PERCEPTION AND ITS OBJECTS

P. F. STRA WSON
Ayer has aIways given the proIe! o" per#eption a #entraI pIa#e in is thin$ing. ReasonaIy so% "ar a phi&osopher's views on this
()estion a"e a $ey oth to is theory o" $nowIe*ge in generaI an* to is !etaphysi#s. The !ove!ent o" Ayer's own tho)ght has
een "ro! pheno!enaIis! to what he *es#ries in is Iatest treat!ent o" the topi# as 'a sophisti#ate* Corro o" rea&is!'.+ The
epithet is *o)Iy apt. No a*e()ate a##o)nt o" the !atter #an e si!pIe% an* Ayer's a##o)nt, whi&e *isting)ishe* y is
a##)sto!e* I)#i*ity an* e#ono!y o" sty&e, is notaIy an* s)t&y responsive to a&I the #o!pIe-ities inherent in the s).e#t itse&"
an* to a&I the press)res o" !ore or Iess pers)asive arg)!ent whi#h ave !ar$e* the #o)rse o" its treat!ent y phi&osophers. /et
the "or! o" rea&is! he *e"en*s has another $in* o" sophisti#ation ao)t whi#h it is possiie to ave reservations an* *o)ts% an*,
tho)gh I a! #ons#io)s o" eing "ar tra! #&ear on the !atter !yseI", I sha&& try to !a$e so!e o" !y own *o)ts an* reservations
as #&ear as I #an. I sha&& ta$e as !y te-t Chapters 0 an* 1 o" The Centrai Questions oj Philosophy; an* I sha&& aiso #onsi*er a
*i""erent $in* o" rea&is!2that a*vo#ate* y
J. 3. 4a#$ie in is oo$ on 3o#$e.5 There a"e points o" #onta#t as we&I as o" #ontrast etween Ayer's an* 4a#$ie's views. A #o!2
parison etween the! wi&& heIp to ring o)t the nat)re o" !y reservations ao)t oth.
A##or*ing to Ayer, the starting2point o" serio)s tho)ght on the !atter o" per#eption #onsists in the "a#t that o)r nor!ai
per#ept)ai .)*ge!ents aIways 'go eyon*' the sensiie e-perien#e whi#h gives
Reprinte* y per!ission o" the a)thor an* 4a#!i&&an, 3on*on an* Basingsto$e,
"ro! 6. 4#Dona&* 7e*.8, Perception and Identity: Essays presented to A. J. Ayer 79:;:8, 092<=.
9 A. J. Ayer, The CentraI Questions oJ Philosophy 73on*on, 9:;>8 #hs. 0 an* 1,
pp. <?2999.
5 J. 3. 4a#$ie, ProblemsJrom Loce 7O-"or*, 9:;<8 #hs. 9 an* 5, pp. ;2;+.
rise to the!% "ar those .)*ge!ents #arry i!p&i#ations whieh wo)&* not e #arrie* y any 'striet a##o)nt' o" that e-perien#e.> Ayer
sees or*inary per#ept)a& .)*ge!ents as re"&e#ting or e!o*ying what he #a&&s the #o!!on2sense view o" the physiea& wor&*,
whieh is, a!ong other things, a rea&ist view% an* he sees that view itse&" as having the #hara#ter o" 'a theory with respe#t to the
i!!e*iate *ata o" per#eption'. 0 @e *evotes so!e spa#e to an a##o)nt o" how the theory !ight e seen as #apa&e o" eing
*eve&ope* y an in*ivi*)aI oserver on the asis o" the *ata avai&a&e to iro% tho)gh he *isavows any intention o" giving an
a#t)a& history o" the theory's *eve&op!en&. The p)rpose o" the a##o)nt is, rather, to ring o)t those "eat)res o" sensi&e e-perien#e
whieh !a$e it possi&e to e!p&oy the theory s)##ess")&&y an* whieh, in*ee*, .)sti"y a##eptan#e o" i&. For it is, he ho&*s, y an*
&arge an a##epta&e theory, even tho)gh the *is#overies o" physi#a& s#ien#e !ay re()ire )s to !o*i"y it in #ertain respe#ts.
Evi*ent&y no in"ant is *e&ivere* into the wor&* a&rea*y e()ippe* with what Ayer #a&&s the #o!!on2sense view o" i&. That view
has to e a#()ire*% an* it is open to the psy#ho&ogist o" in"ant &earning to pro*)#e at &east a spe#)&ative a##o)nt o" the stages o"
its a#()isition. Ayer insists, as I ave re!ar$e*, that is own a##o)nt o" a possi&e &ine o" *eve&op!ent or #onstr)#tion o" the
#o!!on2sense view is not inten*e* as a spe#)&ative #ontri)tion to the theory o" in"ant &earning. It is inten*e*, rather, as an
ana&ysis o" the nat)re o" !at)re or a*)&t per#ept)a& e-perien#e, an ana&ysis *esigne* to show .)st how #ertain "eat)res o" !at)re
sensi&e e-perien#e vin*ieate or s)stain the #o!!on2sense view whieh is e!o*ie* or re"&e#te* in !at)re per#ept)a&.)*ge!ents.
C&ear&y the two ai!s ere *isting)ishe*2the genetie2psy#ho&ogiea& an* the ana&yti#2phi&osophiea&2are very *i""erent in*ee*, an* it
wi&& e o" great i!portan#e not to #on")se the!. In partie)&ar it wi&& e i!portant to r)n no ris$ o" #hara#teriAing !at)re sensi&e
e-perien#e in ter!s a*e()ate at est on&y "or the #hara#teriAation o" so!e stage o" in"anti&e e-perien#e. It is not #&ear that Ayer
entire&y avoi*s this *anger.
