view, since I was explicit

in claiming that we only see what is before our eyes and not what is in them . . . If all the

light from the extinct star is inside my eyeball, then this is light that I no longer see.212

Here the opposition to O’Shaughnessy is stark: for Carrier, no light that I see is inside my

eyeball; for O’Shaughnessy, all the light I see is inside my eyeball. And yet, though stark, it also

is superficial in that, unlike nearly all other recent or current analytic philosophers, both

emphatically proclaim the following central thesis (I state it in Carrier’s words):

Thus I must maintain that in all cases of seeing we see light, either emitted or reflected;

whereas just in some cases we do see physical objects: only in those cases in which these

objects still exist at the time they are seen.213

211 ibidem, 269 212 Carrier, “Time-Gap Myopia”, 55. 213 Carrier, “The Time-Gap Argument”, 270.

142

The question that divides them (“Where is the light I see?”) must also be faced by other

philosophers seriously considering this solution to the time-gap problem. One of them is

Howard Robinson. He says that if in answer to this question (in the case of looking at the sun)

one chose the light that is striking the eye, then one would have the problem of why one

did not seem to see two suns, for there are two spatially separate packages of light

striking the two eyes: and there might, too, be pressure to treat all visual perception in

the same way, and this would be no better than the sense-datum theory.214

When I asked Robinson to explain what he had in mind by this last sentential clause, he

replied: “What I think I meant was that, if one agreed that, in general, one perceived light

striking the eye, rather than the object from which the light was coming, one would be aware of

an intermediary entity—the light at the eye —not the external object. And this looks rather like a

form of Representative Realism, with the light constituting a kind of physical sense-datum.”215

It is in anticipation of such a response and to avoid Representative Realism that Carrier

insists that in all cases of seeing we see light but only in some do we also see physical objects.

He therefore explicitly denies that “[w]e see either the light or the object, but not both”216 and

thereby evades Robinson’s objection that he is treating the light at the eye as “a kind of physical

sense-datum” of the external object.

But “at my eye” is not the only possible answer to the question, Where is the light I see? To

continue the quote from Robinson, if

one chose light at some distance from the eye there would be the problem of deciding

how far away the preferred light was deemed to be. . . . [One answer could be] that in the

case of stars and sun we perceive the light that is that distance away from the eye which

is the maximum distance at which the light reaches the eye in a shorter time than the

minimum time discrimination we are able to make. This may vary according to the

nature and state of the subject. The ‘sun’ and the ‘stars’ that we see will, then, be purely

visual phenomena, rather like the blue of the sky or a rainbow.217

A bit later Robinson raises the question of “whether the naïve realist can cope with the[se]

purely visual phenomena . . . or whether they constitute a foothold for the sense-datum

theorist.”218 He leaves this question unanswered.

214 Robinson, Perception, 82. 215 Personal communication 216 Carrier, “The Time-Gap Argument”, 270. 217 Robinson, Perception, 83. 218 ibidem, 84.

143

4.11. Seeing distant objects vs. seeing objects close-by: Mandelbaum and Noë

Earlier in the history of contemporary philosophy this question was discussed by, among others,

Maurice Mandelbaum. I will quote his remarks in full because they are both interesting and

important in more than one respect:

When we say that we see the sun or a star we presumably do not believe that the sun

which we see is actually to be described as a dazzlingly bright disk, nor that the star

itself is a tiny glittering point of bluish light: in our everyday world we accept sun and

stars as being immensely distant objects which possess characteristics quite different

from those we would attribute to them if we were merely describing what is visible to us

when we look at the heavens. Nevertheless, when we say that we see a particular star we

mean that what we see is a star. These two beliefs—that we do really see a star, and yet

that the star does not possess the properties which we see it as having—are not

inconsistent. What saves then from inconsistency is the fact that in such cases we readily

interpret the relationship between the object which we see and the qualities which that

object appears to us to possess as a causal relationship. Such a causal relationship need

not be clearly conceived, and on the level of common sense it presumably is not: it

merely involves a belief that if there were no object of the sort that we refer to as “a star”

we should not be presented with the particular qualities by means of which we describe

how the star looks to us.219

In a footnote to this passage Mandelbaum approves of C. D.Broad’s claim (in the latter’s

Perception, Physics, and Reality) that, in Mandelbaum’s words, “the difficulties of naïve

realism tend to push common sense toward the acceptance of a causal theory of perception.” In

that footnote Mandelbaum goes on to say:

It is also true that an acceptance of scientific astronomy would lead us to distinguish, as I

have distinguished, between the star’s nature and the appearance which it presents to us.

However, in addition, I should like to point out that there may well be phenomenological

grounds on which—without argument—we distinguish between what we call “things”

and which we regard as “mere appearances,” and that stars (as we see them) may possess

those phenomenological characteristics which lead us to classify them as belonging to

the latter rather than to the former group. Among such phenomenological characteristics

219 Mandelbaum, Philosophy, Science, and Sense Perception, 177-178.

144

I might mention indefiniteness of contour, lack of fixity (e.g., flickering), absence of

perceived microstructure, and color which is not surface-color. . . .220

Mandelbaum, in the main text, continues:

To separate the look of a star—its appearance to us—and the star itself, is not to invent

entities for philosophic purposes. To speak of the faint and shimmering light of a star

when we look at the heavens is a perfectly natural mode of discourse, and one which we

take to be descriptive of our experience. The point of light which we see is a token of the

star’s presence, as a faint and flashing light may be the token of an airplane’s presence.

In neither case need we think—nor need we speak—as if what were there for us to see

was to be identified with the object itself.221

Yet what Mandlebaum is saying here is so tricky that we should be extremely careful. In the

second sentence he tells us that in speaking “of the faint and shimmering light of a star” we are

describing “our experience.” But is it our experience we are describing—or are we describing

what we experience, the object of our experience, and saying of it that it is a “faint and

shimmering light”? I opt for the second alternative, and think that Mandelbaum, upon reflection,

would not object.

In the last sentence Mandelbaum tells us what it is that we need not think or speak. But does

this mean that we cannot or even should not identify the point of light with the star? The faint

and flashing light may indeed be the token of an airplane’s presence: after all, it is just the

airplane’s light working rather than the airplane itself. But isn’t the flashing of the airplane’s

light just the airplane’s presence announcing itself? And the shimmering light in the sky—isn’t

that the star itself? Mandelbaum wants there to be two objects and two processes, but isn’t there

just one?

No, there are two: Mandelbaum is right. But the two are so closely connected that they are

easily (but mistakenly) identified.

Immediately after the above quotation (asserting that there is no need to think that “what

were there for us to see was to be identified with the object itself”), Mandelbaum makes the

following noteworthy point:

However, there must be acknowledged to be a very great difference between what we

take for granted about seeing a star and what we take for granted when, for example, we

see a book. In the latter case we would assuredly reject the view that what we are

actually seeing is caused by the object at which we are looking; rather, what we see

strikes us as being the object. And this is true even in those cases in which we feel there

220 ibidem, 178. 221 ibidem.

145

is something deceptive about the appearance of what we see. For example, if we regard

the illumination as responsible for making a book appear a different color from the color

which we would say that it actually is, we would nonetheless hold that what we have

immediately before us is the book itself, and not some token of it.222

One thing that surprises about this passage is the thought that there is an opposition between

what a perceived object causes and what it is; i.e., what’s surprising is the thought that the view

that what we are actually seeing is caused by the object at which we are looking is somehow

incompatible with the perceived object’s being that object. In other words, what “we would

assuredly reject” is the view that (as Dummett expressed it) there is nothing “more to our

perceiving objects than their causing us to have certain perceptual experiences.”223

Then Mandelbaum continues:

On the other hand, when we look at a star we are willing to accept the view that what we

see is in some sense an image of the star, for if we were to describe (and not merely

name) what we are at present seeing, we would not regard that description as an accurate

description of the star itself: what we see (we might say) is the light of the star. It is

because of this difference between the two cases that direct realists take the case of

seeing a book as paradigmatic for an analysis of perception, while one who holds a

representative theory of perception would wish to devote careful attention to cases such

as those of the star, in which distance substantially alters the appearance of objects.224

This last sentence is also (pleasantly) surprising. And now for my crucial point: isn’t a

shimmering, flickering light always the way a star appears to us? Don’t we (human beings on

Earth) always see a star as just being a “point of light”? Isn’t that the only appearance of a star

available to us in our terrestrial condition? The answer to these rhetorical questions is, I believe,

Yes. And this is so whether or not the star still exists at the time the flickering light is seen.

Alva Noë seems partially to agree with this but puts a unique twist to it and lands himself in a

possible contradiction:

When you look up in the night sky, you don’t actually visually experience the stars;

what you see, rather, are points of light in the night sky, points of light you reasonably

take to be stars (or to be marks or signs or traces of stars). The stars themselves do not

enter into your experience. The direct theory of perception fails for the seeing of stars in

the night sky. We can see why this is so by considering that, for one’s experience of the

night sky to be veridical (as we would like to say), there’s no requirement that stars be

222 ibidem, 178-179. 223 Dummett, “Common Sense and Physics”, in Macdonald, G. F. (ed.), Perception and Identity, Essays Presented to A. J. Ayer, with His Replies, 36. 224 Mandelbaum, Philosophy, Science, and Sense Perception, 179.

146

points of light, or that they really look like points of like, or that they be located where

we seem to see them. That’s why the fact that stars may have gone extinct millions of

years ago does not put the lie to our current experience of the night sky.225

The contradiction emerges from the conjunction of the statements that “what you see . . . are

points of light,” that these points of light may “reasonably take[n] to be stars,” that “you don’t

actually visually experience the stars” and that “[t]he stars themselves do not enter into your

experience” while the points of light do.

To get rid of the contradiction, Noë would have to say one of two things. Either (a) the points

of light are the stars, in which case they (the stars) are ipso facto seen by you and therefore they

do enter into your experience. Or (b) they (the points of light) are not the stars but only “marks

or signs or traces of stars,” which marks or signs or traces you do see and which, again, do enter

into your experience. Which alternative does he choose? For the time being this remains a

mystery.

Noë next compares looking at stars with looking “down from the height of a very tall

skyscraper at cars and people below”:

The people look to be the size of ants! That is to say, you can’t really experience the

people from the top of the skyscraper. After all, there’s nothing in the least ant-like

about people (in respect of size). That they look the size of ants is intelligible, of course.

This means, roughly, that what you see takes up about the same amount of visual field as

an ant would when looked at from a normal upright position. What can’t be denied is

that this is an incorrect experience of the people. For what would make such an

experience veridical? The actual presence of ant-sized people!226

To be sure, much of this makes sense, but saying: You can’t really experience the people from

the top of the skyscraper – come on! What about the terrorist sharpshooter ensconced at the top

of a high-rise and picking down the “ants” one by one? Of course, his “experience” is

undesirable and horrid; nevertheless he has an experience of people that, from that height, is

visually correct. If, looking down from the top of a very high skyscraper, he “saw” people who

looked, not the size of ants but the size of elephants, now that would be an incorrect

(nonveridical) experience!

And the only reason we wouldn’t call an experience of people looking the size of ants

veridical is that we tie veridicality to normal conditions of observation, which looking at people

from atop a skyscraper doesn’t provide.

Later in his paper Noë makes a very good point:

225 Noe, “Real Presence,” 49. 226 ibidem, 49-50.

147

I have proposed that we think of seeings as like touchings. On this analogy, seeing how

things look is a way of coming into contact with them, it is a way of grasping them. This

is why the things themselves enter into visual experience. We see – that is to say, we

come into contact with – them.227

But there are ways in which we may fail “to establish and maintain contact with things in

perception,” says Noë. One of them is allegedly as follows:

When we look up at the night sky, I argue, we just don’t succeed in making contact with

the stars although we do succeed in making contact with the lights in the night sky. The

stars are just too far away! Ditto for the people down below on the ground. . . .

It is relevant that most cases of nonveridical perception are cases of merely partial

nonveridicality. You misperceive the spoon as bent in water, for it isn’t bent. But in thus

misperceiving the spoon, you do succeed in seeing the spoon. In exactly the same way,

you may fail to visually experience the stars, but not because you are hallucinating. You

do experience the lights in the night sky. You are in contact with them. For example,

when you move your eyes away, they go out of view. You modulate your relation to

them in this and many other ways. For the vast majority of cases of nonveridical

perception, the world is at hand, and is present, thus, as a partner in the experience, as

content for the experience.228

But aren’t the lights in the sky as far away as the stars? And if you “do experience the lights in

the night sky” and you “are in contact with them” and you “modulate your relation to them” in

various ways, then why does Noë think you “may fail to visually experience the stars”? Isn’t it

the case that anything you do with respect to the lights in the night sky you can—and do—

accomplish with respect to the stars as well? And the people down below on the ground are so

much closer to you than the lights in the night sky that you can even consciously and

deliberately target them.

Yet in another way Noë is absolutely correct. The people down people you see from atop a

skyscraper you don’t see in any of their individuality—don’t see their faces, clothes, build,

walk, or any other distinguishing features. So the sharpshooter picking down the “ants” below

one by one is not “consciously and deliberately” targeting them as specific people. If he were

instructed to assassinate a specific person and just given a picture of him and told that this

person would be in the crowd at that location and time, it would be a sheer accident if he hit the

person in question. So for us looking at people from the top of a skyscraper is an incorrect way

of experiencing people— Noë is right. But we, earthlings, have few if any other ways of

227 ibidem, 51. 228 ibidem, 52.

148

experiencing stars than by looking up at the night sky and observing the points of light—so here

Noë overstates his case. Besides, our expectations for experiencing stars are different and much

lower than our expectations for experiencing people.

4.12. Carrier et al. on Seeing Light and the Simultaneity-Requirement (III)

There is one more objection to Carrier’s thesis that we see light that we’ve mentioned (in 4.10.)

but not yet answered. Carrier answers it thus:

[E]ven though light may be composed of, say, photons, it is not the case that one must

see photons if he sees light. For a wall is composed of molecules; but we may see a wall

without thereby being forced to see the molecules that compose it, since the latter are

invisible to the naked eye. According to the logical principle of identity, we need only

assume that light is identical with that which is composed of photons, just as the wall is

identical with that which is composed of molecules. In seeing light, then, we do see that

which is composed of photons. But seeing the whole in either case does not require one

to see the parts, since the parts are, ex hypothesi, invisible.229

The logical principle is unexceptionable, but this just forces us to recognize that “light” is

ambiguous: it can either mean

(1) that which is composed of photons,

or it can mean, as Berkeley’s Philonous has it,

(2) light and colors “in the vulgar sense, or taken for the immediate objects

of sight”230; i.e., things secondarily caused by, but not composed of, photons.

Carrier opts for (1); but I will argue that (2) is correct and that Carrier should opt for it. He

then continues as follows:

The consideration above also answers the criticism that we only speak loosely about

seeing a beam of light, whereas what we ‘really’ see are the masses of odds and ends—

bits of dust, etc.—that are illuminated. But we cannot see these odds and ends because,

at such a distance, they are too small to be seen. Perhaps light does bounce off these

particles of dust, but it cannot be these particles we see. If we could see then, then we

might just as well say that whenever we look at a table we can see the microscopic dust

on its surface. But we do not say this. Though what we see—whether it is a beam of light

229 Carrier, “The Time-Gap Argument”, 269-270. 230 Berkeley, Works, 3 D, 187.

149

or a table—may owe part of its visibility to something else, we cannot, logically, see

invisible particles.231

This passage brings out another conflict between Carrier and O’Shaughnessy. Both agree that

we see a torchlight, say; but O’Shaughnessy claims we see the torchlight in virtue of seeing the

specks of dust from which the light bounces off; whereas Carrier claims that in most cases we

don’t see the dust at all, just the light. On this issue I’ve already voiced my misgivings regarding

O’Shaughnessy in 4.7, and they coincide in large measure with Carrier’s points here.

The light that we see, however, is not that which is composed of photons, but that which is

caused by them: a sensible expanse (sense-datum) of light. I support this claim by appeal to the

arguments that Foster presents in The Case for Idealism, Chapters 4, 5, and especially 6. More

on this in my own Chapter Six.

4.13. Presentism vs. Eternalism

Carrier regards it as a virtue of his solution to the time-gap problem—and, by extension, of all

treatments of the TGA that endorse the simultaneity-principle—that it is neutral with respect to

whether a tenseless or a tensed way of speaking describes the world better.

Another way of formulating the latter issue is in terms of the following question. Which is

true: Presentism or Eternalism? As Ned Markosian puts it in the Stanford Encyclopedia of

Philosophy, Presentism “is the view that, necessarily, it is always true that only present objects

exist” (where “present” means “temporally present, as opposed to spatially present”). By

contrast, Eternalism is the view “that objects from both the past and the future exist just as much

as present objects.” Markosian himself, though recognizing that Presentism faces “major

problems,” holds it to be the “common sense view” and defends it.

Now Carrier claims, in effect, that, one who endorses the simultaneity-principle can stay

neutral on the issue of whether Presentism or Eternalism is correct, whereas someone denying

that principle and claiming that what we see in the no-longer-existent star case is some past state

of it, is committed to Eternalism.

This issue has been discussed by three other philosophers in their papers on the TGA. That of

Ronald W. Houts is especially intriguing. Concentrating on George Pitcher’s way of dealing

with the TGA, Houts shows that it has some wildly implausible implications.

One of them is that we are not now at any spatial distance from what we see. Let us start with

“the situation in which we are allegedly presently seeing an extinct star. . . . [W]hat we presently

231 Carrier, “The Time-Gap Argument”, 270.

150

see”, according to Pitcher as Houts explains his view, “is a stage of the star which obtained

many years before we were born. But that stage of the star (as well as the star itself) does not

presently exist. Both existed many years ago, and are presently nonexistent.” However, as Houts

points out, “there is no spatial distance between two physical things which do not coexist [exist

at the same time]: a necessary condition for two physical things to be spatially distant from one

another is for them to coexist.”232 Therefore, it follows, on Pitcher’s view, that, although I

(allegedly) see the no-longer-existent star, my body is at no distance from it. Or, as Houts puts

it, “I (or my body) am not now at any spatial distance from the past, presently nonexistent star I

see. So, sometimes, some of the physical things we see are not at any spatial distance from

us.”233 This is a disconcerting conclusion.