What is #&ear is that i" we a##ept Ayer's starting2point, i" we agree that O)r or*inary per#ept)a& .)*ge!ents #arry i!p&i#ations
not #arrie* y a 'striet a##o)nt' o" the sensi&e e-perien#e whi#h gives rise to the!, then we !)st !a$e aso&)te&y s)re that o)r
a##o)nt o" that e-perien#e, in the "or! it ta$es in o)r !at)re &ire, is in*ee* stri#t2in
> Ayer, Centrai Questions! ?9, ?:. 0 Ii*. 88.
the sense o" stri(t&y #orre#t. On&y so #an we ave any prospe#to" !a$ing a #orre#t esti!ate o" the ")rther *o#trines that the
#o!!onsense view o"the wor&* has the stat)s o" a theory"ith respe#t to a type o" sensi&e e-perien#e whi#h provi*es data "ar the
theory% that this e-perien#e s)pp&ies the e#idence on whi#h the theory is ase*% 1 that the #o!!on2sense view #an e regar*e* as
in$erred or at &east in"erra&e "ro! this evi*en#e% an* that o)r or*inary per#ept)a& .)*ge!ents ave the #hara#ter o"
interpretations!% in the &ight o" theory, o" what sensi&e e-perien#e a#t)a&Iy presents )s with.
B)t #an we2an* sho)&* we2a##ept Ayer's starting2pointB 9 thin$ that, s)ita&y interprete*, we oth #an, an* sho)&*, a##ept it.
Two things wi&& e re()ire* o" a stri#t a##o)nt o" o)r sensi&e e-perien#e or o" any parti#)&ar episo*e or s&i#e o" sensi&e e-peri 2
en#eC "irst, as 9 ave .)st re!ar$e*, that it sho)&* in no way *istort or !isrepresent the #hara#ter o" that e-perien#e as we a#t)a&Iy
en.oy it, that is, that it sho)&* e a tr)e or "aith")& a##o)nt% se#on*&y, that its tr)th, in any parti#)&ar #ase, sho)&* e in*epen*ent
o" the tr)th o" the asso#iate* per#ept)a& .)*ge!ent, that is, that it sho)&* re!ain tr)e even i" the asso#iate* per#ept)a& .)*ge!ent
is "a&se. It is the se#on* re()ire!ent on whi#h Ayer &ays stress when he re!ar$s that those .)*ge!ents #arry i!p&i#ations whi#h
wo)&* not e #arri e* y any stri#t a##o)nt o" sensi&e e-perien#e% or, &ess happi&y in !y opinion, that in !a$ing s)#h .)*ge!ents
we ta$e a step eyon* what o)r sensi&e e-perien#e a#t)a&Iy presents )s with. B)t it is the "irst re()ire!ent to whi#h 9 now wish
to give so!e attention.
S)ppose a non2phi&osophi#a& oserver gaAing i*&y thro)gh a win*ow. To hi! we a**ress the re()est, '6ive )s a *es#ription o"
yo)r #)rrent vis)a& e-perien#e', or '@ow is it with yo), vis)a&Iy, at the !o!entB' Dn#a)tione* as to e-a#t&y what we want, he
!ight rep&y in so!e s)#h ter!s as theseC 'I see the re* &ight o" the setting s)n "i&tering thro)gh the &a#$ an* thi#$&y #9)stere*
ran#hes o" the e&!s% 9 see the *app&e* *eer graAing in gro)ps on the vivi* green grass . . .' an* so on. So we e-p&ain to hi!. We
e-p&ain that we want hi! to a!en* is a##o)nt so that, witho)t any sa#ri"i#e o" "i*e&ity to the e-perien#e as a#t)a&Iy en.oye*, it
neverthe&ess she*s a&I that heavy &oa* o" #o!!it!ent to propositions ao)t the wor&* whi#h was #arri e* y the *es#ription he
gave. We want an a##o)nt whi#h #on"ines itse&" stri#t&y within the &i!its o" the s).e#tive
1 Ii*. ?:.
< Ii*. 81.
episo*e, an a##o)nt whi#h wo)&* re!ain tr)e even i" he ha* seen nothing o" what he #&ai!e* to see, even i" he ha* een s).e#t
to tota& i&&)sion.
O)r oserver is ()i#$ on the )pta$e. @e *oes not start ta&$ing ao)t &ights an* #o&o)rs, pat#hes an* patterns. For he sees that to
*o so wo)&* e to "a&si"y the #hara#ter o" the e-perien#e he a#t)a&Iy en.oye*. @e says, instea*, 'I )n*erstan*. &'ve got to #)t o)t o"
!y report a&I #o!!it!ent to propositions ao)t in*epen*ent&y e-isting o.e#ts. We&I, the si!p&est way to *o this, whi&e
re!aining "aith")& to the #hara#ter o" the e-perien#e as a#t)a&Iy en.oye*, is to p)t !y previo)s report in inverte* #o!!as or
oratio o&i()a an* *es#rie !y vis)a& e-perien#e as s)#h as it wo)&* ave een nat)ra& to *es#rie in these ter!s, ha* I not
re#eive* this a**itiona& instr)#tion. Th)sC EI ha* a vis)a& e-perien#e s)#h as it wo)&* ave een nat)ra& to *es#rie y saying that
I saw, et#. . . . For, to *es#rie in these wor*s, 'I saw . . . et#.'G were it not "ar the o&igation to e-#&)*e #o!!it!ent to
propositions ao)t in*epen*ent&y e-isting o.e#ts.E In this way F#ontin)es the oserverG I use the per#ept)a& #&ai!2the #&ai! it
was nat)ra& to !a$e in the #ir#)!stan#es2in ar*er to #hara#teriAe !y e-perien#e, witho)t a#t)a&Iy !a$ing the #&ai!. I ren*er the
per#ept)a& .)*ge!ent internaI to the #hara#teriAation o" the e-perien#e witho)t a#t)a&&y asserting the #ontent o" the .)*ge!ent.