Houts shows by further argument that Pitcher is committed to denying three other, more

general, and seemingly self-evident theses: namely, that (1) “at a time t, we (or our bodies) are

at some spatial distance from” all “events and stages we perceive at t,” that (2) “all the spatially

noncontiguous events and stages we perceive at a” certain “time are or were at some spatial

distance from one another,” and that (3) “we sometimes perceive at-a-time events and stages

which are or were in a three-dimensional array.”234 In other words, one who accepts Pitcher’s

(and a multitude of other philosophers’) way of dealing with the TGA by denying the

simultaneity-requirement is forced to say that, for example, when at night I look at my hand, the

moon, and the stars, these objects are not spatially related to each other in any three-dimensional

array. This result, claims Houts, is a reductio ad absurdum of such theories.235

He concludes his paper with these words:

It is important for philosophers who agree with Pitcher to recognize that their response to

the time-lag argument entails the falsity of propositions (1)-(3). It is then requisite for

them to face the implications of this result in light of their allegedly "commonsensical"

and "scientific" theories of Direct Realism. This is a problem about which these

philosophers have been reticent too long.236

Thus Houts issued a challenge; but the philosophers challenged continued to be reticent for

another three decades: as far as I am aware, Houts’s gauntlet was picked up only by John

Kardosh in 2008 and Sean Enda Power in 2010. As I find Kardosh’s paper outside my area of

competence, I’ll discuss only Power’s.

232 Houts, “Some Implications of the Time-Lag Argument,” 151. 233 ibidem. 234 ibidem, 156. 235 ibidem. 236 ibidem, 157.

151

Power met the challenge by forthrightly conceding that the TGA refutes Direct Realism, but

only on the assumption of Presentism. He also agreed that, taking Presentism for granted,

Houts’ assumption that things at different times are not related spatially is correct; given

one thing is present, the things at other times are not present; thus they do not exist.

Lacking existence, they do not have spatial relations with anything (even if they once

did). Thus, if I seem to perceive them in any spatial relations, this perception would be

illusory.237

Instead, what Power recommends is to abandon Presentism in favor of Eternalism. Then Direct

Realism can be saved—not, indeed, in Carrier’s way (by holding fast to the simultaneity-

principle and claiming that in all cases of perception we see light) but in the way of his

opponents who (a) deny the simultaneity-principle; (b) claim the superiority of a completely

tenseless way of speeking; and (c) insist that we do see the star itself (in the disintegrated star

case) because, though it exists no longer, it still exists in an “unqualified” sense. In addition,

claims Power, Eternalism is better supported by modern physical theory and is more compatible

with Relativity than is Presentism.

We will return to this issue in 6.1.

4.14. Conclusion about the Simultaneity-Requirement

I am ready now to state explicitly, and to expound in a fully detailed and elaborate way, why I

think the simultaneity-thesis is a conceptual, necessary, and analytic truth.

To see a thing is to come into visual contact with it. That means being able not only to look at

it, but also to look at it more closely, to approach it, and to touch it. Therefore, it must exist. For

all the things (actions) just mentioned to be able to happen, it, the thing seen (the thing one has

come into visual contact with) must exist. And it must exist at the time it is seen: it can neither

cease to exist before it is seen nor start to exist after it is seen. That it cannot start to exist after

it is seen may seem self-evident; it should be no less self-evident that a thing cannot cease to

exist before it is seen.

Seeing something puts me into perceptual contact with it. But if I’m in perceptual contact

with something physical, it should be logically possible for me to enter into some physical

contact with it as well. Yet if it doesn’t exist at t, then I can’t (it’s logically impossible for me to)

enter, at t, into physical contact with it. Therefore, I’m not in perceptual contact with it either.

237 Powers, “Perceiving External Things and the Time-Lag Argument,”

152

A thing, in order to be seen, must be able to be physically looked at, pointed to or at,

approached, touched, prodded, poked, handled, affected, influenced, changed, destroyed, or

otherwise acted upon or interacted with. If it can’t be physically looked at, pointed to or at,

approached, touched, prodded, poked, handled, affected, influenced, changed, destroyed, or

otherwise acted upon or interacted with, then it can’t be seen. But if it can be seen, then it can

also be physically looked at, pointed to or at, approached, touched, prodded, poked, handled,

affected, influenced, changed, destroyed, or otherwise acted upon or interacted with. And it can

only be physically looked at, pointed to or at, approached, touched, prodded, poked, handled,

affected, influenced, changed, destroyed, or otherwise acted upon or interacted with at time t, if

it exists at time t. In other words, it can’t be physically looked at, pointed to or at, approached,

touched, prodded, poked, handled, affected, influenced, changed, destroyed, or otherwise acted

upon or interacted with at t, if it does not exist at t (for example, because it only existed at a time

previous to t, or because it only will exist at a time after t, or because it never existed, exists, or

will exist).

Thus the main reason why a physical thing that no longer exists now cannot be seen now is

that it cannot—logically cannot—be physically looked at, pointed to or at, approached, touched,

prodded, poked, handled, affected, influenced, changed, destroyed, or otherwise acted upon or

interacted with now. And this physical thing that no longer exists now cannot be the object of

such present acts precisely because it existed only in the past. But it must be able to be

physically looked at, pointed to or at, approached, touched, prodded, poked, handled, affected,

influenced, changed, destroyed, or otherwise acted upon or interacted with now, if I or anybody

else is to see it now.

None of these statements seems to me open to the least doubt. I can’t imagine how they

might be questioned. Therefore, I think, these statements constitute a solid proof of the

simultaneity-principle.

But now suppose someone said in objection that there’s a certain star in the night-sky which

you are pointing to and looking at and therefore you think it exists but unbeknownst to you it

has gone out of existence hundreds of years ago. Isn’t this a possible scenario? And if so,

wouldn’t this be a case of your looking at something that no longer exists and thus a case that

falsifies your claim that you can only physically look at something if it exists now?

No, this is not a possible scenario. What makes it impossible is this: if you can physically

look at something, this implies there is a line that can be drawn between yourself and that thing.

But if that thing doesn’t exist, then there is no such line, and thus you are not physically looking

at that thing. It only seems to you that you are pointing to, and physically looking at, a star, one

that exists not in your mind only; what you are really pointing to, and looking at, is the night-sky

153

with some star-like lights which only exist in your mind. (How this is to be understood, and

what it amounts to, will be explained in our conclusion in Chapter Six.)

Pierre Jacob has claimed that

[s]ome of the things that [people] can see are objects that they can also reach, grasp and

manipulate with their hands. Many of the things that they can see, however, are not

objects that they can reach and grasp. As Austin pointed out, humans can see, e.g.,

mountains, lakes, liquids, gases, clouds, flames, movies, shadows, holes, stars, planets,

comets, and events.238

But even if they “are not objects that they can reach and grasp,” at least they are objects that

they can act upon or interact with in any of the other ways enumerated above, or they are events

in which such objects (as can be acted upon and interacted with in these ways) essentially figure.

I conclude that the simultaneity-requirement is entailed by the existence-requirement and that

the proposition “For all (x), x is seen at t if and only if x exists at t” is a conceptual, analytic, and

necessary truth. It is conceptual because its truth follows on reflection from an understanding of

the concepts involved; for exactly the same reason it is analytic; and it’s a necessary truth

because no counterexamples to it can be thought of.

238 Jacob, “Grasping and perceiving objects”, 1.

154

Chapter Five: The Representational and Adverbial Theories of

Perception

So far I have mostly assumed the subject-act-object and subject-state-object (content) models of

perception and mostly focused on the objects of perception, which are either sense-data or

external objects. In this chapter I will also focus on the perceiving of these objects and, in

particular, on two theories about that perceiving: the Representational and the Adverbial

Theories.

5.1. Dretske and the Sensory Core

In conclusion of his chapter on what he calls ‘non-epistemic seeing’ and from which I have been

quoting previously, Fred Dretske says:

One occasionally hears it said that if one systematically strips away from a given

perceptual act all the accretions due to past experience, all the collateral information,

anticipations, interpretive and inferential elements, all the habitual or conditioned

associations, then one will be left with a ‘pure sensory core’—the given of sense

experience. Such ‘stripping operations’ are taken very seriously by those who propose

them. We begin by seeing a plump juicy tomato and finish by being told that all that is

really given is a bulgy red patch, not really edible at all. What I have tried to show in this

chapter is that we have a way of seeing plump juicy tomatoes which, when subjected to

this ‘stripping operation’, leaves, as the sensory core or the directly given, precisely the

same plump juicy tomato with which we began. If S, as we commonly say, sees a tomato,

then we can supply him with the mentality of a one-year-old, take away all his past

experience of tomatoes, subtract whatever beliefs he has about the tomato, allow him no

inferences or interpretations, give him nothing that is not indubitable from the experience

itself, and we are left with a simple residue: S’s seeing a tomato. The tomato is the

sensory core, the directly given, if these phrases are meant to signify what it is that S sees

when this is purified of all inferential, interpretive, and discursive or associational

elements.239

If ‘core’ can be taken to mean ‘content’, and I don’t see why it can’t, then Dretske can be

taken to claim that the content of the sensory experience of the tomato is the tomato itself. This

is exactly analogous to my claiming that the content of the sensory experience of a star is the

239 Dretske, Seeing and Knowing, 75-76.

155

star itself. Recall two premises of my (Robinson-suggested) argument for the simultaneity-

thesis:

(2) The content of a psychological state is what (in whole or in part) constitutes that

state, what (in whole or in part) makes it the state that it is

and

(3) What constitutes the direct perceptual awareness of a star, what makes it the state that

it is, is, in part, the star one is directly perceptually aware of.

I take it that the expressions “direct perceptual awareness of a...”, “act of perceiving a...”,

“sensory experience of a...” are at least roughly equivalent here. Thus if Dretske were to accept

(2) and (3), he would be well on his way to accepting the conclusion (the simultaneity-thesis).

But in an e-mail letter from 2002, speaking of himself, he said that “I am ready to agree . . . that

the content of a mental state constitutes that state . . . but I doubt whether the (external) object

(the external object that the state represents as F), is part of that content.” Rather, “I think the

content of a mental state is completely given by the properties that the state represents

something as having. . . . an experience of a star (a distant object that no longer exists) as a

bright shiny object has the same content as an experience of a nearby object (which still exists)

as a bright shiny object.” Thus, while in 1969 Dretske might have accepted that the star is the

content of a certain mental state in just the way the tomato is the content of another, a few

decades later his view is different: according to that new view, it is no longer the star (tomato)

itself that is part of the content constituting a mental state; rather, this content consists entirely

of the properties the mental state represents the star (tomato) as having.

For what reasons did Dretske change his mind? And are they good ones? These are

legitimate and interesting questions: the first, for (a small part of) the history of philosophy; the

second, more broadly, for the philosophy of mind and of perception, in general. What I

minimally need to do here is only to evaluate how Dretske’s new theory bears on the TGA; but

this will require some verdict on its plausibility, and also an answer to the second question.

5.2. Travis and Alston on Representationalism

I am impressed by some of the arguments given by those philosophers, primarily Charles Travis

and William P. Alston, who think the Representationalist Theory of Perceptual Experience (also

called representationalism, representationism, intentionalism) to be mistaken. Alston in

particular focuses heavily on Dretske and makes some general suggestions as to why he and

other representationalists might have come to find the theory attractive. Travis doesn’t refer to

Dretske at all but formulates the theory untendentiously to fit all representationalists as follows:

156

“Perhaps the most common view of perception today is that it is representational: that in our

perceptual experience—in our seeing, hearing, feeling, smelling, what we do—the world is

represented to us as being thus and so.”240

He then goes on to say that he knows of no argument that any defender of the view has given

for it. “Rather it is assumed from the outset.” Some philosophers perhaps assume it “faute de

mieux, seeing representation as better material than ‘qualia’ for answering a very special

question as to what an experience was like.” Others “may be moved by the thought that

perceptual experience, being mental, is intentional, and that intentionality just is that sort of

aiming at the world which representation is.” Yet intentionality, according to Travis, is “but one

form of the mental: perception, and experience, exemplify another. In any event, perception is

not representational.”241

Travis starts out his argument by harking back to John L. Austin. Of himself and his

relationship to Austin he says “that the points I will make against the representationalist view

differ little, if at all, from points Austin makes in Sense and Sensibilia . . . , a remarkably rich

work. Were Austin not so thoroughly ignored, perhaps I would not have written this.”242 Then

he quotes Austin’s observation that “[i]n fact, of course, our senses are dumb. . . . our senses do

not tell us anything, true or false” and elaborates as follows: “Austin’s idea is that, rather than

representing anything as so, our senses merely bring our surroundings into view; afford us some

sort of awareness of them.” In this respect, Travis maintains, Austin is of one mind with

Descartes:

Sensory experience is, for Descartes, one more case where I am simply confronted with

‘ideas’. I cannot be confronted correctly or incorrectly, veridically or deceptively. I

simply confront what is there. Perception leads me astray only where I judge

erroneously, failing to make out what I confront for what it is. . . . [P]erception is a

source of unmediated awareness. I will call awareness mediated if it is hostage to

awareness of something else: that further awareness is part of what entitles one to take it

that X is so, or present; so part of what qualifies one as aware of that. In unmediated

awareness, one’s entitlement to take it that X is hostage to no more than some form of

awareness of X itself (such as seeing it).243

Travis doesn’t add, but I do, that Austin should also have been (but unfortunately wasn’t) of

one mind with Berkeley who, after all, adopted Descartes’s view on this point. Just recall (for

one of many things on this topic) what Philonous says of the man who thinks an oar with one

240 Travis, “The Silence of the Senses,” 57. 241 ibidem. 242 ibidem, 64. 243 ibidem, 65.

157

end in the water to be crooked: “He is not mistaken with regard to the ideas he actually

perceives; but in the inferences he makes of his present perceptions.”244 The Descartes-Berkeley

view was also endorsed by Kant when he stated that “truth or illusion is not in the object, in so

far as it is intuited, but in the judgment about it, in so far as it is thought. It is therefore correct to

say that the senses do not err—not because they always judge rightly but because they do not

judge at all. . . . In the senses there is no judgment whatsoever, neither a true nor a false

judgment.”245

These philosophers would all have rejected the Representational Theory of Perceptual

Experience, had they thought of it. Travis, in his paper, goes on to argue very carefully and

exhaustively that representationalism “is mistaken: there is nothing in perceptual experience

which makes it so that in it anything is represented as so (except insofar as the perceiver

represents things to himself as so). . . . . Perceptual experience is not as such either veridical or

delusive. It may mislead, but it does not take representation to accomplish that.”246

Let us now turn to Alston, who in one fell swoop criticizes both representationalism and

adverbialism. Having claimed that “Representational theories of perception were very prominent

in the 17th and 18th centuries” and that “contemporary representationists are closer in some

respects to these forebears than they care to admit,” Alston says that the “mid-twentieth century

in Anglo-American philosophy of perception was dominated by a reaction against the sense

datum-theories that had been so prominent in the first third of the century”; and that before

sense-datum theories had “collapsed under the weight of an impressive accumulation of

difficulties,” opposition to them “up until the last few decades of the 20th century” was

“spearheaded” by the Adverbial Theory of Perception.

This is “the view that perceptual experience, instead of being a direct awareness of objects

(public or private), is a way of being conscious and in this respect on a par with feeling anxious,

relieved, or excited and other apparently objectless mental states. This has the advantage of

doing justice to the role of sensory experience in perception, while avoiding any commitment to

non-physical, private objects of awareness.”247

Alston then says that the adverbial theory

itself faces serious difficulties. For one thing it conflicts with the apparent fact that what

is most distinctive to perceptual experience is the presentation of objects to

consciousness. Perceptual experience doesn’t seem at all to be an objectless mode of

consciousness like feeling anxious or relieved. This defect is reflected in the artificial (at

244 Berkeley, Works, 3 D, 238. 245 Kant, Critique of Pure Reason, 1787/1929, A294/B350. 246 Travis, “The Silence of the Senses,” 57. 247 Alston, “Perception and Representation,” 254-255.

158

best) way in which adverbial specifications of sensory consciousness are constructed by

taking familiar ways of specifying perceived objects and mechanically turning them into

adverbs—sensing flowering-crab-applely, blue spherely, bottle of winely, etc. There are

in addition a variety of more technical problems. . . . These difficulties emerge only

when thinkers take the adverbial idea with full seriousness. More often, people who

think of themselves as adverbialists are content to specify a particular perceptual

experience in such ways as “It was just as if I were seeing an apple tree,” without

making any serious attempt to spell out just what kind of similarity with veridical

perception is being attributed.

In this last gambit it is barely below the surface that adverbialism is designed not only

to avoid commitment to sense data, but also to accommodate something taken as a

sacrosanct datum by virtually every account of perception from the 17th century on—

viz., the possibility of a perceptual experience’s being of exactly the same intrinsic

character in veridical perception and in complete hallucination. This supposed possibility

has been thought to render impossible any serious direct realism, in which the

experience involved in veridical perception is a direct awareness of an external physical

object, or, to reverse the direction of description, a direct presentation of an external

object to the subject’s consciousness. For, it is supposed, in an hallucination there is no

such presentation; and if the experience involved there is intrinsically just like one in a

veridical perception of an external object, then in the latter case as well the experience

cannot be constituted as direct realism would have it. An adverbial characterization of

the experience that mentions no object at all is in at least as strong a position to realize

this desideratum as a sense datum theory that builds a relation to a private object into

veridical perception and hallucination alike.

Against this background we can understand the current popularity of

representationism. It seems to its proponents to provide a way of enjoying all the

desiderata mentioned above. By thinking of the experience involved in an object’s

looking a certain way, e.g., smooth, as a matter of the subject’s having a certain kind of

representation of the object as smooth, it takes the experience to be purely intra-mental

(and hence capable of being intrinsically the same whether or not there is an real smooth

object being perceived), while at the same time accommodating the intuitive object

directedness of perceptual experience by construing this as a representation of an object

as bearing certain characteristics, the representation obtaining whether or not this

intentional object turns out to be real and, if it is, whether it is as it is represented as

159

being. At the same time it avoids commitment to an internal direct object of awareness

(sense datum) that mediates the perception of the external object.248

Alston then proceeds to criticize representationism, with great care and at great length, first by

clarifying the basic concepts and then by arguing for the following four claims:

(1) There is no sufficient reason for positing a representative function for perceptual

experience. It doesn’t seem on the face of it to be that, and nothing serves in place of

such seeming. (2) Even if it did have such a function, it doesn’t have the conceptual

resources to represent a state of affairs. (3) Even if it did, it is not suited to represent,

e.g., a physical property of color. (4) Finally, even if I am wrong about the first three

points, it is still impossible for the phenomenal character of the perceptual experience to

consist in its representing what it does. My central argument for this central claim of the

paper is that it is metaphysically, de re possible that one have a certain perceptual

experience without its presenting any state of affairs. And since all identities hold

necessarily, this identity claim fails.249

Since Alston, by his own admission, “lays the greatest weight” on his argument for claim (1),

this is the only one I’ll highlight. It is the argument “that we lack sufficient reason for positing

any such representation” at all and “that in the absence of such a sufficient reason there is no

basis for attributing a representative function to” perceptual experience.

The only other basis there could be is that [perceptual experience] presents itself, is

experienced as, a representation. But that is clearly not the case. When something I see

looks a certain way to me (conical, red...), it doesn’t appear on the face of it be a

representation of anything. The mind is not irresistibly conveyed to something it is

representing the way the mind is when one looks at a (realistic) painting or a photograph.

The experience is not of that sort. Phenomenologically it has the character of a

presentation of an object as being such-and-such. The experience terminates in the

object presented without, so far as it appears, functioning to put [the subject] in mind of

something else. Hence we need a reason beyond the phenomenological character of the

experience to take it to be a representation.250

Dretske and other representationalists do give “a reason of this sort”; they, says Alston,

suppose that perceptual experiences

must represent facts about perceived objects in order to perform their function of

providing the belief-desire system with information concerning those objects. Everyone

248 ibidem, 255-256. 249 ibidem, 253. 250 ibidem, 275-276.

160

recognizes that this is (at least a basic part of) what perception is for. And, they claim,

[perceptual experience] could not perform this function without constructing such

representations.251

But this account, though possible, is not the only available. Alston presents another:

furnish[ing] the belief-desire system with information about perceived objects . . . could

also be done by [the perceptual experience’s] simply being what it introspectively seems

to be, viz., the object’s presenting itself as such-and-such . . ., provided the belief-desire

system has the capacity to read off of that the ways the object presents itself and to

encode those ways in the form of one or more propositions, i.e., encode them as

representations of states-of-affairs. And why should it not have this capacity? That is

what it is fitted to do—construct propositionally shaped representations and make use of

them in thought and motivation. On this alternative account all the representing is done

on the recipient side of the transaction between experience and thought. There is no need

for the donor side to construct any representation. I take this alternative to the

[representationalist] picture to be superior on the grounds of simplicity and economy.

Provided the belief-desire system has the capacity it posits, there is no need for any

representing on the [perceptual experience] side; it would be a fifth wheel that is making

no contribution to the outcome.252

With respect to Alston’s final claim (4), I’ll just remark that it revolves around the very

plausible consideration that some very young human perceivers, as well as some non-human

ones, could have what we would “recognize as a perceptual experience without having the

conceptual wherewithal to represent any facts, physical or otherwise, concerning putatively

perceived objects.” As Alston says, and I can’t help but agree,

when we think of the likes of frogs and insects (and even dogs and cats) . . . it does seem

eminently possible that an object can look a certain way to a frog without the frog’s

representing the object as having the kind of physical property that is responsible for its

looking that way, or indeed any physical property at all. Even if frogs can do something

that could properly be called ‘experientially representing some states-of-affairs

concerning perceived objects’, it strains credibility to the utmost to imagine those

representations picking out some particular physical property of the object.253

251 ibidem, 276. 252 ibidem, 276-277. 253 ibidem, 285-286.

161

This point is one that Dretske might find especially bruising—in view of his insistence, in

Seeing and Knowing, that the account of perception he gives there applies to infants and animals

as well. I will not pursue this point further here.

5.3. Jackson on Representationalism

Frank C. Jackson, who for much of the second half of the last century used to be a champion of

sense-data and in 1977 even wrote a classic defense of the Representative Theory of Perception,

towards the end of that century changed his mind about sense-data. In a draft paper from 2000

called “Some reflections on representationalism” he explained why. Since this paper wasn’t

published in a journal or book, but just put on the Internet explicitly as a draft, my emphasis,

too, will not be so much on Jackson personally as on the ideas contained therein. Since many of

them are not Jackson’s alone but have been floating around in representationalist circles for

years, they’re worthy to be criticized (even if Jackson himself might put them a little or even

very differently in 2011).

Still, I’ll start with a personal confession of Jackson’s. Speaking of himself, he says:

The reason I abandoned the sense datum theory was my belated realisation that it fails to

capture the representational nature of perceptual experience.254

Earlier in the paper he had stated that it should be

no surprise that, as a former sense datum theorist, I find representationalism very

attractive. Both theories see the nature of experience as lying in the properties of the

objects of experience, with the big difference that, for representationalists, the properties

of the objects of experience reside in the way that experience represents things as being.

There need be nothing actually having the properties; the ‘objects’ are intentional

objects.255

He also states that both the sense-datum and the representationalist theory subscribe to

the act-object account of sensory experience. The act-object account of sensory

experience captures the nature of experience through the properties of the objects of

sensory awareness. On the act-object view, the difference between an itch and a pain lies

in the difference between what one is aware of, and not in the mode of awareness as in

adverbial theories.256

254 Jackson, “Some reflections on representationalism,” 2. 255 ibidem, 1 256 ibidem.

162

Adverbial theories in his opinion are wrong, and Jackson now likes how representationalism

“handles the many-property problem” he “raised for adverbial theories of sensory experience

some years ago.” At that time he

asked how theories that talk of ways of sensing can capture the fact that having a red,

round after-image involves, in some sense, the properties of redness and roundness

attaching to the very same thing. It is, for example, wrong to analyse having a red, round

after-image simply in terms of sensing redly and sensing roundly. For we must

distinguish having a red, round after-image from having a red after-image at the same

time as having a distinct round one. Representationalism handles this problem nicely in

terms of the difference between having an experience that represents that there is

something that is both red and round before one, and having an experience that

represents that there is something red before one and that there is something else round

before one.257

But Jackson’s main foe is now the sense-datum theory, about which he says:

The central objection to the sense datum theory is that it is a classic piece of buck

passing. We should all agree the experience of, say, there looking to be something red in

front of one represents that there is, in the world, something red in front of one.258

Now my question is: Why should we all agree on this? Is it because it is self-evident? A truism?

Indisputable? No one in fact disputes it? Jackson doesn’t let on. As a matter of fact, there are

some serious philosophers who do dispute it. Perhaps Jackson is saying that even if it isn’t

exactly self-evident, a truism, or indisputable, there are some good arguments supporting it. But

what are they? Again, Jackson doesn’t give a hint.

This is in line with the observation made by Alston and Travis that representationalists often

don’t give arguments for why they are representationalists. In any case, let’s look at this

proposition that Jackson claims we should all agree on:

the experience of there looking to be something red in front of one represents that there

is, in the world, something red in front of one.

What does this actually mean? Is it, and why is it, something worth saying? What is gained by

pointing out that the experience of there looking to be something red in front of one represents

that there is, in the world, something red in front of one? What word could be substituted for

“represents” here? “Shows”? “Portrays”? “Means”? “States”? “Carries the information”? They

all, though similar, come to something just a tad different. And why is any of them needed?

257 ibidem, 1-2. 258 ibidem, 2.

163

Instead of saying what he did, couldn’t Jackson simply have said the following: “If there

looks to be something red in front of one, then there is, in the world or in one‘s mind, something

red in front of one”? This would have been simpler and clearer. To speak bluntly, Jackson’s

phrase

the experience of there looking to be something red in front of one represents that there

is, in the world, something red in front of one

sounds stilted and wordy: the expressions experience of and represents that sound like

circumlocutions; they seem to be empty verbiage.

Immediately after the quoted phrase Jackson continues:

It is definitive of perceptual experience that it represents how our world is. We can all

agree on this independently of whether we think that representational content exhausts

the phenomenology. This means that sense datum theorists’ analysis of the experience of

there looking to be something red in front of one in terms of direct awareness of

something mental that actually is red, the sense datum, must preserve this feature. But

then the key relation of direct awareness to something mental must be explicated as

involving representing that there is something red in the world, and we have gained

nothing. It is like analysing knowledge as belief that constitutes knowledge. What

confuses sense datum theorists, or confused me anyway, is the thought that the

requirement that there be something which is red of which the subject is directly aware,

automatically captures, or part way captures, the key representational notion. This is a

mistake.259

I would object once more to the the very first sentence of this quotation. I would say that what

is definitive of perceptual experience is, at most, that it presents (not: represents) how our world

is. Consequently, I would also reject the second, third, fourth, and fifth sentences. There simply

is no such thing as the “representational content” of a perceptual experience; a fortiori it is not a

feature that any sense-datum analysis must preserve, nor must the “key relation of direct

awareness” to a sense-datum “be explicated as involving representing that there is something

red in the world.” In fact, the sense-datum theory seems to be all right as it stands: it doesn’t

need to “preserve” or to “explicate” what Jackson (currently) thinks it does.

Hence, with respect to the last two sentences of the above quote, I would say the following:

it is indeed a mistake to think “that the requirement that there be something which is red of

which the subject is directly aware, automatically captures, or part way captures, the key

representational notion”; but the mistake is not that it doesn’t capture this notion (indeed it

259 ibidem, 2-3

164

doesn’t), the mistake is thinking that it, a “representational notion,” is required at all—and that

it is “key”.

Yet not being able to let go of this notion Jackson elaborates:

It is true that I can represent how I am representing something to be by using the actual

way something is. For example, I might represent to you the colour I remember the

murderer’s coat to be by holding up an actual sample of the colour. Here I would be

using the actual colour of one thing, the sample, to represent how my memory represents

the colour of something else, the murderer’s coat, to be; a colour which the coat, of

course, may or may not have. In that sense, we have a model for understanding the sense

datum theory (and it is noteworthy that it would be an obvious mistake to run the

‘reification of appearances’ objection against this way of representing how I remember

things to be). But, and this is the crucial point, the fact that I am using an actual sample

of the colour cuts no representational ice per se. I could be using the sample to represent

the one colour I do not think the murderer’s coat to be. Or I could be following the

convention of holding up a sample with the colour complementary to that I remember the

murder’s coat to be. In the same way, standing in a certain direct-awareness relationship

to a mental item with such and such properties says nothing, represents nothing, per se,

about how the world is. The—I now think, extraordinary—failing of the sense datum

theory is that it does not start to address the representational nature of perceptual

experience. It somehow manages to leave out the most important part of the story.260

But this is “the most important part of the story” and leaving it out is an “extraordinary failing”

only if you are in the grip of a false picture—that perceptual experience has a “representational

nature”. It doesn’t; its nature is presentational.

This picture is additionally false in having the world’s objects be represented, rather than

idealistically constituted out of the sense-data. What is really the case, in the true story, is that

perceptual experience directly presents sense-data; and through them, in most cases, it directly

presents the objects themselves. In other words, on the idealist story, “standing in a certain

direct-awareness relationship to a mental item with such and such properties” is perfectly

enough for showing “how the world is”; no representing whatsoever is needed.

When he was a traditional Representationalist (in the spirit of his 1977 book Perception),

Jackson could have moved toward idealism (without doing too much violence to his then-

avowed beliefs) while still adhering to sense-data. But upon becoming a physicalist a few

decades later, he, in his own words, “had another good objection to the sense datum theory.

260 ibidem, 3.

165

When something looks square to one, there need be nothing physical that is square in existence.

But, for the physicalist, the physical things are all the things.”261 To me this sounds not like an

objection to sense-datum theory, but, on the contrary, like a reason to posit sense-data: the

objection would only be to physicalism.

Frank Jackson’s move from traditional Representationalism to its contemporary version is

perhaps best reflected in an article he wrote in 1992262:

I am watching a football game on television. My senses, in particular my eyes and ears,

‘tell’ me that Carlton is winning. What makes this possible is the existence of a long and

complex causal chain of electro-magnetic radiation running from the game through the

television cameras, various cables, my television set and a region of space between my

eyes and the television screen. Each stage of this process carries information about

preceding stages in the sense that the way things are at a given stage depends on the way

things are in preceding stages. . . . There needs to be systematic covariance between the

state of my brain and the state of the march, and that will not obtain unless it obtains

between intermediate members of the long causal chain. . . .

A few of the stages in this transmission of information between game and brain are

special in the sense that I am in some sense perceptually aware of them. . . . For instance,

I am perceptually aware of the images on the screen. I am also perceptually aware of the

game. Otherwise I could not be said to watch the game on television. Now my perceptual

awareness of the march depends on my perceptual awareness of the screen. The former

goes via the latter. . . . if you suddenly covered the screen with a cloth and asked me (1)

to report on the images, and (2) to report on the game, I might well find it easier to report

on the game than on the images. But that does not mean that my awareness of the game

does not go via my awareness of the images on the screen. It shows that I am more

interested in the game than in the screen, and so am storing beliefs about it in preference

to beliefs about the screen.

But if my awareness of the game goes via my awareness of the images on the screen, what does

my awareness of the images on the screen go by? The latter awareness is certainly not direct. So

what does it causally depend on? Jackson goes on to say that

[o]ur initial statement of Representative Realism talked of the information acquired in

perceiving an object being most immediately about the perceptual experience caused in

us by the object, and only derivatively about the object itself. In the act/object, sense-

261 ibidem, 3 (footnote 3). 262 Frank Jackson, “Representative Realism,” in: Jonathan Dancy and Ernest Sosa (eds.), A Companion to Epistemology: 445-448. What follows are quotations from the end of that article (pp. 447-448).

166

datum approach, what is held to make that true is the fact that what we are immediately

aware of is a mental sense-datum. But instead, Representative Realists can put their view

this way: just as awareness of the match goes via awareness of the screen, so awareness

of the screen goes via awareness of experience, and in general when subjects perceive

objects, their perceptual awareness always goes via awareness of experience.

These passages illustrate nicely the beginning of Jackson’s slide from traditional to

contemporary Representationalism (as well as Alston’s claim that the difference between them

isn’t all that great). In Jackson’s encyclopedia article, “experience” has replaced “mental sense-

data” as the object of immediate awareness. But this isn’t quite yet contemporary

Representationalism in full dress, for thae latter does not require that the perceiver be aware of

the experience at all.

As William Seager and David Bourget explain in their “representation-friendly” paper

“Representationalism about Consciousness” from 2007, contemporary or Modern

Representationalism (MR), as they call it, claims that

perception is representational in the sense that it is intentional, not in the sense that it is

mediated by “internal pictures.”. . . When someone perceives a cardinal they are

perceiving a bird and not one of their own mental states. While dependent upon there

being an active mental representation within the perceiver, proponents of MR deny that

perception is indirect, proceeding via an apprehension of this representation. . . . There is

no “maple leaf in the head.” When one imagines a Canadian flag, one is aware of the

content of a flag-representation which encodes shape and color information. This

encoding does not have to be flag-shaped and colored red. Whatever the vehicle of this

representational content might be, perhaps a neural state, there is no need for there to be

any awareness of it.263

Let me end this section by returning to traditional Representationalism and its refutation. Even

though thinking of himself as a traditional Representational Realist, O’Shaughnessy, as we’ve

seen in 3.4., denies that seeing something, say a theater play, on TV is really seeing the play (as

you would if you were sitting in the theater itself rather than in front of the TV). This accords

well with what Foster says in criticizing the view that “someone watching a live football match

on television” is really seeing the match itself. Foster, I think, has most effectively disposed of

any such view as Jackson’s on p. 212-218 of The Nature of Perception. It is a terse but

compelling dispatch of Indirect Realism as a whole, as is David Smith’s even shorter argument,

263 Seager and Bourget, “Representationalism about Consciousness,” 3,

167

on p. 13-16 of The Problem of Perception, attempting to show that, as “a matter of the first

philosophical importance, . . . Indirect Realism is incoherent.”

5.4. Valberg and Noë on Representationalism

I think the best take on these issues is Jerry Valberg’s. He, too, distinguishes between two kinds

of Representationalism. One, that of the 1977 Jackson, is usually called the Representational

(Representative) Theory of Perception (or Representative Realism). Valberg calls it “the

traditional Representationalism of Descartes and Locke.” The other, what we’ve been calling the

Representational Theory of (Perceptual) Experience, he calls “contemporary

Representationalism” (much like Alston does in his paper, discussed above). This is the doctrine

now upheld by Jackson and Dretske, among many others, and argued against by Travis and

Alston. Valberg also argues against (both sorts of) Representationalism. With respect to what he

takes to be its contemporary version, he says the following:

I think that, when we conceive of experience as something which occurs in our heads (or

souls), it is extremely natural to adopt a Representationalist view of experience, that is,

to think of what is going on in the head as representing what is out in the world, in the

very way that a picture or image represents an object in the world. On this model, the

sensory content of experience will play the same role as the coloured shapes in a picture.

It will function, not as the object, but as the vehicle, of representation. The object, as we

said, will be the thing in the world—the external object.264

Now, according to Valberg, contemporary and traditional Representationalism differ as

follows: “[t]raditional Representationalism takes the vehicle of representation (the sensory

content, the ideas, etc.) to be what is present in experience.” But it does it in such a way that

“[t]he object in the world is represented by what is present in experience, but is not itself

present.” By contrast, “contemporary Representationalism does not place anything ‘between’

the subject and the world.”265 However, does contemporary Representationalism preserve the

presence of the world? Valberg says:

Suppose I make a picture of the book on my desk. The coloured shapes in the picture

represent the book (they are the vehicle), and in that sense the book is ‘in’ the picture.

Now consider the idea that my experience has a sensory content, and that, like the

coloured shapes in the picture, the sensory content of my experience represents the book.

So we could say that, in a representational sense, the book is ‘in’ my experience. . . . But,

264 Valberg, The Puzzle of Experience, 130-131. 265 ibidem, 131.

168

if the book is ‘in’ my experience in the sense that the sensory content of my experience

represents the book, what has happened to the fact that the book is present in my

experience? The danger is that we may forget about the fact of presence, or suppose we

are dealing with that fact in dealing with the fact of representation. These are, I want to

insist, utterly different facts; there is no way of equating them.266

Valberg goes on to repeat a previous point (made both by him and by me), namely, that

we have a strong inclination to view what is present in experience as existing when it is

present: if it is present now, it exists now. We have no comparable inclination with

respect to representation. In our example, the book represented by my picture exists now.

Would it matter if the book were destroyed? Would we be any less inclined to say that

the picture represents the book? Or think of it from the standpoint of experience: suppose

that the sensory content of my experience represents the book, and that the book is

present in my experience. Then God eliminates the book. We will want to say that the

book can no longer be present in my experience. But there seems no reason to deny that,

if the sensory content of my experience remains the same, it continues to represent the

book.267

Besides Valberg, Alva Noë, too, seems to have a good grasp of some of these issues. According

to him, perceivings are “episodes of grappling with the world itself”. He recommends we should

think of

perceptual experiences as temporally extended patterns of engagement with the world,

not as things that happen in us. We enact perceptual experience; it doesn’t happen to us.