An* this is rea&&y the est possi&e way o" #hara#teriAing the e-perien#e. There a"e perhaps a&ternative &o#)tions whi#h !ight
serve the p)rpose, so &ong as they a"e )n*ersto o* as eing to the sa!e e""e#t2on the who&e, the !ore arti"i#ia& the etter, sin#e
their arti"i#ia&ity wi&& he&p to !a$e it #&earer .)st to what e""e#t they a"e inten*e* to e. Th)s we !ight aveC EIt sensi&y see!e*
to !e .)st as i" I were seeing s)#h2an*2s)#h a s#eneE or E4y vis)a& e-perien#e #an e #hara#teriAe* y saying that I saw what I
saw, s)pposing I saw anything, as a s#ene o" the "o&&owing #hara#ter . . .'E
I" !y oserver is right in this2an* I thin$ he is2then #ertain generaI #on#&)sions "o&Iow. O)r per#ept)a& .)*ge!ents, as Ayer
He!ar$s, e!o*y or re"Ie#t a #ertain view o" the wor&*, as #ontainIOg o.e#ts, vario)s&y propertie*, &o#ate* in a #o!!on spa#e
an* #ontin)ing in their e-isten#e in*epen*ent&y o" o)r interr)pte* an* re&ative&y "Ieeting per#eptions o" the!. O)r !a$ing o"
s)#h .)*gerne,nts i!p&ies o)r possession an* app&i#ation o" #on#epts o" s)#h OHe#ts. B)t now it appears that we #annot give a
veri*i#a& #hara#tenAation even o" the sensi&e e-perien#e whi#h these .)*ge!ents, as Ayer e-presses i&, 'go eyon*', witho)t
re"eren#e to those .)*ge!ents the!se&ves% that o)r sensi&e e-perien#e itse&" is thoro)gh&y per!eate* with those #on#epts o"
o.e#ts whi#h "ig)re in s)#h .)*ge!ents. This *oes noi !ean, that is, it *oes noi "o&&ow *ire#t&y "ro! this "eat)re o" sensi&e
e-perien#e, that the generaI view o" the wor&* whi#h those .)*ge!ents re"&e#t !)si e tr)e. That wo)&* e too short a way with
s#epti#is!. B)t it *oes "o&&ow, I thin$, that o)r sensi&e e-perien#e #o)&* noi ave the #hara#ter it *oes ave )n&ess2at &east e"ore
phi&osophi#a& re"&e#tion sets in2 we )n()estioning&y too that generaI view o" the wor&* to e tr)e. The #on#epts o" the o.e#tive
whi#h we see to e in*ispensa&e to the veri*i#a& #hara#teriAation o" sensi&e e-perien#e si!p&y wo)&* noi e in this way
in*ispensa&e )n&ess those whose e-perien#e it was initia&Iy an* )nre"&e#tive&y too$ s)#h #on#epts to ave app&i #ation in the
wor&*.
This has a ")rther #onse()en#eC the #onse()en#e that it is ()ite inappropriate to represent the generaI, rea&ist view o" the wor&*
whi#h is re"&e#te* in o)r or*inary per#ept)a& .)*ge!ents as having the stat)s o" a theory with respe#t to sensi&e e-perien#e% that
it is inappropriate to represent that e-perien#e as s)pp&ying the data "or s)#h a theory or the e#idence on whi#h it is ase* or "ro!
whi#h it is in$erred or in$errable; that it is inappropriate to spea$ o" o)r or*inary per#ept)a& .)*ge!ents as having the #hara#ter
o" an interpretation! in the &ight o" theory, o" the #ontent o" o)r sensi&e e-perien#e. The reason "or this is si!p&e. In or*er "or
so!e e&ie" or sei o" e&ie"s to e #orre#t&y *es#rie* as a theory in respe#t o" #ertain *ata, it !)si e possi&e to *es#rie the *ata
on the asis o" whi#h the theory is he&* in ter!s whi#h *o noi pres)ppose the a##eptan#e o" the theory on the pari o" those "or
who! the *ata ore *ata. B)t this is .)st the #on*ition we ave seen noi to e satis"ie* in the #ase where the so2#a&&e* *ata a"e the
#ontents o" sensi&e e-perien#e an* the so2#a&&e* theory is a generaI rea&ist view o" the wor&*. The '*ata' a"e &a*en with the
'theory'. Sensi&e e-perien#e is per!eate* y #on#epts )nre"&e#tive a##eptan#e o" the generaI app&i#ai&ity o" whi#h is a #on*ition
o" its eing so per!eate*, a #on*ition o" that e-perien#e eing what it is% an* these #on#epts a"e o" rea&isti#a&Iy #on#eive*
o.e#ts.
I !)si !a$e it ()ite *ear what I a! saying an* what I a! no& saying ere. I a! ta&$ing o" the or*inary non2phi&osophi#a& !an.
I a! ta&$ing o" )s a&I e"ore we "e&t, i" ever we *i* "ee&, any in*i nation to respon* to the so&i#itations o" a generaI s#epti#is!, to
regar* it as raising a pro&e!. I a! saying that it "o&Iows "ro! the #hara#ter o" sensi&e e-perien#e as we a&I a#t)a&Iy en.oy it that
a #o!!on2sense rea&ist view o" the wor&* *oes noi in generaI ave the stat)s o" a theory in respe#t o" that e-perien#e% whi&e Ayer,
as I )n*erstan* iro, ho&*s that it *oes. B)t I a! noi *enying that to one who has seen, or thin$s he has seen, that sensi&e
e-perien#e !ight ave the #hara#ter it *oes ave an* yet a rea&ist view or the wor&* e "a&se, to him the i*ea !ay we&I present
itse&" that the est way o" a##o)nting "ar sensi&e e-perien#e as having that #hara#ter is to a##ept the #o!!on rea&ist view o" the
wor&* or so!e variant o" it. &e !ight e sai* to a*opt, as a theory, the *o#trine that the #o!!on rea&ist view o" the wor&* is, at
&east in so!e asi# essentia&s, tr)e. B)t this wi&& e a phi&osopher's theory, *esigne* to *ea& with a phi&osopher's pro&e!. 7I sha&&
noi ere *is#)ss its !erits as s)#h.8 What I a! #on#erne* to *isp)te is the *o#trine that a rea&ist view o" the wor&* has, "ar any
!an, the stat)s o" a theory in re&ation to is sensi&e e-perien#e, a theory in the &ight o" whi#h he interprets that e-perien#e in
!a$ing is per#ept)a& .)*ge!ents.
To p)t the point s)!!ari&y, whereas Ayer says we ta$e a step eyon* o)r sensi&e e-perien#e in !a$ing o)r per#ept)a& .)*ge2
!ents, I say rather that we ta$e a step a#$ 7in generaI8 "ro! o)r per#ept)a& .)*ge!ents in "ra!ing a##o)nts o" o)r sensi&e
e-perien#e% "ar we ave 7in generaI8 to in#&)*e a re"eren#e to the "or!er in "ra!ing a veri*i#a& *es#ription o" the &atter.
It !ay see!, on a s)per"i#ia& rea*ing, that Ayer ha* anti#ipate* an* answere* this o.e#tion. @e intro*)#es, as ne#essary "ar
the #hara#teriAation o" o)r sensi&e e-perien#e, #ertain #on#epts o" types o" pattern, the na!es "ar whi#h a"e orrowe* "ro! the
na!es o" or*inary physi#a& o.e#ts. Th)s he spea$s o" vis)a& &ea" patterns, #hair patterns, #at patterns, an* so on.; At the sa!e
&i!e, he is #are")&, i" I rea* hi! right&y, to g)ar* against the i!pression that the )se o" this ter!ino&ogy #o!!its hi! to the view
that the e!p&oy!ent o" the #orrespon*ing physi#a&2o.e#t #on#epts the!se&ves is ne#essary to the #hara#teriAation o" o)r sensi&e
e-perien#e.? The ter!ino&ogy isappropriate 7he ho&*s8 si!p&y e#a)se those "eat)res o" sensi&e e-perien#e to whi#h the
ter!ino&ogy is app&ie* a"e the "eat)res whi#h govern o)r i*enti"i#ations o" the physi#a& o.e#ts we thin$ we see. They a"e the
"eat)res, 'i!p&i#it&y noti#e*' , : whi#h provi*e the !ain #&)es on whi#h o)r every*ay JD*ge!ents o" per#eption a"e ase*.
; Ii*. :9.
? Ii*. :<.
: Ii*. :9
This is ingenio)s, )i I *o noi thin$ it wi&& *o. This we #an see !ore #&ear&y i" we )se an invente*, rather than a *erive*
ter!ino&ogy "or these s)ppose* "eat)res an* then *raw )p a ta&e o" e-p&i#it #orre&ations etween the invente* na!es an* the
physi#a&o.e#t na!es. Ea#h arti"i#ia& "eat)re Da!e is sei against the Da!e o" a type o" physi#a& o.e#tC o)r per#ept)a&
i*enti"i#ations o" seen o.e#ts as o" that type a"e he&* to e governe* y i!p&i#it noti#ings o" that "eat)re. The nat)re an*
signi"i#an#e o" the "eat)re na!es is now ()ite #&ear&y e-p&aine* an* we ave to as$ o)rse&ves whether it is these rather than the
asso#iate* physi#a&2o.e#t ter!s that we o)ght to )se i" we a"e to give a ()ite stri#t an* "aith")& a##o)nt o" o)r sensi&e
e-perien#e. I thin$ it is #&ear that this is noi so% that the i*ea o" o)r or*inary per#ept)a& .)*ge!ents as eing invaria&y ase*
)pon, or invaria&y iss)ing "ro!, awareness o" s)#h "eat)res is a !yth. The sit)ation is rather, as I ave a&rea*y arg)e*, that the
e!p&oy!ent o" o)r or*inary, ")&&2&oo*e* #on#epts o" physi#a& o.e#ts is in*ispensa&e to a striet, an* stri#t&y veri*i#a&, a##o)nt
o" o)r sensi&e e-perien#e.
On#e again, I !)si !a$e it #&ear what I a!, an* what I a! noi, saying. I ave een spea$ing o" the typi#a& or stan*ar* #ase o"
!at)re sensi&e an* per#ept)a& e-perien#e. I ave no interest at a&I in *enying the thesis that there a&so o##)r #ases o" sensi&e
e-perien#e s)#h that the e!p&oy!ent o" ")&&2&oo*e* #on#epts o" physi#a& o.e#ts wo)&* noi e in*ispensa&e, an* !ay e
inappropriate, to giving a striet a##o)nt o" the e-perien#e. S)#h #ases a"e o" *i""erent types, an* there is ODe in parti#)&ar whieh
is o" interest in the present #onne#tion. An oserver, gaAing thro)gh is win*ow, !ay perhaps, y an e""ort o" wi&&, ring hi!se&"
to see, or even wi&&&ess&y "in* hi!se&" seeing, what he $nows to e the ran#hes o" the trees no &onger as ran#hes at a&&, )i as an
intrieate pattern o" *ar$ &ines o" #o!p&e- *ire#tions an* shapes an* vario)s siAes against a a#$gro)n* o" varying sha*es o" grey.