Perceptual experiences, then, should not be thought of as representations, as internal

states that are about a scene. Rather, they are episodes of contact with a scene. This non-

representationalist conception of experience squares with perceptual phenomenology, I

would say. After all, perceptual experiences don’t feel like representations. It doesn’t

seem to us, when we see, as if what we experience is represented in our head. Rather, it

seems to us as if what we see is out there in the world. And it seems to us as if we have a

special kind of access to what is out there. Our sense of the presence of objects and

properties around us, in perceptual experience, is understood in terms of our being

skillfully poised to reach out and grasp them. Instead of thinking of perceiving on the

model of seeing, which is in turn understood on a kind of quasi-photographic or optical-

projective model, we should think of perceiving on the model of touching.268

266 ibidem. 267 ibidem, 131-132. 268 Noë, “Real Presence,” 47-48.

169

This is quite in line with my own reflections expounded earlier, especially in connection with

my rejection of John Bengson’s and Kendall Walton’s assimilation of seeing photographs to the

seeing of non-photographic reality. I would also endorse the following remarks (with the

exception of the second sentence) of Noë:269

For philosophers there may be a temptation to think of experiences as a kind of logical

act, comparable to an act of judgement or to assertion. We find it natural to think of

experiences as representations. But experiences are not acts, in this sense; they are not

representations; they are activities, events themselves; they are temporally extended

patterns of skilful engagement. When you perceive an event unfolding, it is not as if you

occupy a dimensionless point of observation. You live through an event by coupling

with it. What you experience is the event, as it plays out in time. You experience the

singer’s song, and the ball player’s play, and the dancer’s dance, by tracking what they

do over time. The very experience is a world-involving achievement of control and

attention.

All in all, I think that perceptual experience doesn’t represent anything as being so or as not

being so: perceptual experiences just aren’t representations. They’re not like thoughts, which

represent (whether truly or falsely) reality or irreality; they’re not like sentences, which express

thoughts and in so doing likewise represent reality or irreality; they’re not like pictures, which

can likewise represent reality or irreality, or which create their own, “artistic,” reality.

Perceptual experiences neither represent nor create anything; they just present the world, period.

Here I largely agree with Charles Travis when he says that “perception, as such, simply

places our surroundings in view; affords us awareness of them. There is no commitment to their

being one way or another. It confronts us with what is there, so that, by attending, noting,

recognizing, and otherwise exercising what capacities we have, we may, in some respect or

other, make out what is there for what it is—or, again, fail to. It makes us aware, to some extent,

of things (around us) being as they are. It is then up to us to make out, or try to, which particular

ways that is. Perception cannot present things as being other than they are. It cannot present

some way things are not as what is so. That would not be mere confrontation. So it cannot

represent anything as so. Representing, by nature, is liable to be of what is not so. . . . [I]n

perception things are not presented, or represented, to us as being thus and so. They are just

269 Noë, “Experience of the world in time,” 31–32.

170

presented to us, full stop. It is in making out, or trying to, what it is that we confront that we take

things, rightly or wrongly, to be thus and so.”270

The only reservation I have is regarding the very last sentence. When in perception, in

perceptual confrontation with things, they “are presented to us, full stop,” there is one respect in

which quite automatically, in that very perception, we do take them to be thus and so. Namely

(and here I would agree with John Foster), we take (interpret) these things as being located,

together with ourselves (the perceiving subjects), “in a common three-dimensional space,” and

located “in a way which, with varying degrees of specificity, purports to characterize the spatial

relationship between them from the subject’s standpoint (in the perspective of his

viewpoint).”271

But perhaps it isn’t necessary to see Foster (and myself) as really disagreeing with Travis. If

perception “simply places our surroundings in view” and “affords us awareness” of them, as

Travis has it; and if we and our surroundings exist in a common three-dimensional space with

various perceived spatial relationships between us, as Foster and I have it, then isn’t it the case

that when Travis speaks of the presentation of things to us, “full stop,” he need not be saying

something incompatible with Foster’s talk of “[t]he notion of an interpretation blending with a

presentation to form an integrated, conceptually enriched, episode of awareness” and of “cases

where the interpretation is an integral part of the whole experience, rather than an extraneous

judgement about it, or about the object perceived.”272 We will return to this issue at the end of

the next chapter.

5.5. The Adverbial Theory of Perception

In 5.2. I’ve already mentioned William Alston’s objections to adverbialism, the gist of which

lies in the claim “that it conflicts with the apparent fact that what is most distinctive to

perceptual experience is the presentation of objects to consciousness. Perceptual experience

doesn’t seem at all to be an objectless mode of consciousness like feeling anxious or relieved.

This defect is reflected in the artificial (at best) way in which adverbial specifications of sensory

consciousness are constructed by taking familiar ways of specifying perceived objects and

mechanically turning them into adverbs.”

270 Travis, “The Silence of the Senses,” 65. 271 Foster, The Nature of Perception, 151-152. 272 ibidem, 155.

171

The latter artificiality is due to what Frank Jackson calls “the many-property problem,”

Howard Robinson calls the problem of “structural complexity,” and John Foster calls the

problem of “sensible complexity” faced by adverbialism.

In an extended discusion of the Adverbial Theory of Perception, John Foster argues that the

linguistic objection can be overcome. In particular, he argues that “by recognizing an ontology

of episodes of sensing—an ontology of sensory acts”273 in place of sensory objects, “the

adverbialist can deal with the phenomenon of sensible complexity.”274

Nevertheless, the theory is “vulnerable at two other, and more fundamental points—points

which concern the substance of its philosophical claims, rather than the descriptive adequacy of

the associated language.”275

One of these is the point that what the Adverbial Theory “takes to be the true character of our

experiential situation is at variance with what an introspective analysis seems to disclose.”

Foster thinks that

when we focus on the character of our experiences introspectively, we do not become

aware of the supposed adverbial nature of their sensory core. If we accept the distinction

between the sensory and interpretative components of phenomenal experience—a

distinction recognized by both the Sense-Datum Theory and Adverbial Theory

accounts—and if we introspectively focus on the character of our experiences in the light

of this distinction, we can, indeed, come to identify what (in that framework) qualifies as

the sensory core. And my becoming aware of this core, we become aware of the sensible

qualities which feature in its content. But we are always aware of this core as the

sensing, or sensory awareness, of some item (for example, a colour-array, a felt patch of

texture, a sequence of sounds), and are aware of the sensible qualities involved as

aspects or elements of this item. We never manage to achieve, introspectively, an

explicitly adverbial view, where we are aware of the sensory episode as merely a sensing

in a certain manner (a manner not defined by the character of an object of sensory

awareness), and are aware of the sensible qualities as the modes of sensing which

compose this manner. But why not, if an adverbial view would be correct?276

After trying several other ways, and failing, to uncover “the supposed adverbial character of

the sensing,” Foster concludes “that the introspective situation shows the adverbial theory to be

highly implausible, even if it does not refute it absolutely.”277

273 Foster, The Nature of Perception, 175. 274 ibidem, 178. 275 ibidem. 276 ibidem, 179. 277 ibidem, 181.

172

What does so is the second point referred to above, namely, that the Adverbial Theory

“leaves us without any explanation of how phenomenal experience, in the relevant sense, comes

to occur. For if the sensory core is truly adverbial—the sensing in a certain manner, rather than

the sensing of a certain object—it becomes impossible to understand how we manage to

experience it as presentationally perceptive, or indeed as perceptive in any way.”278

I leave out the details of Foster’s critique and conclude, in his words, that “I do not see how

we can continue to regard the adverbial theory as a serious option.”279

5.6. Conclusion about the Representational and Adverbial

Theories

In Chapters One through Four I assumed the subject-state-object/content model of perception,

but made this explicit only in 4.5., insofar as I agreed with Tom Stoneham that this is Berkeley’s

default model. In this chapter I rehearsed objections to the intentional model, otherwise known

as the Representational Theory of Perception. Now I must record and justify my conviction that,

though false, it is, after all, compatible with both the simultaneity-principle and the postulation

of sense-data, as required by the TGA.

According to this model, which regards mental states as being like measuring instruments, a

mental state has content just like a thermometer can have content—by being in a

representational state. As Dretske explained, “a thermometer is in the same representational

state (has the same content) if it registers 95 degrees in location L (representing L as being 95

degrees) as it does when it registers 95 degrees in location L’ (representing L’ as 95 degrees).

Likewise, an experience of a star (a distant object that no longer exists) as a bright shiny object

has the same content as an experience of a nearby object (which still exists) as a bright shiny

object.” If that’s so, then the “bright shiny object” of which the experience is of can be a

physical object or a mental sense-datum; it can actually exist now, or in the past, or never have

existed (in which case it would be a hallucination); it can be an indeterminate distance away

from me; and it can either be a star, or my neighbor’s light, or whatnot—as long as it’s bright

and shiny.

Now, as it happens, and if I were to speak in this way, my experience (switching to

Armstrong’s example in 2.7.) of what I take to be a star’s vanishing and of a bird’s flight much

closer to me represents the star’s vanishing and the bird’s flight as occurring simultaneously,

even though the star vanished eons earlier. So the representation is false, but that is no skin off

278 ibidem, 185. 279 ibidem.

173

the Representational Theory’s back. It leaves virtually all options open. What it doesn’t do, of

course, is provide a premise for the TGA; but it is not incompatible with all of its premises and

conclusion.

The Adverbial Theory of Perception, by contrast, is not only thoroughly false (this quality it

shares with Representationalism), but it also is both useless for, and irreconcilable with, the

TGA. Cornman needed it so as to avoid sense-data. If he could have been convinced that the

Adverbial Theory is, though initially perhaps appealing, but ultimately quite hopeless, his case

against sense-data would have collapsed, and the TGA might have had one more unambiguously

supportive proponent.

174

Chapter Six: What We See

In Section 1.3. I mentioned some Direct Realists who attack the TGA but not by denying the

simultaneity-requirement. Besides Len Carrier, who presumed to avoid the TGA’s conclusion

(that we see sense-data) by claiming that we always see light, there are David Lewis and Hanoch

Ben-Yami who simply deny that a star can go out of existence before light from it reaches the

eyes of a seeing subject. According to Lewis, the stars “as I now see them are not

straightforwardly past; for light-like connection has as good a claim as simultaneity-in-my-rest-

frame to be the legitimate heir to our defunct concept of absolute simultaneity.”280 Thus we can

see stars as they are now.

6.1. Apparent Simultaneity

This view has been worked out more fully by Ben-Yami in his paper “Apparent Simultaneity.”

He claims that backward light-cone simultaneity, which he prefers to call apparent simultaneity,

has certain advantages over what he calls “Einstein simultaneity.” In a section of his paper titled

“Appearance as Reality,”281 Ben-Yami writes that

According to our pre-scientific approach, what we are now seeing is happening now.

This view presupposes that the speed of light from source to observer is infinite; and

indeed, most natural philosophers from Aristotle to Descartes argued that that is light’s

speed. With the advent of modern science, however, with its claim that light’s speed is

finite, this pre-scientific view had to be given up. Our improved knowledge of the world

(science) seemed to establish that the world is not as it appears to be. The things we now

see are not only not happening now, but they did not even happen together at all: we now

see the sun where and as it was eight minutes ago, but the moon where and as it was just

over a second ago. In this way a gap opened between how the world appears to us and

how we think it really is. The scientific image of the world became radically different

from its manifest or apparent image. And, since in our nonreflective moments, which

constitute the great majority of our active life, we act on the basis of the world’s apparent

image, we have to concede that our everyday attitude to the world involves an illusion.

280 D. Lewis, “Veridical Hallucination and Prosthetic Vision.” In: J. Dancy, ed. Perceptual Knowledge (OUP, 1988): 83. 281 All quotations of Ben-Yami are from his paper “Apparent Simultaneity.“ In [2007] EPSA07: 1st Conference of the European Philosophy of Science Association (Madrid, 15-17 November, 2007).

175

However, if one adopts backward-light-cone simultaneity, this gap partly closes . . .

According to backward-light-cone simultaneity, what we now see through vacuum is

what is happening now. The world is as it appears to be, in this respect. Science does not

force us to reject our everyday view of the world.

Since according to backward-light-cone simultaneity what appears to be happening

now, that is, what we now see, is indeed happening now relative to us, this simultaneity

can also be called Apparent Simultaneity. This is the phrase I shall use below.

Ben-Yami then discusses how the notion of “apparent simultaneity” may be fruitfully applied

to criticizing the TGA. He lays out the argument in the following steps:

1. Light reaches the seeing subject after it has left the seen object.

2. During that time-lag the seen object may have ceased to exist.

3. Necessarily, what we are directly aware of when we see an object,

exists at that moment.

4. Therefore, what we are directly aware of when we see an object is

not the object seen.

Ben-Yami notes that some “have tried to criticize” this argument “by claiming that what we are

directly aware of need not exist when we are aware of it. But I don’t find this plausible,” he

remarks and explains why:

When we see anything, whether near-by or far away, the qualities of which we are

directly aware are there for us to inspect and study, and we often can, if we wish, pay

them more or less attention. It seems preposterous to claim that they may not be

instantiated while these cognitive processes, focusing on them, are taking place. If the

argument is unsound, its fault lies somewhere else.

I fully agree with Ben-Yami’s explanation of why premise (3) is correct. Therefore, recognizing

the truth not only of (3), but also of (2), and wishing to avoid the conclusion, he has no choice

but to claim that

Apparent Simultaneity suggests a different way out: according to it, the argument’s first

premise is false. Relative to the observer, the event of seeing something is simultaneous

with the event being seen. Accordingly, what we are directly aware of when we see an

object may be the seen object. Direct Realism is unharmed by the Time-Lag Argument.

Ben-Yami adds the following:

To show that the Time Lag Argument is unsound, we do not have to accept Apparent

Simultaneity as the true simultaneity; it is enough that it is an optional simultaneity

definition. If that is so, then the argument’s first premise in not true simpliciter; its truth

is a matter of convention. Yet the argument assumes that its first premise is simply true.

176

I lack the competence in physics to assess whether Ben-Yami’s and Lewis’s response to the

TGA is adequate. However, I’m inclined to agree with what Bengson says (see 3.3.) when he

first states and then criticizes Lewis’s view. The view, according to Bengson, is that “Lewis

appears to believe that if is seen now, then because it is within an observer’s light cone, is

simultaneous with what is happening now— is “not straightforwardly past”, but present. If this

is so, then an explosion of Polaris that occurred almost seven centuries ago is present so long as

we see it now.”282 (Lewis himself claims that he got this idea from the physicist Eric

Mellum.283)

Bengson’s criticism of this view is that it is just plain wrong, for “presumably, such an

explosion does not occur in the present. Indeed, it does an injustice to our concept of the present

to apply it to that which occurred almost seven centuries ago: pace Lewis, we clearly consider

such an event to be past.”284

6.2. Intra-Organismic Time-Gaps

But even if there is no time-gap between the time light leaves a star and the time it arrives at our

eyes, there is another gap before we actually have a visual experience. As Ben-Yami explains,

Some might try to save the argument by claiming that even if light cannot be said to

leave the seen object before reaching the seeing subject, it still takes the nervous system

some time to transfer the signal from retina to brain, and during that time the seen object

may have ceased to exist.

In other words, if the TGA can be circumvented, along the lines of Lewis and Ben-Yami, for

the first time-gap, then the second one (between the time the light hits the retina and the time it

has been finished being processed in the brain) certainly cannot, and the TGA can successfully

move to its conclusion: what we directly see are sense-data and Direct Realism is false. To this,

Ben-Yami objects as follows:

But this response presupposes a Cartesian model of perception, as if the perceiving

subject is a homunculus located somewhere deep in our brain. If, following an approach

developed by Dennett [1991], we maintain that the time we see anything—we, the

embodied human beings—cannot be determined more accurately than the vague interval

between the activation of our retinas by the light and the subsequent activation of the

relevant parts of our brain, then this line of response is no longer available.

282 Bengson, “How to Perceive the Past with Your Eyes Shut,” 3. 283 Lewis, “Veridical Hallucination and Prosthetic Vision.,” 83, note 9. 284 Bengson, ibidem.

177

Fortunately, Dennett’s approach is not unassailable, and it has been assailed by Selmer

Bringsjord. He first claims that “the cornerstone of Daniel Dennett's case for his “multiple

drafts” view of consciousness in his Consciousness Explained is a set of inferences he draws

from the phi phenomenon.” Then he explains “why these inferences are incorrect, in part by

explaining phi (as replicated in our own lab) in such a way that it is consistent with the standard

conceptions of consciousness Dennett seeks to supplant.”285

If Bringsjord is correct, then Dennett’s rejection of the Cartesian model is unjustified and the

response criticized by Ben-Yami is still viable. But if so, then the TGA is effective with respect

to the intra-organismic time-gap as well.

Still, rather than delving into the Dennett and Bringsjord controversy, I suggest reconsidering

what we said, in 2.3., about Leibniz’s original TGA. There we claimed that when the distance

between the perceived object and the perceiver is too short and thus the time-gap between the

light’s leaving the object and its hitting the perceiver’s eye is too brief (as in the case of

someone’s looking at a painting), then the TGA cannot properly get a foothold. We also

mentioned a reply on Leibniz’s behalf as well as a rejoinder to that reply. The reply went as

follows:

the fact that we cannot discriminate very small time intervals is simply irrelevant.

Whether or not I can discriminate such an interval, I do in fact perceive [the picture as it

was some minuscule part] of a second ago; and it could, conceivably, have changed in

some perceptible respect just in that time, so that by the time I in fact perceive it it is

different from how I perceive it to be. The context is relevantly extensional and not

intentional: my cognitive abilities in minute time discriminations are not to the point.

And the rejoinder to this reply was that it misses the point: the perception and the content

perceived “must be phenomenologically contemporaneous, which means that it must not be

possible to discriminate any time distinction between perception and content by direct

experience. This condition is satisfied provided that the physical time difference is insufficient

to be noticed.”

But now we can bring up another point, namely, that this condition must be satisfied and

phenomenological contemporaneity is required only for the TGA to succeed. Other arguments

for the same conclusion (that in all cases of perception we perceive sense-data) require neither

phenomenological contemporaneity nor an unnoticeable time difference between the perceiving

and the perceived. Therefore, if there are any other successful arguments for this conclusion,

then these arguments will work in cases of the intra-organismic time-gap as well, and reliance

285 S. Bringsjord, “How to Explain the Phi Phenomenon Without Dennett's Exotica: Good Ol' Computation Does Just Fine, Abstract.