The "ra!e o" !in* in whi#h we en.o/%2i"2we ever *o en.oy, this $in* o" e-perien#e is a rare an* sophisti#ate*, noi a stan*ar* or
nor!a&, "ra!e o" !in*. Perhaps the "a#t, i" it is a "a#t, that we #an ring o)rse&ves into this "ra!e o" !in* when we #hoose !ay
e he&* to give a sense to the i*ea o" o)r 'i!p&i#it&y noti#ing' s)#h patterns even when we a"e noi in this "ra!e o" !in*. I" so, it is
a sense very "ar re!ove* "ro! that whi#h Ayer's thesis re()ires. For that thesis re()ires noi si!p&y the possi i&ity, )i the a#t)a&
o##)rren#e, in a&I #ases o" per#eption, o" sensi&e e-perien#e o" this $in*. ODe &ine o" retreat !ay see! to &ie open at this pointC a
retreat to the position o" saying that the o##)rren#e o" s)#h e-perien#es !ay e in$erred! even tho)gh we *o not, in the h)rry o"
&ire, genera&&y noti#e or re#a&& their o##)rren#e. B)t s)#h a retreat wo)&* e the "ina& irony. The ite!s in ()estion wo)&* ave
#hange* their stat)s ra*i#a&&yC instea* o" *ata "ar a #o!!on2sense theory o" the wor&*, they wo)&* appear as #onse()en#es o" a
sophisti#ate* theory o" the !in*.
This #on#&)*es the "irst stage o" !y arg)!ent. I ave arg)e* that !at)re sensi&e e-perien#e 7in generaI8 presents itse&" as, in
Iantian phrase, an immediate #ons#io)sness o" the e-isten#e o" things o)tsi*e )s. 'Immediate! o" #o)rse, *oes not !ean
in$allible.( @en#e, the #o!!on rea&ist #on#eption o" the wor&* *oes not ave the #hara#ter o" a 'theory' in re&ation to the '*ata o"
sense'. I ave not #&ai!e* that this "a#t is o" itse&" s)""i#ient to 're")te' s#epti#is! or to provi*e a phi&osophi#a& '*e!onstration' o"
the tr)th o" so!e "or! o" rea&is!% tho)gh I thin$ it *oes provi*e the right starting point "ar re"&e#tion )pon these enterprises. B)t
that is another story an* I sha&& not try to te&& it ere. 4y point so "ar is that the or*inary h)!an #o!!it!ent to a #on#ept)a&
s#he!e o" a rea&ist #hara#ter is not proper&y *es#rie*, even in a stret#he* sense o" the wor*s, as a theoreti#a& #o!!it!ent. It is,
rather, so!ething given with the given.
II
B)t we a"e phi&osophers as we&& as !en% an* so !)st e-a!ine !ore #&ose&y the nat)re o" the rea&ist s#he!e to whi#h we a"e pre2
theoreti#a&&y #o!!itte* an* then #onsi*er whether we a"e not rationa&&y #onstraine*, as 3o#$e an* 4a#$ie wo)&* !aintain we
a"e, to !o*i"y it ()ite ra*i#a&&y in the &ight o" o)r $now&e*ge o" physi#s an* physio&ogy. Sho)&* we not a&so, as phi&osophers,
#oosi*er the ()estion o" whether we #an rationa&&y !aintain any "or! o" rea&is! at a&&B Perhaps we sho)&*% )t, as a&rea*y
re!ar$e*, that is a ()estion I sha&& not #onsi*er ere. 4y !ain o.e#t, in the present se#tion, is to get a #&ear view o" the !ain
"eat)res o" o)r pre2theoreti#a& s#he!e e"ore #onsi*ering whether It IS *e"ensi&e, as it stan*s, or not, I go in a so!ewhat
ro)n*ao)t Way to wor$.
I have spo$en o" o)r pre2theoreti#a& s#he!e as rea&ist in #hara#ter. phi&osophers who treat o" these ()estions #o!!on&y
*isting)ish I""erent "or!s o" rea&is!o So *o oth Ayer an* 4a#$ie. They oth !ention, at ODe e-tre!e, a "or! o" rea&is! whi#h
4a#$ie #a&Is 'naive' an* even 'very naive', )i whi#h !ight !ore appropriate&y e #a&&e* '#on")se* rea&is!'. A s)""erer "ro!
#on")se* rea&is! "aiJs to *raw any *istin#tion etween sensi&e e-perien#es 7or 'per#eptions'8 an* in*epen*ent&y e-isting things
7or 'o.e#ts per#eive*'8 )i is sai* 7y 4a#$ie e-po)n*ing @)!e8 to #re*it the "or!er with persistent )noserve* e-isten#e. IO It
sho)&* e re!ar$e* that, i" this is an a##)rate way o" *es#riing the naive rea&ist's #on#eption o" the !atter, he !)si e very
#on")se* in*ee*, sin#e the e-pression ')noserve*' a&rea*y i!p&ies the *istin#tion whi#h he is sai* to "aiJ to !a$e. Spea$ing in
is own person, 4a#$ie gives no positive a##o)nt o" the naive rea&ist's view o" things, )i si!p&y says that there is, histori#a&&y,
in the tho)ght o" ea#h o" )s, a phase in whi#h we "ai& to !a$e the *istin#tion in ()estion.&9 It !ay in*ee* e so. The point is ODe
to e re"erre* to the e-perts on in"anti&e *eve&op!ent. B)t in any #ase the !atter is noi ere o" any #onse()en#e. For we a"e
#on#erne* with !at)re per#ept)a& e-perien#e an* with the #hara#ter o" the s#he!e to whi#h those who en.oy s)#h e-perien#e a"e
pre2theoreti#a&&y #o!!itte*. An* it see!s to !e as #ertain as anything #an e that, as an integrai pari o" that s#he!e, we
*isting)ish, nat)ra&&y an* )nre"&e#tive&y, etween o)r seeings an* hearings an* "ee&ings2o)r per#eivings2o" o.e#ts an* the
o.e#ts we see an* hear an* "ee&% an* hen#e ()ite #onsistent&y a##ept oth the interr)pte*ness o" the "or!er an* the #ontin)an#e
in e-isten#e, )noserve*, o" the &atter.