178

on the TGA to deal with that minuscule time-gap will not be needed. (But its over-all point that

during the combined time-gap between the time the light from an object leaves the object and

the time that a visual experience caused by it occurs, a sense-datum of the object is immediately

seen, still stands, of course.)

Let us, then, ask, Are there any other good arguments for this conclusion?

6.3. Other Arguments for Sense-Data

At least two others have been extensively (much more extensively than the TGA) discussed in

the literature. They are the Argument from Illusion and the Causal Argument (sometimes also

called the Argument from Hallucination.) Of these, the first, if successful, is the one that should

have the greater weight put on it, because illusion occurs incomparably more frequently than

does hallucination. The Argument from Illusion is, I believe, entirely successful: the most

productive (though not uniformly supportive) recent discussions of it in the literature are those

of Valberg, Smith, and Robinson. The Causal Argument is also successful, but it presents an

unexpected challenge that reinforces Berkeley’s and Foster’s theistic brand of Idealism. Then

there is a great variety of arguments for sense-data presented by O’Shaughnessy; though I have

agreed with their conclusion, I have not discussed any of the arguments themselves; but I think

it is safe to say that the probability of some of them being sound is high. Finally, there are

exceedingly simple arguments that can be thought of as basically depending on just one fact

about perception.

6.3.1. The Simplest Arguments

For example, the mere fact that hallucinations occur shows the existence of sense-data: what

else can a hallucinated pink rat be if not what has been called a “wild” sense-datum? Then

there’s an argument from the perception of color that the decidedly anti-Idealist philosopher

Peter van Inwagen introduces almost surreptitiously in the course of critically discussing

Berkeley.286 He plausibly attributes to Berkeley the assumption that “[t]he sensible properties of

common objects are the same things as the sensations we have when we see or touch (or hear or

taste or smell) common objects” (where by “common objects” van Inwagen means “ordinary

material ojects”). Now, says van Inwagen, when “I look at a blank sheet of white paper . . . I

have a certain sensation—a sensation of whiteness. And the paper, everyone agrees, is white, or

286 Peter van Inwagen, Metaphysics, 52.-54.

179

has the property of whiteness. The word “whiteness” occurs both in my description of my

sensation and in my description of the piece of paper. It seems, therefore, to designate

something that is both a property of the paper and is present in my mind.” This, says van

Inwagen, is what Berkeley thinks. He goes on to ask:

But is this right? Isn’t it far more plausible to suppose that the word “whiteness”

designates two different things when we are talking about my sensation and when we are

talking about the piece of paper? When we use the words “white” and “whiteness” to

talk about sensations, we are talking about something that can exist only as a part of

some person’s conscious experience. When we use these words to talk about a piece of

paper, however, we are using them to talk about a feature of the piece of paper, a feature

of the paper that it would have even if it were locked up in a dark room. . . . The feature

or property of the paper that we call “whiteness” does not vanish from the paper when no

one is experiencing the paper; the paper can lose this feature only if it is dyed or

undergoes some other treatment that we—of course—describe as “changing the paper’s

color.”

A bit later van Inwagen describes this physical property of the paper thus:

We experience the property whiteness by having sensations of whiteness (provided that

these sensations are caused by the presence of a white object and are caused on the

normal way). We could say that these sensations represent to us the presence of an object

that has the property. And we can easily form a conception of this property: it is the

property of common objects that (in appropriate circumstances) causes observers of the

objects to experience sensations of whiteness.

And just before this passage van Inwagen explained that “[a] person experiences sensations of

whiteness when that person looks at an object that has the property of whiteness—at least if the

person’s eyes, optic nerve, and brain are in good working order and if the object is being viewed

under normal conditions”.

It seems from all these passages that van Inwagen is subscribing to a Representative (Indirect

Realist) theory of perception, and he’s doing so because it offers what he thinks is a good

account of color perception. According to such an account, in any veridical perception,

sensations (sense-data) of whiteness “represent to us the presence of an object that has the

property” of whiteness; and more generally, sensations of any color represent to us the presence

of an object that has that color, if the perception is veridical. And if it’s not veridical, then the

sensations represent to us the presence of an object that as a matter of fact has a different color

from the color of the sensations that we are experiencing. the wrong color, but the sensations we

experience are of that color.

180

This is not only a Representative Realist theory of perception, but also a dispositional

account of physical color. But the issue between Berkeley and van Inwagen isn’t whether

sensations and sensed properties are or aren’t the same. Or rather, that’s one issue; but there’s

another issue as van Inwagen has framed it here, namely, whether the two occurrences of

“whiteness” (in “the whiteness the sensation is of” and in “the whiteness of the paper seen”)

designate the same item. On this issue van Inwagen says “no”, whereas Berkeley and I say

“yes.”

My argument is as follows. The whiteness that the sensation is “of” is the whiteness of the

sensed paper. But no sensations are such that there is no property which they are “of.” So

whenever a sensation occurs, there is a property it is “of” that gets instantiated. Therefore, if

there is no such property instantiated in the external world, then it must be instantiated in some

person’s conscious experience (i.e., his mind). But every property has a bearer. Therefore, a

property that is instantiated in the mind has a mental bearer, i.e., a sense-datum bearing that

property.

Thus a dispositional account of color perception entails the existence of sense-data provided

only that sensations of color sometimes occur.

6.3.2. Valberg on the Argument from Illusion

According to Valberg, there are experiential facts (facts that presuppose experience), to be

“contrasted with facts of existence, at least where the existence of external objects is

concerned.”287 To go back to his favorite example of a physical object, a book in front of him on

the table: “The book can be present only from within experience; but the book exists on its own,

whether or not it is present. The fact of the book’s existence is not an experiential fact.”288

Experiential facts come in three sorts. First, there are facts of presence: “it is only from

within experience that something can be present, demonstratively available.”289 Second, there

are facts of appearance: “it is only from within experience that it can look, or appear, as if

something is present.”290 Third, there facts of manifestation, to be distinguished from facts of

appearance. The way an object looks, or appears, to us is its way of being manifest to us. In the

classical case of the stick immersed in water looking bent,

the fact that the stick looks bent is a fact of manifestation; the fact that it looks as if a

bent stick is present is a fact of appearance. The fact of manifestation, but not the fact of

287 Valberg, The Puzzle of Experience,” 77-78. 288 ibidem, 78. 289 ibidem, 77 290 ibidem, 77-78.

181

appearance, entails that a stick is present. But neither fact could obtain unless something

is present.

An object cannot appear (be manifest) one way or another without being present. But,

equally, an object cannot be present without appearing one way or another. Visually (if

this is relevant), the object must look one way or another. And, as we know, the way an

object looks need not be the way it is. In the bent-stick example, the fact of appearance

(versus manifestation) is that it looks as if something bent is present.

A fact of appearance […] must be grounded on a fact of presence. In the example

under consideration, it looks as if a stick is present because a stick is present. But [this]

ground of the fact of appearance has another component, namely, the fact of

manifestation: that the stick looks bent (that it is manifest in a bent way). If the stick did

not look bent, it would not appear that something bent is present in my experience. When

the stick is removed from the water, it looks straight. Now the relevant fact of

appearance is that it looks as if something straight is present. And, as before, the ground

of this fact has two components: the fact of presence (the presence of the stick) and the

fact of manifestation (that the stick looks straight—which it is).291

Now Valberg says that “[o]ne reason for going into these matters” is that he wants to

disassociate himself from “a certain familiar argument for sense-data,” the so-called Argument

from Illusion. This argument

starts from the fact objects can appear other than they are. The stick, say, looks bent.

Now here an appeal is made to a principle which sounds rather like our point about

grounding. Nohing, it will be said, can look F unless there is present in experience

something which is F. In the bent-stick case, then, there must be something which is bent

present in my experience. Since the bent object is not the stick, it must be an internal

object (a sense-datum).

This argument rests on a confusion between facts of appearance and facts of

manifestation. If in the bent-stick case the object present is not the stick but a sense-

datum, then it cannot be true (as we are supposing) that the stick looks bent; for the stick

cannot look bent unless it is present. The argument treats a fact of manifestation (that the

stick looks bent) as if it were a fact of appearance. Thus it seeks to ground the fact of

manifestation on a fact of presence. This is a mistake. If we try to ground a fact of

manifestation on a fact of presence, we undermine its status as a fact of manifestation.

The correct view, I think, is that a fact of manifestation is the ground, or part of the

291 ibidem, 78-79.

182

ground, of a fact of appearance; but a fact of manifestation itself has no ground. In this

respect it is like a fact of presence, ungrounded.292

This is what Valberg says, and most of it seems just right. But there are a couple of

contentious points.

First, why does Valberg think a fact of manifestation has, or perhaps even needs, no ground?

Admittedly, let a fact of presence be ungounded; it’s just a brute fact. But why can’t we say, in

the case of the stick taken out of the water, that the fact of its manifestation (that it looks

straight) is grounded on the fact that it is straight?

Secondly, in the case of the straight stick looking bent when immersed in water, why can’t

we say the fact of its manifestation (that it looks bent) is grounded on the fact it presents a bent

sense-datum? Why can’t it be the case that, in this instance, there are two connected objects

present: the stick that looks bent and the sense-datum that really is bent? The idea here would be

that the stick is a whole of which the sense-datum is a part—so that in seeing the latter you’re

seeing the former. In other words, just as you can see a man by seeing his back, hand, or face, so

you can see a stick by seeing some of its sense-data.

Someone might reply that this idea illegitimately conflates two senses of “part” in play here:

sense-data are “logical parts” of sticks and stones; whereas backs, hands, and faces are “physical

parts” of human beings. But even if there are these two senses (of course there are), why should

this make any difference here?

For example, John Foster in The Immaterial Self discusses the dualist option that each human

being is a logical creation of two “parts” or “natures”: a soul and a body. As Foster puts one

version of this view,

each person is a unitary entity whose existence is wholly constituted by the existence of

a non-physical basic subject, the existence of a purely corporeal object, and the

attachment-relationship between them, and . . . this unitary entity takes on the twin

natures of the entities which feature in its constitution.293

Views of this kind once dominated the Western philosophical scene; some sectors of the latter

still take them seriously; and I see little in the way of fatal objections to them.

Now, in the bent-looking stick case, we also have, in a certain way, more than one thing

present. We have the stick, and we have the many sense-data that constitute it, one of which is

the sense-datum as of a bent stick. But in another way, we just have the stick. The stick is a

unitary entity whose existence is wholly constituted by the existence of many actual and

292 ibidem, 79. 293 Foster, The Immaterial Self, 238-239.

183

potential sense-data, some of which are sense-data as of a straight stick, while others are sense-

data as of a bent stick.

This is an Idealist view, and nothing that Valberg says in this context seems to rule it out.

In a footnote to Valberg’s last sentence above, after quoting the following passage from C. D.

Broad’s Scientific Thought, p. 240:

When I look at a penny from the side I am certainly aware of something; and it is

certainly plausible to hold that this something is elliptical in the same plain sense in

which a suitable bent piece of wire, looked at from straight above, is elliptical. If, in fact,

nothing elliptical is before my mind, it is very hard to understand why the penny should

seem elliptical rather than of any other shape,

Valberg comments:

The last sentence of this quote should make us pause, since it employs something very

much like our point about grounding. But it is a misapplication of the point. To say that

the penny looks elliptical is to say how it, the penny, appears (is manifest) in our

experience. It is to express a fact of manifestation. If we attempt to ground this fact on

the presence of an internal object, we treat it in a manner appropriate to a fact of

appearance; and, by inserting the internal object, we lose the fact with which we

began.”294

But why, by “inserting the internal object,” should “we lose the fact with which we began”?

What is this fact? It’s the fact that the penny looks elliptical. We shouldn’t lose this fact, and we

don’t lose it, if we ground it not only on the presence of an elliptical internal object (sense-

datum), but also on the presence of the round but elliptical-looking penny. In other words, why

can’t we have at least two objects present here: both the penny that merely looks elliptical and

the penny-related sense-datum that actually is elliptical? If it is replied that there’s just one

penny-like object present in this example, just as there was, in a way, one stick-like thing in the

previous one, the answer would again be: yes, in a certain sense, there’s just one thing present in

each case; but what this shows is that the penny and the stick is each in some way composed of

the sense-data; hence, in another sense, there are, in both cases, multiple objects present.

This, to repeat, is an Idealist position, arrived at by the Argument from Illusion.

6.3.3. Smith on the Argument from Illusion

In the recent philosophical literature it is David Smith who has concerned himself most

extensively with the Argument from Illusion, first setting it up as carefully as possible and

294 Valberg, 100, Note 3.

184

defending it, then attempting to refute it.

Smith expounds the argument in four steps. The first just is “the premise that perceptual

illusion can occur,”295 that, putting it more explicitly,

there is no type of physical feature that may not appear differently from the way it really

ease to any sense that could possibly perceive it. The next step . . . is an inference from

this: that whenever something perceptually appears to have a feature when it actually

does not, we are aware of something that does actually possess that feature. So, if you

are looking at a white wall, which because of the illumination looks yellow to you, you

are aware of something yellow. This inference is commonly known as the “sense-datum

inference,” with the immediate object of awareness that the inference introduces termed

a “sense-datum.” This is the heart of the [a]rgument. . . .

The third step in the [a]rgument consists in pointing out that since the appearing

physical object does not possess that feature which, according to the previous step, we

are imnmediately aware of in the illusory situation, it is not the physical object of which

we are aware in such a situation; or, at least, we are not aware of it in the direct,

unmediated way in which we are aware of whatever it is that possesses the appearing

feature—that direct way in which we formerly took ourselves to be generally aware of

normal physical objects. In the previous example, since the wall is white, not yellow, but

what we are immediately aware of is yellow, not white, what we are immediately aware

of cannot be the wall. . . .

The conclusion of the [a]rgument thus far is that in no illusory situation are we

directly aware of the physical object that, as we should initially have put it, “appears” to

us other than it is. The final step in the [a]rgument is what we may call the “generalizing

step”: we are immediately aware of sense-data, and only at best indirectly aware of

normal physical objects, in all perceptual situations, veridical as well as illusory.296

The reason for taking this final step

is the subjective indiscernibility of veridical and possible illusory situations. To put it

crudely, being aware of a sense-datum is exactly like perceiving a normal object. But a

sense-datum, whatever ot may turn out precisely to be, is clearly a radically different

type of thing from a normal physical object—at least as the latter are usually (that is,

realistically) conceived. So how could awareness of two such radically different types of

object be experientially identical? How could we mistake one for the other?297

295 Smith, The Problem of Perception, 23. 296 ibidem, 25-26. 297 ibidem, 26.

185

Here Smith refers to a point made also by H. H. Price “that these two radically different sorts

of perceptual object may form, subjectively, a smoothly connected series: that there could, as

Price put it, be a “sensibly continuous transition” from a state of veridical perception to an

illusory one, and conversely.”298

Finally, Smith emphasizes that the Argument from Illusion does not require that illusions

actually happen (though, of course, they do!), but only that they “be barely possible.” From such

“a mere possibility” the argument then proceeds (to repeat) as follows:

It is of the very nature of illusion, as opposed to hallucination, that were we to perceive

something illusorily, we should be aware of the same kind of object as we could be

aware of veridically; by the sense-datum inference, we should be aware of a sense-datum

in the possible illusory situation; so such a sense-datum, not a normal physical object, is

the kind of object we are aware of when perceiving veridically.299

Smith then criticizes a number of popular ways of disposing of the Argument from Illusion.

These need not be surveyed here except to say that Smith fully supports his claim that “the only

way to block” this Argument “is subtle and non-obvious” and that the “accounts of perception

that are generally purveyed today” (i.e., most of those published before the appearance of

Smith’s book) “simply do not face up to the Argument’s puzzling power.”300

This would be a ringing and unqualified endorsement of the Argument from Illusion—were

it not for the fact that Smith thinks he has found a way to block it. Thus the Argument, in his

view, can be resisted and Non-Idealist Direct Realism can be sustained. The way to proceed is to

put the standard accounts of perceptual consciousness behind us, along with aprioristic

assumptions about what such consciousness “must” be like, and turn, as Husserl put it, to

“the things themselves.” What we need is a careful phenomenological appreciation of the

lived character of perceptual experience. . . . [I]f Direct Realism is true we should be

able to discern a non-sensuous aspect to such experience, though one that is not a matter

of the exercise of concepts. Let us, therefore, take a fresh, unbiased look at perceptual

consciousness, to see if we can discern anything that makes it, simply qua mode of

sensory consciousness, distinct from sensation.301

In fact Smith thinks he can discern three basic phenomena that suffice for perceptual

consciousness and that distinguish it from the mere having of sensations. To put it briefly, the

first is the “spatiality” of perceptual consciousness, understood as “the phenomenal, three-

298 ibidem, 26-27. 299 ibidem, 29. 300 ibidem, 21. 301 ibidem, 133.

186

dimensional locatedness of the objects of awareness in relation to a sense-organ.”302 The second

is kinetic structure, “the way in which perception is integrated with movement—specifically,

movement on the part of the perceiving subject . . . the movement of sense-organs in relation to

perceived objects . . . the bodily movements . . . by which we come to enjoy different

perspectives on perceptible objects.”303 And the third is what borrowing a term of Fichte’s, he

calls the Anstoss: that’s the phenomenon “of a check or impediment to our active movement: an

experienced obstacle to our animal striving, as when we push or pull against things.”304

Smith then says that the first two basic phenomena of perceptual consciousness—

phenomenal three-dimensionality and kinetic structure—can “be comprehended by a single

account,”305 on which they can be seen as achieving “an identical phenomenon—one that alone

can confer perceptuality upon sensuous modes of awareness.”306 This phenomenon is what

psychologists term “perceptual constancy” and Smith sometimes calls “phenomenological

constancy.”

Hereby Smith deems himself to have located

a feature of perceptual experience . . . that is both sufficiently basic in all conscious

animal perception, and that will allow us to distinguish between an object of perceptual

experience and either sensations or sense-data. In relation to sensuously presentational

perception, this is to be found in perceptual constancy, elsewhere in the Anstoss, and

nowhere else. Here and here alone do we find within sense-experience itself the

phenomenological independence of object from subject that is the hallmark of perceptual

consciousness. Here we find the distinctive intentionality of perception: one that is

distinct from, and indeed more basic than, that to be found in thinking.307

It is from such “merely phenomenological observations” that Smith endeavors to construct an

original and needed refutation of the Argument from Illusion—needed in his view because in its

absence nothing else will block this argument.