At the opposite e-tre!e "ro! naive rea&is! stan*s what !ay e #a&I e* s#ienti"i# or 3o#$ian rea&is!o This "or! o" rea&is!
#re*its physi#a& o.e#ts on&y with those o" their properties whi#h a"e !entione* in physi#a& theory an* physi#a& e-p&anation,
in#&)*ing the #a)saI e-p&anation o" o)r en.oy!ent o" the $in* o" per#ept)a& e-perien#e we in "a#t en.oy. It has the #onse()en#e
that we *o not, an* in*ee* #annot, per#eive o.e#ts as they rea&Iy a"e. It !ight e sai* that this #onse()en#e *oes noi ho&* in an
)n()a&i"ie* "or!. For we per#eive 7or see! to per#eive8 o.e#ts as having shape, siAe, an* position% an* they rea&Iy *o ave
shape, siAe, an* position an* !ore or &ess s)#h shape, siAe, an* position as we see! to per#eive the! as having. B)t this rep&y
!is#onstr)es the inten*e* "or#e o" the a&&ege* #onse()en#e. We #annot in sense per#eption2the point is
an o&* one2e#o!e aware o" the shape, siAe an* position o"
IO 4a#$ie, Problems! <;. I& Ii*. <?.
physi#a& o.e#ts e-#ept y way o" awareness o" o)n*aries *e"ine*
in so!e sensory !o*e2"or e-a!p&e, y vis)a& an* ta#ti&e ()a&ities s)#h as s#ienti"i# rea&is! *enies to the o.e#ts the!se&ves% an*
no #hange in, or a**ition to, o)r sensory e()ip!ent #o)&* a&ter this "a#t. To per#eive physi#a& o.e#ts as, a##or*ing to s#ienti"i#
rea&is!, they rea&&y a"e wo)&* e to per#eive the! as &a#$ing any s)#h ()a&ities. B)t this notion is se&"2#ontra*i#tory. So it is a
ne#essary #onse()en#e o" this "or! o" rea&is! that we *o not per#eive o.e#ts as they rea&&y a"e. In*ee*, in the sense o" the pre2
theoreti#a& notion o" per#eiving2that is, o" i!!e*iate awareness o" things o)tsi*e )s2we *o not, on the s#ienti"i#2rea&ist view,
per#eive physi#a& o.e#ts at a&I. We a"e, rather, the vi#ti!s o" a syste!ati# i&&)sion whi#h ostinate&y #&ings to )s even i" we
e!ra#e s#ienti"i# rea&is!o For we #ontin)e to en.oy e-perien#e as o" physi#a& o.e#ts in spa#e, o.e#ts o" whi#h the spatia&
#hara#teristi#s an* re&ations a"e *e"ine* y the sensi&e ()a&ities we per#eive the! as having% )t there a"e no s)#h physi#a&
o.e#ts as these. The on&y tr)e physi#a& o.e#ts a"e ite!s syste!ati#a&&y #orre&ate* with an* #a)sa&&y responsi&e "or that
e-perien#e% an* the on&y sense in whi#h we con e sai* to per#eive the! is .)st that they #a)se )s to en.oy that e-perien#e.
These re!ar$s a"e inten*e* on&y as a description o" s#ienti"i# rea&is!o I *o not #&ai! that they show it to e )ntena&e. I sha&&
ret)rn to the topi# &ater.
In etween the 'naive' an* the 's#ienti"i#' varieties, Ayer an* 4a#$ie ea#h re#ogniAe another Corro o" rea&is!, whi#h they ea#h
as#rie to '#o!!on sense'. B)t there is a *i""eren#e etween Ayer's version o" #o!!on2sense rea&is! an* 4a#$ie's. For 4a#$ie's
version, )n&i$e Ayer's, shares ODe #r)#ia& "eat)re with s#ienti"i# rea&is!o
The theory o" per#eption asso#iate* with s#ienti"i# or 3o#$ian rea&is! is #o!!on&y an* reasona&y *es#rie* as a
representative theory. Ea#h o" )s see!s to hi!se&" to e per#ept)a&&y aware o" o.e#ts o" a #ertain $in*C o.e#ts in spa#e o)tsi*e
)s with vis)a& an* ta#ti&e ()a&ities. There a"e in "a#t, on this view, no s)#h o.e#ts% )t these o.e#t appearan#es #an in a roa*
sense e sai* to e representative o" those a#t)a& o.e#ts in spa#e o)tsi*e )s whi#h a"e syste!ati#a&&y #orre&ate* with the
appearan#es an* #a)sa&&y respons&&e "or the!. The interesting "eat)re o" 4a#$ie's version o" #H!!on2sense rea&is! is that the
theory o" per#eption asso#iate* With it is no &ess a representative theory than that asso#iate* with 3o#$ian rea&is!o The
*i""eren#e is si!p&y that #o!!on sense a##or*ing to 4a#$ie, views o.e#t appearan#es as !ore "aith")K representatives o" a#t)a&
physi#a& o.e#ts than the 3o#$ian a&&owsC in that #o!!on sense, grat)ito)s&y y s#ienti"i# stan*ar*s, #re*its a#t)a& o.e#ts in
spa#e o)tsi*e )s with vis)a& an* ta#ti&e as we&& as pri!ary ()a&itieso As 4a#$ie p)ts it, #o!!on sense a&&ows '#o&o)rs2as2we2see2
the! to e resemblances o" ()a&ities a#t)a&&y in the things' =95 On oth views, sensi&e e-perien#e has its Own, sensi&e o.e#ts%
)t the #o!!on2sense view, a##or*ing to 4a#$ie, a&&ows a $in* o" rese!&an#e etween sensi&e an* physi#a& o.e#ts whi#h the
s#ienti"i# view *oes not.
I hope it is a&rea*y #&ear that this version o" #o!!on2sense rea&is! is ()ite *i""erent "ro! what I ave #a&&e* o)r pre2
theoreti#a& s#he!eo What we or*inari&y ta$e o)rse&ves to e aware o" in per#eption a"e not rese!&an#es o" physi#a& things
the!se&veso This *oes not !ean, as a&rea*y re!ar$e*, that we ave any *i""i#)&ty in *isting)ishing etween o)r e-perien#es o"
seeing, hearing an* "ee&ing o.e#ts an* the o.e#ts the!se&ves. That *istin#tion is as "ir!&y a part o" o)r pre2theoreti#a& s#he!e as
is o)r ta$ing o)rse&ves, in generaI, to e i!!e*iate&y aware o" those o.e#ts. Nor *oes it !ean that we ta$e o)rse&ves to e
i!!)ne "ro! i&&)sion, ha&&)#ination, or !ista$e. We #an, an* *o, per"e#t9y a*e()ate&y *es#rie s)#h #ases witho)t what is, "ro!
the point o" view o" the pre2theoreti#a& s#he!e, the ()ite grat)ito)s intro*)#tion o" sensi&e o.e#ts interpose* etween )s an*
the a#t)a& physi#a& o.e#ts they a"e s)ppose* to represent.
The o** thing ao)t 4a#$ie's presentation is that at ODe point he shows hi!se&" to e per"e#t9y we&& aware o" this "eat)re o"
the rea& rea&is! o" #o!!on sense% "ar he writes, 'What we see! to see, "ee&, hear an* so on o . o ore seen as real thin)s "ithout
)s2that is, o)tsi*e )s. We .)st see things as eing si!p&y there, o" s)#h2an*2s)#h sorts, in s)#h2an*2s)#h re&ationso . . ' 9> @e
goes OD, o" #o)rse, to say that 'o)r seeing the! so is &ogi#a&&y *istin#t "ro! their eing so', that we !ight e, an* in*ee* a"e,
wrongo B)t he wo)&* s#ar#e&y *isp)te that what is th)s seen as rea& an* o)tsi*e )s is a&gH seen as #o&o)re*, as possessing vis)a&
()a&ities% that what is $elt as a rea& thing o)tsi*e )s is a&so "e&t as har* or so"t, s!ooth or ro)gh2s)r"a#e*2as possessing ta#ti&e
()a&itieso The rea& rea&is! o" #o!!on sense, then, *oes in*ee* #re*it physi#a& things with vis)a&
an* ta#ti&e properties% )t it *oes so not in the spirit o" a notion o"
95 Ii*. <0.
I> Ii*. <9.
representative per#eption, )t in the spirit o" a notion o" *ire#t or i!!e*iate per#eption.
4a#$ie's version o" #o!!on2sense rea&is! is, then, I !aintain, a *istortion o" the a#t)a& pre2theoreti#a& rea&is! o" #o!!on
sense, a *istortion whi#h wrong&y assi!i&ates it, in a ")n*a!enta& respe#t, 9= the 3o#$ian rea&is! he espo)ses. I *o not "in* any
#o!para&e *istortion in Ayer's version. Re apt&y *es#ries the physi#a& o.e#ts we see! to o)rse&ves, an* ta$e o)rse&ves, to
per#eive as 'vis)ota#t)a& #ontin)ants'. The s#he!e as he presents it a&&ows "ar the *istin#tion etween these ite!s an* the
e-perien#es o" per#eiving the! an* "ar the #a)saI *epen*en#e o" the &attee on the "or!er% an* *oes so, as "ar as I #an see, witho)t
intro*)#ing the a&ien "eat)res I ave *is#erne* in 4a#$ie's a##o)nt. It is perhaps *eata&e whether Ayer #an #onsistent&y
!aintain the s#he!e's "ree*o! "ro! s)#h a&ien e&e!ents whi&e #ontin)ing to represent it as having the stat)s o" a 'theory' in
re&ation to the '*ata' o" sensi&e e-perien#e. B)t, having a&rea*y set o)t !y o.e#tions to that *o#trine, I sha&& not p)rs)e the
point. .
So!ething !ore !)st e sai*, however, ao)t the position, in the #o!!on2sense s#he!e, o" the #a)saI re&ation etween
physi#a& o.e#t an* the e-perien#e o" per#eiving it. A&tho)gh Ayer a*!its the re&ation to a pIa#e in the s#he!e, he see!s to
regar* it as a so!ewhat sophisti#ate* a**ition to the &attee, a &ate#o!er, as it were, "ar whi#h roo! has to e !a*e in an a&rea*y
sett&e* arrange!ent.90 This see!s to !e wrong. The i*ea o" the presen#e o" the thing as a##o)nting "ar, or eing responsi&e "ar,
o)r per#ept)a& awareness o" it is i!p&i#it in the pre2theoreti#a& s#he!e "ro! the very start. For we thin$ o" per#eption as a way,
in*ee* the asi# way, o" in"or!ing o)rse&ves ao)t the wor&* o" in*epen*ent&y e-isting thingsC we ass)!e, that is to say, the
generaI re&iai&ity o" O)r per#ept)a& e-perien#es% an* that ass)!ption is the sa!e as the ass)!ption o" a generaI #a)saI
*epen*en#e o" o)r per#ept)a& e-perien#es on the in*epen*ent&y e-isting things we ta$e the! to e o". The tho)ght o" !y "&eeting
per#eption as a perception o" a #ontin)o)s&y an* in*epen*ent&y e-isting thing i!p&i#it&y #ontains the tho)ght that i" the thing ha*
not een there, I sho)&* not even have seemed to per#eive it. It rea&&y sho)&* e ovio)s that with the *Ist!#tion etween
in*epen*ent&y e-isting o.e#ts an* per#ept)a& awareness o" o.e#ts we a&rea*y ave the generaI notion o" #a)saI *epen*en#e o"