6.3.4. Robinson’s Response to Smith on the Argument from Illusion

302 ibidem, 133-134. 303 ibidem, 140-141. 304 ibidem, 153. 305 ibidem, 170. 306 ibidem. 307 ibidem, 186.

187

It is just at this juncture that Robinson jumps in to defend the Argument against Smith’s claim

that it doesn’t actually prove that we fail to be perceptually aware of ordinary physical

(“normal”) objects. Robinson cites (part of) Smith’s complaint that

[t]he only proof that is offered . . . proceeds by pointing out that the sensory qualities that

feature in illusory perceptual experience are inherent in sense-experience itself. It forces

us to recognize sensory qualia. This, however, does not constitute the proof; it is but one

stage of it. What has to be shown in addition is that we are directly aware of these

sensory qualities—or, rather, of whatever it is that possesses them—as objects. Only

then will the normal object be edged out of its presumed position as object of immediate

awareness. Only then will “transcendence” have been lost. So all we have to show, in

order to block the Argument, is that we are not directly aware of whatever it is that

possesses such qualities, so that awareness of the latter does not cognitively mediate our

awareness of normal physical objects. Showing that such bearers are sense-data would

carry the Argument through; but that, I have suggested, is what has not been, and cannot

be, demonstrated.308

According to Robinson, Smith has made “an illicit move” here. It is

the move from demanding that the quality be an object of awareness, to demanding that

we be aware of what it is that posseses the quality, not just of the quality itself, and

treating these as equivalent. That this is a mistake can be seen as follows. When

something looks other than it is, we are usually aware of how it looks, so we are directly

aware of the qualities involved in illusion. Because this is how the physical object looks

to be, there is a sense in which we are not directly aware of what actually possesses these

qualities, insofar as that is not the physical object. We are not, that is, directly aware of

what it is that possesses these qualities. But this does not matter, for it is enough for the

proponent of the Argument to show (i) we are aware of the qualities presented in

illusions as objects of awareness, and (ii) these are in fact possessed by something other

than the physical objects. It is not necessary that (iii) we are directly aware of them as

possessed by something other than the physical objects. So once it is conceded that we

are aware of the qualities presented in illusions, and that they are in fact possessed by

something, which must be something other than the physical object, then it follows that

we are aware of that thing.

The difficulty of Smith’s position becomes especially clear if combined with his

account of secondary qualities. Smith is a subjectivist about such qualities. Such

308 ibidem, 186-187.

188

qualities have the same status as the various qualities of bodily sensations. “So there is as

much chance of a physical object being sensuously red as there is of a pin that is stuck

into my body possessing the same quality as my resulting pain.”

But it can hardly be doubted that sensuous red is an object of awareness. It follows

that, in a de re sense, it is the thing that exhibits the quality of sensuous red—which is

not the physical object—that is the object of awareness.309

Robinson also criticizes Smith’s appeal to perceptual constancy. Admitting the attractiveness

of this idea, Robinson agrees “that there is a sense in which, when I see a white wall as white,

ignoring the various patternings of shadow, . . . then I am not aware of the ‘deviant’ qualia, only

of the standard property.”310 But there’s also a sense in which one is “aware of the varieties of

shadows,” only “one does not take notice of them in certain ways.”311 More importantly,

many cases of illusion are not subject to constancy. When I take off my glasses, things

become blurred and very noticeably so: the blurred shapes, not the actual properties of

any physical objects, become the objects of my awareness. Nor do I spontaneously

ignore the bentness of the oar in water.312

Robinson also briefly discusses Smith’s notion of the Anstoss, which he calls “interesting and

difficult,”313 but insufficient for a defense of direct realism because what the Anstoss does is

frustrate our wills, and that “alone could not create more than a very abstract picture of the

world.”314 I think Robinson is correct here and that rather than helping Direct Realism the

Anstoss helps us see the absurdity of a certain kind of scepticism.

I am, therefore, inclined to agree with Robinson verdict: “. . . it seems to me that Smith fails

to refute the argument, in the very strong form in which he states it.”

That the Argument from Illusion is indeed irrefutable seems also clear from Foster’s

discussion of it in his The Nature of Perception. I will not rehearse his arguments here, just state

that the only criticism of Foster’s discussion that I have come across in the literature is Robert

Hopkins’s. And that criticism is totally unconvincing.315

309 Robinson, Review of Smith’s The Problem of Perception, 522-523. 310 ibidem, 523. 311 ibidem. 312 ibidem. 313 ibidem, 524. 314 ibidem. 315 Hopkins asserts that Foster’s “case against direct realism fails to distinguish between imprecision and inaccuracy in how experience represents the world. Once that distinction is made, I think it is clear that the case really depends on the claim that [Direct Realism] cannot allow for imprecision. But if this is so, the argument relying on that claim is otiose. Perceptual content is never wholly precise: there is always a point beyond which, although the object is characterized in one way or another, our experience of it fails to commit itself on which.” But Hopkins fails to distinguish between slight imprecision and radical distortion; and it is this difference—and the continuity between precision, slight imprecision, and radical distortion—on which Foster’s argument turns, and not on any claim that direct realism cannot allow for imprecision.

189

6.3.5. The Causal Argument

The Causal Argument for Sense-Data has recently been used by both Robinson and Foster. I

will state it as it occurs in Robinson’s Perception:

It is clearly true that:

1. It is theoretically possible by activating some brain process which is

involved in a particular type of perception to cause an hallucination

which exactly resembles that perception in its subjective character.

2. It is necessary to give the same account of both hallucinating and

perceptual experience when they have the same neural cause. Thus, it

is not, for example, plausible to say that the hallucinatory experience

involves a mental image or sense-datum, but that the perception does

not, if the two have the same proximate—that is, neural—cause.

These two propositions together entail that perceptual processes in the brain produce

some object of awareness which cannot be identified with any feature of the external

world—that is, they produce a sense-datum.316

This argument has been attacked, mainly but not exclusively by attempting to discredit the

premise (2). In his book, Robinson refutes these attacks, including that stemming from the so-

called disjunctive analysis, masterfully. Therefore, I would unhesitatingly conclude, in concert

with Robinson, that the Causal Argument is successful, were it not for the fact Foster, who once

approved of it too, later came to reject it, for reasons that seem to me to have considerable force.

As a result, it is not easy to make up one’s mind about the Causal Argument, for the situation

is indeed more complicated than at first it might seem, and this in two respects. The first respect

is the following:

On the one hand, this argument is intuitively appealing; there is very much to be said for it;

and both Robinson and (the earlier) Foster have said it well. On the other hand, there also is a

good case against it, which is made all the stronger by its dovetailing, as we’ll see presently,

with my (and Foster’s, and Berkeley’s) ultimate thesis that God (or an equipotent agent)

guarantees the veracity of our perceptions. Let me, therefore, turn directly to Foster’s recent

criticism of the argument (again, though his discussion is lengthy and minute, mine will be as

compact as possible).

It is admitted by all sides that the Causal Argument depends heavily on, and draws much of

its strength from, the principle of same type of proximate cause, same type of immediate effect.

316 Robinson, Perception, 151.

190

Thus it is possible for me to either see the sun, or to hallucinate it. In both cases the proximate

cause (neural event) is the same; but in the first case the immediate effect is my seeing the sun

(the remote cause), and in the second case the immediate effect is my hallucinating the sun by

seeing a sun-like sense-datum (where the remote cause is an evil demon, a cunning scientist, or

whomever or whatever else caused the hallucination). The intervening process (what happens

between the remote cause and the proximate cause) does not matter.

Now Foster thinks that there is a logical and convincing way both to preserve this principle

and to reject the argument. All the Direct Realist has to do is to say that

although the remote cause does not stamp its character on the intervening process, it

causally affects the outcome directly. In other words, he can claim that, at the point when

the [perception or hallucination] is about to occur, the factors which directly contribute,

causally, to its occurrence and character include not just the current state of the brain, but

also certain aspects of the preceding causal process, including, crucially, certain aspects

of the way in which the neural process leading up to the realization of the brain state has

itself been brought about. This would mean that there was, after all, no violation of the

principle same type of proximate cause, same type of immediate effect, since what we

have hitherto been describing as the remote cause would become directly relevant to the

outcome, and therefore an element of the proximate cause. More precisely, the process

which leads up to the neural process would play a double causal role. It would still

continue to play, in the straightforward way, the role of what brings about the neural

process, and, on that score, would indeed count as the remote cause with respect to the

psychological outcome. But it would also, as part of the whole physical process leading

up to the even in the brain, combine with this event to exert a direct influence on the

psychological outcome—in particular, to fix it as something perceptive or as something

hallucinatory—and, on that score, would count as part of the proximate cause.317

This is, admittedly, a bizarre-seeming explanation “of why the nature of the psychological

outcome varies, systematically, according to the way in which the neural process is brought

about,” says Foster. For

[e]verything else we know about the world suggests that causation does not work in this

sort of way: it always works in a way which is not only temporally directed (from earlier

to later), but temporally continuous, so that earlier events only have an influence on non-

contiguous later events by affecting the chain of events that intervene. The idea that

earlier events may have a direct influence on what happens after a temporal interval—an

317 Foster, The Nature of Perception, 35-36.

191

interval in which any record of the relevant features of those earlier events have been

lost—seems very strange, and perhaps hardly credible.318

It is the sense that this postulated causal mechanism is extremely unusual that makes the Causal

Argument seem natural and plausible. But, cautions Foster, in the present case there are two

good reasons why “our ordinary modes of causal explanation may fail to apply.”319

One is that, while our usual paradigms of causation “are drawn exclusively from the physical

realm: they concern the ways in which one physical event or set of conditions brings about

another,” what it is that we are seeking in this case is an account that “concerns the relationship

between the physical and the mental—between the types of physical process leading up to

perception and hallucination and the nature of the psychological outcome itself.” And we cannot

just assume without further ado “that forms of causality which are operative in the physical

realm will be operative in the psychophysical realm as well.”320

The second, and “more crucial,” reason is that in “standard cases, when some physical object

is caused to come into a certain state . . . this state does not consist in, or inherently include, a

relationship to something earlier. . . . But in the case of perception,” as the Direct Realist

“construes it, the situation is quite different.

The subject is caused to come into a psychological state which . . . is inherently a

relationship with an earlier item—a state which, on its own, suffices to put the subject

into perceptual contact with a physical event or object-stage at an earlier time. And this

does not just mean that the relevant state is one which cannot be realized without putting

the subject into contact with some earlier physical item. It means that there is a particular

earlier item of which the state is, in itself, perceptive.321

Now what the champion of the Causal Argument fails to realize, according to Foster, is that

“the sort of situation which arises with respect to the distinction between perception and

hallucination also arises, routinely, with respect to the distinction between different

perceptions.”322 Thus, to take an example of my own, if I am inside a giant circular enclosure

and am surrounded on all sides by a huge, uniformly white wall, then, if I look at the wall and

keep my gaze steady, I will be seeing that portion of it from which light is streaming into my

eyes. If a second later I move my head (and eyes) a little bit to the right but keep on looking, I

will see an adjacent, numerically different but qualitatively indistinguishable portion—because

light is now streaming into my eyes from it, the different portion of the uniformly white wall

318 ibidem, 37. 319 ibidem. 320 ibidem, 37-38 321 ibidem, 38. 322 ibidem, 43.

192

I’m looking at now (and not from the previous portion I was looking at a second ago). If I

change the direction of my gaze yet again but keep on looking, I will see yet another portion of

white wall: this portion is again reflecting light to my eyes though this portion is numerically

different from the first two. Let me call the three portions of uniformly white wall I’ve

mentioned Portions 1, 2, and 3.

The same goes mutatis mutandis for all cases of ordinary seeing. In all cases the identity of

what I see is causally determined, in part, by the identity of the item that reflects or emits the

light that is likewise causally responsible, in part, for what I see. As Foster puts it, “even where

the causal process from item to brain are qualitatively identical on [three] occasions,” as they are

in the examples I just gave, “the psychological outcomes [perceptions of Portions 1, 2, and 3]

will be qualitatively different, simply because the [three] psychological states involved will be

inherently perceptive of different things [here, three different portions of the white wall].”323 He

then draws the following conclusion:

If the Causal Argument had been initially viewed in this broader perspective, it would

never have seemed compelling. For, as it occurs in the perceptual context, the situation

does not create even a prima facie problem for the theorist. Rather, it serves to make

clear the sort of causal account that his distinctive understanding of perception

requires—an account which sees the whole causal process, from item to brain, as directly

responsible for the nature of the perceptual outcome, and, in particular, sees the identity

of the initiating item as causally fixing the identity of the perceptual target. This is an

account which the [Direct Realist] can accept, without embarrassment, prior to any issue

over the treatment of hallucination, and can then suitably deploy, to dispose of the

Causal Argument, once this issue has been raised.324

Yet it is just here that the other, and absolutely essential, complication arises. Above I quoted

Foster as stating that, according to the Direct Realist, in virtually all ordinary perception, “[t]he

subject is caused to come into a psychological state which . . . is inherently a relationship with

an earlier item—a state which, on its own, suffices to put the subject into perceptual contact

with a physical event or object-stage at an earlier time.” In an e-mail from 2003, citing this

passage, I asked Foster:

However, can any subject, at time t, be put into direct, unmediated perceptual—as

opposed to conceptual or recollective—contact with an object-stage that, at t, no longer

exists? Isn’t it just obvious that no human subject can directly perceive the past, as

opposed to thinking about it or remembering it? In other words, isn’t the direct realist

323 ibidem. 324 ibidem.

193

immediately refuted by the TGA—an argument according to which the subject perceives

something caused by, and (in some respects) qualitatively similar to, but not identical

with, the physical object in question?

I then added that I agreed with Foster’s point (which I quoted above and now repeat) that the

Direct Realist requires “an account which sees the whole causal process, from item to brain, as

directly responsible for the nature of the perceptual outcome, and, in particular, sees the identity

of the initiating item as causally fixing the identity of the perceptual target.”

But then, quoting Foster’s further remark that “[t]his is an account which the [Direct Realist]

can accept, without embarrassment, prior to any issue over the treatment of hallucination, and

can then suitably deploy, to dispose of the Causal Argument, once the issue has been raised,” I

asked Foster:

But isn't the time-gap argument precisely the embarrassment this account cannot

overcome?

To this he replied:

You may be right. Certainly many philosophers would share your intuition that the

notion of direct perceptual awareness of something in the past makes no sense. I don’t

share that intuition myself, but I guess the issue is one I should have discussed.

He never did discuss it in anything that he wrote later (including his last book, A World For Us).

However, what I wish to record here is that Foster did admit that the TGA may pose a real

difficulty for the defense that he gave, on the (Non-Idealist) Direct Realist’s behalf, against the

Causal Argument.

I conclude that because the TGA stands unrefuted it really does pose this difficulty and that

the Direct Realist is not saved from the Causal Argument by Foster’s defense against it.

Therefore, the (Non-Idealist) Direct Realist is refuted both by the intact Causal Argument and

by the TGA, the latter saving the former from Foster’s attempt to weaken it. And, of course, the

Causal Argument, like the Argument from Illusion discussed earlier, can be of service to the

TGA by helping it prove that even in minuscule time-gaps we immediately perceive not

physical objects, but sense-data (as discussed in 6.2.).

On the other hand, there is one point on which both the TGA (or, more exactly, its Berkeleian

version) and (part of) Foster’s defense against the Causal Argument agree. To see this, we

should recall that neither the TGA nor the Causal Argument shows that we do not perceive

physical objects; what each shows is that we do directly perceive sense-data and that we may, in

addition, perceive physical objects as well. But in the latter case, if we do perceive physical

objects, then their perception obeys certain strictures—those imposed both by the Berkeleian

TGA and Foster’s defense against the Causal Argument.

194

It might be thought that the defense is only deployed on behalf of the Direct Realist, with

Foster showing him how he can disarm the Causal Argument. But we must remember that

Direct Realism comes in two varieties: the much more common Non-Idealist kind and the

Idealist one, which, following in the footsteps of both Berkeley and Foster, we are defending

here. Now I’ve shown that Foster’s defense against the Causal Argument doesn’t work for the

Non-Idealist Direct Realist: he is thoroughly refuted by that argument. As for the Idealist Direct

Realist, he doesn’t need this defense against it: after all, the Causal Argument is one of the

arguments he can happily use against his non-Idealist fellow. But Foster’s defense has one

element that all Direct Realists, whether Idealist or not, can and must agree on: it’s the claim

that “the perceived item”—any perceived item—plays “a direct causal role in the targeting of

the perception onto it.” Let us call this the Perceptual Target Assumption (PTA). How it works

and why it’s justified I will explain more fully in the next section.

For now, let’s just note that since the Causal Argument is sound, what is perceived in all

cases is a sense-datum. This may be a logical part of a physical object, which, if it still exists,

will be perceived as well, although if it does not, then the sense-datum will be a trace of the no-

longer-existent object. Or it may be a “wild” sense-datum, in which case it is part of a

hallucination. But in all cases what sense-datum is seen causally depends on how and by whom

it was produced.

This, the PTA, is what is correct in Foster’s defense. What is incorrect (and, to repeat, is

shown to be so by the TGA) is the Non-Idealist Direct Realist’s claim (seconded in Foster’s

defense) that the perceiving “subject is caused to come into a psychological state which . . . is

inherently a relationship with an earlier item—a state which, on its own, suffices to put the

subject into perceptual contact with a physical event or object-stage at an earlier time.”

We’ve already established that this claim is false and that its negation, the simultaneity-

thesis, is true, but most recently additional support for the latter has come from an unexpected

quarter.

6.4. Seeing the Present

In most cases the perceptual time-gap is extremely small: that’s why the TGA can get started

only by considering examples of objects at astronomic distances from us. Yet its conclusion

holds for objects perceived at short distances as well. These are the close-by objects, motions

and changes in which matter to us practically every day. Many types of human activity, for

example, swatting a fly; catching a ball; ducking a blow; avoiding a swerving car; and, to take a

nowadays more “exotic” example, stepping into and out of an open moving lift (such as used to

195

be, and in some places still are, common in office-buildings throughout Europe)—all these

require the perceiver, as one evolutionary neurobiologist puts it, “to see the present.” Otherwise,

he or she will not succeed at such a (mundane) activity. The reason why most human beings are

good or excellent at most of these activities is that their perceptual systems have evolved to

perceive, not “events or object-stages at an earlier time,” as Foster claims, but events or object-

stages as they occur now.

That neurobiologist is Mark A. Changizi, previously a professor at Rensselaer Polytechnic

Institute, now director of 2AI Labs in Boise, Idaho. In his view, as expounded in his 2009 book

The Vision Revolution, “we evolved to see moving objects, not where they are, but where they

are going to be. Without this ability, we couldn’t catch a ball because the brain’s ability to

process visual information isn’t fast enough to allow us to put our hands in the right place to

intersect for a rapidly approaching baseball. “If our brains simply created a perception of the

way the world was at the time light hit the eye, then by the time that perception was elicited—

which takes about a tenth of a second for the brain to do—time would have marched on, and the

perception would be of the recent past,”325 Changizi explained in that book. In an interview with

David DiSalvo on the Neuronarrative site, he reiterated his theory as follows:

When light hits our retina, what our brains would like to do is instantaneously generate a

perception of what the world looks like. Alas, our brain can’t do this instantaneously.

Our brains are slow. It takes around a tenth of a second for your perception to be built,

and that’s a long time when you’re moving about. If you perceived the world the way it

was when light hit your eye, you’d be having a tenth-of-a-second old view of the world.

Because of this, visual systems have evolved mechanisms to try to generate a perception

not of the way the world was when light hit the eye, but generate a perception of the way

the world will be by the time the perception occurs in a tenth of a second. By the time the

perception is elicited, the anticipated future will have arisen, and the perception will be

of the present. That is, in order to perceive the present (have perceptions at time t that are

of the world at time t), our visual systems must anticipate the near-future.326

Now this quotation is just what my thesis needs, but it has to be amended a bit. Take the third-

to-last sentence. A “perception” of the way the world will be a tenth of a second later just can’t

be a perception in the strict sense of that word, since this forbids perceiving the future, i.e., what

is not there yet, just as it forbids, on the basis of the simultaneity-principle (which Changizi

accepts), perceiving the past, i.e., what was, but no longer is, there. So what the visual

325 http://itunes.apple.com/us/book/the-vision-revolution/id380289384?mt=11 326 http://neuronarrative.wordpress.com/2010/05/05/everything-we-knew-about-human-vision-is-wrong-author-mark-changizi-tells-us-why/

196

mechanisms try to generate—and actually elicit—are sense-data in the present, which sense-

data, since they aren’t “wild” but of a physical object, mediate the actual perception of the

object in the present.

When I suggested my emendation (with an argument for positing sense-data) to Changizi, he

responded that there was

something deep and nice about this argument. Usually I get by ignoring the time delay

between the event and the light hitting the retina, but what is the right way to think about

it when that delay is non-zero, whether a fraction of a second, eight minutes, or a billion

years? I'd prefer not to have to say one is perceiving sense-data, but I'll have to ponder

this.327

In the paper from which I’ll cite next, Changizi uses the terminology of a Representative Theory

of Perception and speaks of “visual percepts,” which term is sometimes used in the

philosophical literature a synonym for “sense-data.” Thus interpreting his views along sense-

data-ist lines seems natural and called-for, as is the remark I’ve italicized. I’m talking about a

2008 paper, entitled “The trade-off between speed and complexity” and published in Behavioral

and Brain Sciences. In it he stated that proponents, such as himself, of the hypothesis that

the brain has mechanisms for perceiving the present (i.e., mechanisms designed to

generate a perception at time t that is representative of the scene at time t) typically say

that the advantage is that it helps overcome inevitable neural delays. That is, “latency

compensation” and “perceiving the present” have gone hand in hand. The implicit

assumption can often seem to be that natural selection has attempted to minimize neural

delays – by shortening wires, speeding up signal propagation, and using rapid

computational algorithms for generating a visual percept – and whatever latency between

retina and perception is left is handed over to the compensation mechanisms to deal with.

Although this is an open possibility, the hypothesis that we perceive the present is not

committed to this possibility; it is only committed to the idea that perceptions belong to

the present [my italics – M. J. D]. What is left open is how long the delay is, and

whether it is all “inevitable” or whether the delay may be much longer than it would be if

selection for short processing times trumped all other selection pressures.328

Changizi then suggests considering computer software as an analogy.

Computer processing speed has risen by many orders of magnitude over the course of

the last 20 years, but you may have noticed that many of your programs still take

considerable time to start up. Computer designers know how long a wait we are willing

327 Personal communication. 328 Changizi, “The trade-off between speed and complexity,” 203.

197

to endure, and use that time to carry out fancier computations. That is, when faster

computers arrive, computer designers do not appear to be saying, “Now we can compute

the same old things nearly instantaneously!” Instead, they seem to be saying, “Now think

about how much more we can compute while the user waits!” Just as computer software

delay is a consequence of a trade-off between shorter delay and more complex

computations, our perceptual delay is a trade-off between shorter delay and fancier

visual computations. For example, if evolution can find a new clever trick for

extrapolating farther out into the future – say from 30 msec to 120 msec – then it could

utilize this trick and allow itself four times the amount of computation time to build

sophisticated useful perceptions. The resultant latency of 120 msec would not be

understood as an inevitable delay left over after trying to reduce it as much as possible.

Instead, it would be better to say that there is selection pressure to maximize the delay

for which the nervous system is able to compensate, thereby buying more time to prepare

the perception. Counterintuitively, then, it may well be that the slower-to-react brains are

the “smarter” ones. Visual prediction as indicated by perceptual adaptation to temporal

delays and discrete stimulation.329

Changizi’s hypothesis (that there is an evolutionary perceptual adaptation to temporal delays

whereby visual prediction kicks in so as to enable perceptions of the present) still awaits more

testing and confirmation, and it is, as far as I know, not yet universally accepted. On the other

hand, my cursory examination of the relevant literature suggests that so far it has not been

seriously challenged. In any case, if it were true, it would explain why in the vast majority of

ordinary situations phenomenal simultaneity is enough to enable us to execute motion-involving

tasks punctiliously. In matters our health and safety vitally depend on, we perceive (things in)

the present, even though the signals that condition our perceptions were emitted or reflected by

(momentarily) past stages of these things.

Of course, there is almost no chance that evolution will produce visual systems allowing us

to see stars as they are now; but that is most likely because far-away things are, in most cases,

much less important to our well-being. That’s why, if we don’t want to say that most if not all of

the things we see are in the past, we only have the choice of saying EITHER that we see present

sense-data but not physical things in the past OR that from these sense-data we get information

about what these past things were like, without actually seeing them in the past.

Incidentally, that we actually see all things, including stars, not as they were in the past, but

as they are now, was a thesis advanced by some Direct Realists early in the last century. One of

329 ibidem.

198

them was G. Dawes Hicks. His view, as I indicated in section 1.5. above, was presented in a

symposium called “The Time Difficulty in Realist Theories of Perception,” published in the

Proceedings of the Aristotelian Society 11 (1911-12), and roundly criticized by A. O. Lovejoy in

his Revolt Against Dualism (p. 76-85). The criticism was so convincing on both philosophical

and scientific grounds that, on this count, Hicks was, as far as I can tell, neither defended nor

emulated.

To return to the much more interesting Changizi hypothesis: it entails what in the last section

I called the Perceptual Target Assumption (PTA)—namely, that “the perceived item plays a

direct causal role in the targeting of the perception onto it.” This assumption says nothing about

the time of the perception targeted (by something or someone) in the circumstances involving

the perceived item. To this assumption Changizi only adds the claim that thanks to evolution

“we see the present,” thus yielding Changizi’s hypothesis. Through quotations from his texts I

explained how Changizi’s hypothesis—and by implication, the PTA—works. I conclude that

it’s probable that evolution has made the PTA true. But evolution has been enabled by God,

according to some modern theists (those who do not deny evolution). Foster was one of the

latter: no wonder he championed the PTA. Berkeley, too, no doubt, would have had a theistic

view of evolution, if he had heard of, and accepted, it. And he also doubtless would have

approved of the PTA if he had thought of it. Berkeley, of course, would have insisted that it is

God who, in the circumstances involving the perceived item, targets the perception thereof onto

that item, and does it in such a way that the perceived item (the idea or sense-datum) and the

perception thereof are simultaneous.

6.5. Conclusion about What We See

Before bringing this chapter and my whole dissertation to a close, let me briefly return to three

contemporary philosophers whom we have already discussed at great length. They accept the

simultaneity-requirement but stand in different relations to Berkeley’s philosophy. Two of them

are stumped by questions to which Berkeley’s idealism has an answer, and the third comes very

close to the truth about what we see.

Jerry Valberg, in his The Puzzle of Experience, presents an amended and highly simplified

TGA, which, he thinks, gives rise to an insoluble puzzle. Suppose that after the light-rays have

left an external object, God eliminates that object but lets the light-rays reach our eyes and set in

motion the usual cortical processes so that everything in our experience remains the same. What

we then see is not, of course, the external object, which has been eliminated, but a sense-datum,

an internal object, which exists only in so far as it is perceived. That’s what we are forced to

199

conclude philosophically because of the TGA, which is irrefutable and which exploits, most

fundamentally, the “potential irrelevance” of the external thing to our experience.

On the other hand, says Valberg, it’s just obvious that external objects do exist; one need

only be open to how things are in one’s experience. So what we have is an irresolvable

antinomy, which he calls “the puzzle of experience.”

But, I think, Berkeleian (Phenomenalistic) Idealism does solve this puzzle. For according to

it, it isn’t the object itself (except secondarily) that causes my experience of it through any

activity in my visual cortex, but God Himself who primarily and directly causes all external

objects and all our experiences of them under certain appropriate circumstances. Hence, external

things are actually (not just potentially) irrelevant to our experience. That’s exactly what the

Berkeleian TGA leading to Idealism shows; it shows that all perceivable things are internal, but

not that all things are internal: perceivers (including God) are all external with respect to each

other. Valberg, of course, doesn’t believe that Berkeleian Idealism can be proven; moreover, he

believes that even if it could be, “the puzzle will break out” within that Idealism itself; that’s

why his puzzle stays a puzzle, for him.

A bit later I will say more about the provability of Berkeleian Idealism, but for now I wish to

pinpoint exactly where I think Valberg goes wrong. First, he has it exactly right when he asserts

the following: “Phenomenalism begins with the idea that, strictly speaking, the object of

experience is always internal. So how could the conclusion of the problematic reasoning pose a

difficulty for Phenomenalism? The conclusion of the reasoning and Phenomenalism seem to be

made for each another.”330

But then, asks Valberg, how would the Phenomenalist distinguish (as he must, on his own

insistence) between my hallucinating a book, Macbeth-style, and my really seeing it? In such a

case, “I am uncertain whether the object I pick out is external (part of the world) or internal. The

Phenomenalist has his own way of explicating such uncertainty. . . . We might say that the

Phenomenalist distinguishes two kinds of pattern or sequence of internal objects: external object

patterns and internal object patterns. The Macbeth-type uncertainty can then be represented as

uncertainty about which kind of pattern, external or internal, this object belongs to.”331

“Now,” says Valberg, “my point is this. Once we have a Phenomenalistic interpretation of

the problematic reasoning, the conclusion will be not just that the object present in experience is

always internal (that goes without saying), but that the object present is always part of an

330 Valberg, The Puzzle of Experience, 163. 331 ibidem.

200

internal object pattern. And clearly, this conclusion is, for the Phenomenalist, just as impossible

to accept as it is for us . . . to accept that the object present in experience is always internal.”332

But what Valberg misses here is that for the theistic Phenomenalist the problematic reasoning

would not so much as arise. Let us hark back to 4.4. and recall what Valberg says in expaining

one of its steps: “God (it is handy to bring God in here) . . . might eliminate the object while

maintaining the activity in my visual cortex. Is that not a coherent state of affairs, something

God might bring about?” The theistic Phenomenalist, taking God utterly seriously (and not just

as a handy peg to hang a philosophical point on), says: No, God might not bring that about. He

(the theistic Phenomenalist) believes that God, being no deceiver, would never actualize this

logical possibility, never “eliminate the object while maintaining the activity in my visual

cortex,” unless perhaps He wished to teach us a moral lesson.333 Thus, if theistic Idealism

(Phenomenalism) is correct, the puzzle of experience is solved simply because the problematic

reasoning does not go through.

James Cornman holds that Berkeleian “phenomenalism is plausible enough to be one of the

leading contenders for the most reasonable theory of perception and the external world,” but he

thinks that it faces an objection from the time gap as revealed by the following argument:

“If Berkeleian phenomenalism is true, then someone, s, sees a star at time t, if and only if at t

he sesnses a sensum that is in the group of sensa which is identical with the star. But if at t

someone senses a sensum that is in a group of sensa which is identical with a star, then the

group and, therefore, the star exists at t. But just as some stars exist unseen, some stars are seen

at times they do not exist. Therefore Berkeleian phenomenalism is false. It seems that a

Berkeleian has at most two ways to counter this argument: he can either deny that some things

are seen when they do not exist, or deny that every sensum sensed when an external object is

perceived is in the group of sensa that is identical with the object.”334

This is absolutely right, and what our Berkeleian does is take the first alternative—he denies

that some things are seen when they do not exist. He says that all things are seen when and only

when they exist (if not by finite beings, then by God) and that what I see when I see something

in the case of the no-longer existent (because exploded) star is its light. More exactly, since

“star” is ambiguous as between “tangible star”, “visible star,” and “physical star”, what no

longer exist (and therefore can no longer be felt or otherwise observed) are the tangible star and

332 ibidem, 163-164. 333 But in that case, for the lesson to take hold, He presumably would have to let us know when we are hallucinating something and when the object seen is really there. This would involve (1) producing internal object patterns in the case of a visually hallucinated book (so that if I tried to pick it up my fingers wouldn’t grasp it or feel anything book-like) and (2) producing external object patterns in case the book was really there (so that I could pick it up and handle it). 334 Cornman, Perception, Common Sense, and Science, 353-354.

201

the physical star; but the visible star still exists; it is seen; and it is nothing but a speck of light.

That, by the way, is what we human beings always see in the way of physical stars, even when

they’re not exploded and still exist: in all cases we see just pinpricks of twinkling yellowish

light—“visible stars” that have been emitted or caused (“vulgarly” speaking) by the “physical

stars.”

And as Len Carrier has said, “this saves what is correct in Berkeley.”335 Len Carrier is

another Non-Idealist Direct Realist who accepts the simultaneity-requirement but, because of his

deep Australian Materialist convictions, cannot follow Berkeley all (or even much of) the way.

Still, by taking the TGA as proving that we always see light he plays directly into Berkeley’s

hands. He stops where he does because he thinks (a) light is physical and (b) physical things do

not depend on being perceived. But Berkeley would have no trouble accepting (a); he only

denies (b), and his arguments for that denial are familiar and controversial. In the imagined

Berkeleian TGA, some of them would come as support for premises (7), (8), (9), (11), (12),

(14), (15), (19), (20), and (21). I hope I have said enough to show that had Berkeley deployed

his arguments against (b) in the course of presenting and defending something like the TGA,

i.e., had he used TGA as a scaffold or framework for all his relevant arguments, his case for

Idealist Direct Realism would have been even stronger and more perspicuous. In any event, the

perceptual time-gap is an ace in the hole for any metaphysics in the vicinity of Berkeley.

But can Berkeleian Idealist Direct Realism be proven? One way would be by continuing on

the way I started in this work—by providing more support for all the insufficiently supported or

altogether unsupported premises of this Berkeleian TGA. Another way would be to follow

Foster. I think he has, in effect, already proven Idealism. In The Nature of Perception he has

shown, with arguments that I think are absolutely conclusive (as conclusive as any arguments

for an initially unobvious and complex conclusion can be) that Non-Idealist Direct Realism and

Indirect (Representative) Realism are false insofar as they cannot account for the fact that we do

see things and have perceptual knowledge of them. Therefore, since it is absurd to deny the

latter fact and since Non-Idealist Direct Realism, Indirect Realism, and Idealism are the only

options for explaining this fact, it follows that if the first two views are false, the third has got to

be true! This reasoning, I think, is perfectly cogent.

Of course, it says nothing about the specific nature of the Idealism that has been shown to be

true. Foster fills this lacuna as well, and also gives additional arguments for Idealism (basically

varieties of one specifically Fosterian argument) in the last part of The Nature of Perception and

in the whole of The Case for Idealism and of A World For Us. Shorter versions of these

335 Carrier, “The Time-Gap Argument,” 271.

202

arguments appear in several of the papers he left behind. In addition, there is the defense of

idealism in the last chapter of Robinson’s Perception, a defense prepared for in the book’s

earlier chapters, as well as in (some of) his papers and those of John Bolender. All in all, the

proposition that Idealism with respect to the material world is the right view has been, I believe,

adequately supported.

Problems for further investigation, of course, remain in droves. I mention just one. The

Berkeleian TGA I constructed contains the following premise:

(19) Then in the very act of immediately seeing the visible sun I am also perceiving,

whether immediately or mediately, the undestroyed sun of physics, which is a larger

collection of, or a more complex “fiction of our own brain” dependent on, ideas of which

collection or fiction the visible sun is a part.

How is that to be understood? In my dissertation, I do not explain this, but leave it as a task for

future investigation. Chapters 12 and especially 13 of Foster’s Case for Idealism provide the

(difficult) materials for one answer, to some extent distinct from the answer suggested by the

last three chapters of A World For Us. Then there is the completely different account of how

physical objects can be constructed out of sense-data given in Chapters Three (Berkeleyan

Phenomenalism) and Five (Theoretical Phenomenalism) of Cornman’s Perception, Common

Sense, and Science. None of these accounts is straightforward. Building on them, an Idealist

worth his salt would try as best they can to make it clear how physical objects are related to the

sense-data that consitute them.

Thus, to finally conclude, I hope I have given good reasons for my thesis that when we go

through the world with open eyes we always see present sense-data which much more often than

not make up physical objects. What we invariably and always see are sense-data of light; almost

always we also see other sense-data as well as the physical objects they constitute. That sense-

data exist is certainly proven by the Time-Gap Argument, but it is also proven by the Argument

from Illusion and the Causal Argument, and probably by other arguments (i.e.,

O’Shaughnessy’s) as well. Therefore, on the basis of the Time-Gap Argument and other

considerations there is much more to be said for Idealist Direct Realism than for either Non-

Idealist Direct Realism or Representative Realism. But both the Representational Theory of

Perception and the Adverbial Theory of Perception especially have little plausibility, though the

former is compatible with the Time-Gap Argument.

203

Bibliography:

Alston, W. P., Perceiving God: the epistemology of religious experience (Ithaca and London:

Cornell University Press, 1991): 21, 27-28, 37, 55-57.

Alston, W. P., “Back to the Theory of Appearing,” Philosophical Perspectives 13 Epistemology

(1999): 181-203.

Alston, W. P., “Perception and Representation,” Philosophy and Phenomenological Research

LXX (2005): 253-289.

Armstrong, D. M., Perception and the Physical World (New York: Humanities Press, 1961):

144-151.

Armstrong, D. M., Bodily Sensations (London: Routledge & Kegan Paul Ltd, New York:

Humanities Press, 1962): 4-8.

Armstrong, D. M., “In Defence of the Cognitivist Theory of Perception,” Pufendorf lecture

series (2004)Aune, Bruce, Knowledge, Mind, and Nature (New York: Random House, 1967): 4-

18, 63-83, 135-154.

Ayer, A. J., The Problem of Knowledge (Baltimore, Md.: Penguin Books, 1956): 87-95, 148-

164.

Ayer, A. J., “Replies,” in Macdonald, G. F. (ed.), Perception and Identity, Essays Presented to

A. J. Ayer, with His Replies (Ithaca, N. Y.: Cornell University Press, 1979): 277-333.

Belfrage, Bertil, “The Scientific Background of George Berkeley’s Idealism,” in Stephen Gersh

and Dermot Moran (eds.), Eriugena, Berkeley, and the Idealist Tradition, University of Notre

Dame Press, 2006.

Bengson, John, “How to Perceive the Past with your Eyes Shut,” 14th Annual Harvard/MIT

Graduate Student Philosophy Conference (March 2006).

204

Ben-Yami, Hanoch, “Apparent Simultaneity.“ In [2007] EPSA07: 1st Conference of the

European Philosophy of Science Association (Madrid, 15-17 November, 2007).

Berkeley, George, The Works of George Berkeley, 9 vols., A. A.

Luce and T. E. Jessop (eds.), London: Nelson, 1948-1957).

Bolender, John, “Factual Phenomenalism: A Supervenience Theory,” Sorites: An Electronic

Journal of Analytic Philosophy (1998) vol. 9: 15-30.

Bolender, John, “An Argument for Idealism,” Journal of Consciousness Studies, Vol. 8, No. 4

March 2001): 37-61.

Brain, W. Russell, Mind, Perception and Science (Oxford: Blackwell, 1951): 16.

Brewer, Bill, “Perception and content,” European Journal of Philosophy 14 (2) (2006): 165-

181.

Brewer, Bill, “Perception and its objects,” Philosophical Studies 132 (2007): 87–97

Bringsjord, Selmer, “How to Explain the Phi Phenomenon Without Dennett's Exotica: Good Ol'

Computation Does Just Fine,” presented at the 1996 annual Computing and Philosophy

Meeting, August, Pittsburgh, PA. www.rpi.edu/~brings/select.html

Broad, C. D., Scientific Thought (London: Routledge & Kegan Paul, 1923): 367f

Broad, C. D., Mind and its Place in Nature (London: Routledge & Kegan Paul, 1925): 249 ff.

Broad, C. D., “Professor Marc-Wogau’s Theorie der Sinnesdaten,“ Mind 56 (1947): 120-124.

Broad, C. D., "Some Elementary Reflexions on Sense-Perception," Philosophy, 27, 100 (1952):

3-17.

Broad, C. D. “A Reply to My Critics,” in The Philosophy of C.D. Broad, edited by Paul Arthur

Schilpp (New York: Tudor Publishing Company, 1959).

205

Butterfield, Jeremy, 1984, ‘Seeing the Present’, Mind, 93: 161–76. Reprinted in Le Poidevin,

Robin (ed.) Questions of Time and Tense (Oxford: OUP, 1998).

Carrier, L. S., “The Time-Gap Argument,” Australasian Journal of Philosophy 47 (1969): 263-

272.

Carrier, L. S., “Time-Gap Myopia,” Australasian Journal of Philosophy 50 (1972): 55-57.

Carroll, Noël, “Towards an Ontology of the Moving Image,” in: Cynthia A. Freeland and

Thomas E. Wartenberg (eds.), Philosophy and Film (New York: Routledge, 1995)

Changizi, Mark A., “The trade-off between speed and complexity,” Behavioral and Brain

sciences/Volume 31/Issue 02, 2008: 203

Changizi, Mark, The Vision Revolution: How the Latest Research Overturns Everything We

Thought We Knew About Human Vision (Dallas, Tex.: BenBella Books, 2009)

Chisholm, Roderick M., Perceiving: A Philosophical Study, (New York, Cornell University

Press, 1957): 153.

Chisholm, Roderick M., The Foundations of Knowing (Brighton, Sussex: The Harvester Press,

1982): 109-210

Chisholm, Roderick M., “Keith Lehrer and Thomas Reid,” Philosophical Studies 60: 33.

Chrucky, Andrew, “The Alleged Fallacy of the Sense-Datum Inference.” Paper read before the

New Jersey Regional Philosophy Conference (Seton Hall University, 1992) and the Eastern

Pennsylvania Philosophical Association (Bloomsburg University, 1992).

www.ditext.com/chrucky/sensa.html

Clark, Romane, “The Sensuous Content of Perception,” In: Hector-Neri Castaneda, ed., Action,

Knowledge, and Reality: Critical Studies in Honor of Wilfrid Sellars (Indianapolis: The Bobbs-

Merrill Company, 1975): 109-127.

Clark, Romane, “Sensuous Judgments,” Nous 7 (1973): 45-56.

206

Cornman, James W., Materialism and Sensations (New Haven and London: Yale University

Press, 1971): 218-227.

Cornman, James W., Perception, Common Sense, and Science (New Haven and London: Yale

University Press, 1975): 47-50; 93-213; 243-247; 353-355.

Currie, Gregory, Image and Mind: Film, Photography, and Cognitive Science (Cambridge:

Cambridge University Press, 1995)

Dancy, Johnathan, An Introduction to Contemporary Epistemology (Oxford, Cambridge, MA:

Basil Blackwell, 1985): 146-147; 152-153; 184-185.

Daniels, Charles B., “Seeing through a time-gap,” Australasian Journal of Philosophy, Vol. 48,

No. 3 (1970): 354-359.

Dennett, Daniel, Consciousness Explained (Boston: Back Bay Books, Little Brown and

Company, 1991): 101-170.

Dewey, John, “Naive Realism vs Presentative Realism”, Chapter 9 in Essays in Experimental

Logic. Chicago: University of Chicago (1916): 250-263.

Djucik, G. and Popescu, V. B., “A Critique of Langsam's ‘The Theory of Appearing

Defended’”, Philosophical Studies, Vol. 112, No. 1 (Jan. 2003): 69-91.

Dicker, Georges, Perceptual Knowledge (Dodrecht: D. Reidel Publishing Co., 1980): 48.

Dicker, Georges, Review of Smith, A. D., The Problem of Perception, in: European Journal of

Philosophy 14:3 (2006):

DiSalvo, David, http://neuronarrative.wordpress.com

Dretske, Fred, Seeing and Knowing (Chicago: University of Chicago Press, 1969): 4-77, 245-

257.

207

Dretske, Fred, Naturalizing the Mind (Cambridge, Mass.: MIT Press, 1995): 1, 12 ff, 19, 88, 93

Dretske, Fred, Personal communication (2003)

Dummett, Michael, “Common Sense and Physics,” in Macdonald, G. F. (ed.), Perception and

Identity, Essays Presented to A. J. Ayer, with His Replies (Ithaca, N. Y.: Cornell University

Press, 1979): 1-40.

Farnell, Derrick, “This Is A Simulation,” chainsofreason.org

Firth, Roderick, In Defense of Radical Empiricism. Essays and Lectures (ed. Troyer, John,

Lanham and Oxford: Rowman & Littlefield Publishers, Inc., 1998).

Foster, John, In Defence of Idealism (London: Routledge & Kegan Paul, 1982): 3-309.

Foster, John, Ayer (London: Routledge & Kegan Paul, 1985): 144-196.

Foster, John, “Berkeley on the physical world,” in Foster, John and Robinson, Howard (eds.),

Essays on Berkeley (Oxford: Oxford University Press, 1985): 83-108.

Foster, John, The Immaterial Self (London: Routledge, 1991).

Foster John, “The Succinct Case for Idealism,” in Robinson, Howard (ed.), Objections to

Physicalism (Oxford: Clarendon Press, 1993): 293-313.

Foster, John, “In Defence of Phenomenalistic Idealism,” Philosophy and Phenomenological

Research, Vol 54, No. 3 (Sep. 1994): 509-529.

Foster, John, The Nature of Perception (OUP, 2000).

Foster, John, Personal communication (2002-2005).

Foster, John, The Divine Lawmaker (OUP, 2004).

208

Foster, John, A World For Us: The Case for Phenomenalistic Idealism (OUP, 2008).

García-Carpintero, Manuel, “Sense data: The sensible approach,” Grazer Philosophische

Studien 62 (1) (2001): 17-63.

Gram, Moltke S., Direct Realism (The Hague: M. Nijhoff, 1983): 132-135; 137-139; 141-143.

Hahn, Lewis Edwin, ed., The Philosophy of A. J. Ayer (La Salle, Ill.: Open Court, 1992): 179-

200; 489-515; 602.

Harrison, Jonathan, (1993) “Science, souls and sense-data,” in Wright, E. L.

Representationalisms: Essays in the Philosophy of Perception (Brookfield: Avebury, 1993): 15-

44.

Henson, R. G., “Ordinary Language, Common Sense, and the Time-Lag Argument,” Mind 76

(1967): 21-33.

Hirst, R. J., The Problems of Perception (London and New York: George Allen and Unwin,

1959): 22; 65-66; 133.

Hopkins, Robert, Review of Foster, The Nature of Perception, in: The Philosophical Quarterly

Vol. 21, No. 205 October 2001: 552-555.

Houts, Ronald W., “Some Implications of the Time-Lag Argument,” Philosophy and

Phenomenology Research, 41 (1980): 150-157.

Huemer, Michael, Skepticism and the Veil of Perception (Rowman and Littlefield, 2001): 131-

135.

Inwagen, Peter van, Metaphysics (Boulder, San Francisco: Westview Press, 1993): 42-55.

Jackson, Frank, Perception (Cambridge: Cambridge University Press, 1977).

Jackson, Frank, “Representative Realism,” in: Jonathan Dancy and Ernest Sosa (eds.), A

Companion to Epistemology (Blackwell, 1992): 445-448.

209

Jackson, Frank, “Some reflections on representationalism” Draft at 5 April 2000 for NYU

seminar on 25 April 2000

http://www.nyu.edu/gsas/dept/philo/courses/consciousness/papers/RepresentationalismNYU5A

pril00.PDF

Jacob, Pierre, “Grasping and perceiving objects,” in Andrew Brook (ed.), Cognition and the

Brain: The Philosophy and Neuroscience Movement. Cambridge: Cambridge University Press

(2005)

Johansson, Ingvar, “Perception as the Bridge between Nature and Life-World.” In C. Bengt

Pedersen and N. Thomassen (eds.) Nature and Life-World, Odense University Press, 1998, pp.

113-137.

Johnston, Mark, “The Obscure Object of Hallucination,” Philosophical Studies 120 (2004): 113-

183.

Kant, Immanuel, Critique of Pure Reason, Norman Kemp Smith, trans. (New York: St.

Martin’s Press; Toronto: Macmillan, 1965).

Kardosh, John, “Special Relativity and the Time-lag Argument,” Sophia XI (2008): 23-32.

Kelly, S. D. (forthcoming). “Time and experience.” In A. Brooks & Kathleen Akins (eds.),

Philosophy and the Neurosciences. Cambridge.

Langsam, Harold, “The Theory of Appearing Defended,” Philosohical Studies, vol. 87, No. 1,

July (1997)

Lean, M., Sense-Perception and Matter (London: Routledge & Paul, 1953).

Leibniz, G. W., New Essays on Human Understanding, tr. P. Remnant and J. Bennett

(Cambridge: Cambridge University Press, 1982): 134-135.

Le Morvan P., “Arguments against Direct Realism and How to Counter Them,” The American

Philosophical Quarterly 41 (2004): 223-224.

210

Lewis, D., “Veridical Hallucination and Prosthetic Vision,” reprinted in Dancy, J. (ed.)

Perceptual Knowledge (OUP, 1988): 83.

Locke, Don, Perception and our Knowledge of the External World (London: Allen & Unwin;

New York: Humanities Press, 1967; Reissued by Routledge, 2004): 118.

Locke, John, An Essay Concerning Human Understanding,

http://oregonstate.edu/instruct/phl302/texts/locke/locke1/Essay_contents.html

Lockwood, Michael, Mind, Brain & the Quantum (Oxford, UK & Cambridge, USA: Blackwell

Publishers, 1989): 138-171.

Lovejoy, A. O., The Revolt Against Dualism (La Salle, IL: The Open Court Publishing Co.,

1960): 23-25; 75-85; 147.

Macdonald, G. F. (ed.), Perception and Identity, Essays Presented to A. J. Ayer, with His

Replies (Ithaca, N. Y.: Cornell University Press, 1979): 277-333.

Mackie, J. L., Problems from Locke (Oxford: Clarendon Press, 1976): 7-71.

Maclachlan, D. L. C., Philosophy of Perception (Englewood Cliffs, NJ: Prentice Hall, Inc.,

1989): 26, 30, 33-48.

Mandelbaum, M., Philosophy, Science, and Sense Perception (Baltimore: The John Hopkins

Press, 1964): 147, 177-182.

Martin, Michael W. (2000). Beyond dispute: Sense-data, intentionality, and the mind-body

problem. In Tim Crane & Sarah A. Patterson (eds.), The History of the Mind-Body Problem

Myers, G. E., “Perception and the ‘Time-Lag’ Argument,” Analysis 17 5 (1957): 97-102.

Mundle, C. W. K., Perception: Facts and Theories (OUP, 1971): 44-45; 92.

Nathan, N. M. L. (2005). Direct Realism: Proximate Causation and the Missing Object. Acta

Analytica 20 (36):3-6.

211

Noë, Alva, “Causation and Perception: The Puzzle Unravelled,” Analysis 63,2, April, 2003.

Noë, Alva, “Real Presence,” Philosophical Topics 33:1, 2005.

Noë, Alva, “Experience of he world in time,” Analysis 66.1, January 2006: 26–32.

Olding, Alan, “The Time-Gap Argument,” Metaphilosophy, Vol. 9, No. 1, Jan. 1978: 44-57.

O’Shaughnessy, Brian, Consciousness and the World (Oxford: Clarendon Press, 2000): 439-

514.

O’Shaughnessy, Brian, “Sense Data.” In: Barry Smith (ed.) John Searle (Cambridge University

Press, 2003): 169-188.

Pappas, George S., Berkeley’s Thought (Ithaca, N. Y.: Cornell University Press, 2000): 147-182.

Perkins, Moreland, Sensing the World (Indianapolis, Ind.: Hackett Publishing Co., 1983).

Pitcher, G. A, Theory of Perception (Princeton, N.J.: Princeton University Press, 1971): 44-50,

114-115.

Power, Sean Enda, “Perceiving External Things and the Time-Lag Argument,” European

Journal of Philosophy, 19: no. doi: 10.1111/j.1468-0378.2010.00436.x November 2010.

Quinton, Anthony, The Nature of Things (London and Boston: Rouledge & Kegan Paul, 1973):

194-202.

Reid, Thomas, Essays on the Intellectual Powers of Man (Cambridge, Massachusetts, and

London, England: The M.I.T. Press, 1969).

Robinson, Howard (ed.), Objections to Physicalism (Oxford: Clarendon Press, 1993): 1-326.

Robinson, Howard, Perception (London and New York: Routledge, 1994): 80-84, 89.

212

Robinson, Howard, Editor’s Introduction and Explanatory Notes to George Berkeley: Principles

of Human Knowledge, Three Dialogues (OUP, 1996)

Robinson, Howard, Personal communication (2002-2006).

Robinson, Howard, Review of Smith, A. D., The Problem of Perception, in: Mind. Vol. 117,

No. 466, April 2008: 520-524.

Russell, Bertrand, The Problems of Philosophy (London, 1912): 16-17, 21-22.

Russell, B., An Outline of Philosophy (W. W. Norton & Company, Inc., 1927): 138-142; 150-

156.

Russell, B., Human Knowledge: Its Scope and Limits (New York: Simon and Schuster, 1948):

203-209.

Sanford, David H., Review of Carrier, Leonard S., Experience and the Objects of Perception

(Washington: University Press of America, 1981). In: Nous, Vol. 21, No. 3 (Sep. 1987): 435-

438.

Seager, William E. & Bourget, David (2007). Representationalism about consciousness. In Max

Velmans & Susan Schneider (eds.), The Blackwell Companion to Consciousness. Blackwell.

Siegel, Susanna, “The Contents of Perception,” Stanford Encyclopedia of Philosophy (2005,

online)

Smith, A. D., “O’Shaughnessy’s Consciousness,” The Philosophical Quarterly, Vol 51, No.

205, Oct. 2001: 532-539

Smith, A. D., The Problem of Perception (Cambridge, Mass. and London, England: Harvard

University Press, 2002).

Smith, A. D., Personal communication (2003).

Smythies, John R., Analysis of Perception (London: Routledge and Kegan Paul, 1956):

213

Solso, Robert L., Cognition and the Visual Arts, MIT Press, 1996: 8, 26

Sorensen, Roy, “We See in the Dark,” Nous 38:3 (2004): 456–480

Suchting, W. A., “Perception and the Time-Lag Argument,” Philosophical Quarterly 19 (1969):

46-56.

Thrane, Gary, “Berkeley’s “Proper Object of Vision,” Journal of the History of Ideas, Vol. 38,

No. 2 (1977): 243-260.

Travis, Charles, 2004. “The Silence of the Senses,” Mind, 113 (449): 57-94.

Valberg, J. J., The Puzzle of Experience (Oxford: Clarendon Presss, 1992).

Valberg, J. J., Personal communication

Walton, Kendall L., “Transparent Pictures,” Critical Inquiry, 11:2 (1984), 246-77.

Also in Nous, Vol. 18, No. 1 (Mar., 1984), 67-72.

Walton, Kendall L., “On Pictures and Photographs: Objections Answered,” in: Film Theory and

Philosophy (Richard Allen, Murray Smith, eds.) Oxford University Press 1997.