the &attee on the "or!er, even i" this is not a !atter
90 Ayer, Centrai Questions! ?;2?.
to whi#h we give !)#h re"&e#tive attention in o)r pre2theoreti#a& *ays.
Two things see! to ave i!pe*e* re#ognition o" this point. One is the "a#t that the #orre#tness o" the *es#ription o" a
per#ept)a& e-perien#e as the per#eption o" a #ertain physi#a& thing lo)ical*y re()ires the e-isten#e o" that thing% an* the lo)ical is
tho)ght to e-#&)*e the causai #onne#tion, sin#e on&y 9=gi#a&Iy *istin#t e-isten#es #an e #a)sa&&y re&ate*. This is noi a serio)s
*i""i#)&ty. The sit)ation has !any para&&eis. 6ion wo)&* noi e the historian o" the *e#&ine an* "a&& o" the Ro!an E!pire )n&ess
there ha* o##)rre* so!e a#t)a& se()en#e o" events !ore or &ess Correspon*ing to is narrative. B)t it is noi eno)gh, "ar hi! to
!erit that *es#ription, that s)#h a se()en#e o" events sho)&* ave o##)rre* an* he sho)&* ave written the senten#es he *i* write.
For hi! to ()a&i"y as the historian o" these events, there !)si e a #a)saI #hain #onne#ting the! with the writing o" the
senten#es. Si!i&ar&y, the !e!ory o" an event's o##)rren#e *oes noi #o)nt as s)#h )n&ess it has its #a)saI origin in that event. An*
the re#ent&y !)#h #anvasse* '#a)saI theory o" re"eren#e' !ere&y #a&Is attention to another instan#e o" the #a)saI &in$ whi#h
otains etween tho)ght an* in*epen*ent&y 7an* anterior&y8 e-isting thing when the "or!er is right&y sai* to ave the &atter as its
o.e#t.
The se#on* i!pe*i!ent is s&ight&y !ore s)t&e. We a"e phi&osophi#a&Iy a##)sto!e*2it is a @)!ean &ega#y2to thin$ing o" the
si!p&est an* !ogi ovio)s $in* o" #a)saI re&ation as ho&*ing etween types o" ite! s)#h that ite!s o" oth types a"e oserva&e
or e-perien#ea&e an* s)#h that oservation or e-perien#e o" either ter! o" the re&ation is *istin#t "ro! oservation or e-perien#e
o" the otherC that is, the #a)sa&&y re&ate* ite!s a"e noi on&y *istin#t e-isten#es, )i a&gH the o.e#ts o" *istin#t oservations or
e-perien#es. We !ay then #o!e to thin$ o" these #on*itions as #onstit)ting a re()ire!ent on a&I pri!itive e&ie" in #a)saI
re&ations, a re()ire!ent whi#h #o)&* e !o*i"ie* or aan*one* on&y in the interests o" theory. Sin#e we ovio)s&y #annot
*isting)ish the oservation o" a physi#a& o.e#t "ro! the e-perien#e o" oserving it2"or they a"e the sa!e thing2we sha&& then e
&e* to #on#&)*e that the i*ea o" the #a)sai *epen*en#e o" per#ept)a& e-perien#e on the per#eive* o.e#t #annot e even an
i!p&i#it pari o" o)r pretheoreti#a& s#he!e. )i !)si e at est an essentia&Iy theoreti#a& a**ition to it.
B)t the *i""i#)&ty is sp)rio)s. By *ire#ting o)r attention to #a)sai re&ations etween objects o" per#eption, we ave si!p&y een
&e* to over&oo$ the spe#ia& #hara#ter o" per#eption itse&". O" #o)rse, the re()ire!ent ho&*s "ar #a)sai re&ations etween *istin#t
o.e#ts o" per#eption% )t not "ar the re&ation etween per#eption an* its o.e#t. When + is a physi#a& o.e#t an* y is a per#eption
o" +! then + is obser#ed an* y is enjoyed. An* in ta$ing the en.oy!ent o" y to e a per#eption o" +! we ore i!p&i#it&y ta$ing it to e
#a)se* y +.
This #on#&)*es the se#on* phase o" !y arg)!ent. I ave trie* to ring o)t so!e !ain "eat)res o" the rea& rea&is! o" #o!!on
sense an* o" the asso#iate* notion o" per#eption. Fro! the stan*point o" #o!!on2sense rea&is! we ta$e o)rse&ves to e
i!!e*iate&y aware o" rea&, en*)ring physi#a& things in spa#e, things en*owe* with vis)a& an* ta#ti&e properties% an* we ta$e it "ar
grante* that these en*)ring things a"e #a)sa&&y responsi&e "ar o)r interr)pte* per#eptions o" the!. The i!!e*ia#y whi#h
#o!!on sense attri)tes to per#ept)a& awareness is in no way in#onsistent either with the *istin#tion etween per#ept)a&
e-perien#e an* thing per#eive* or with the #a)sai *epen*en#e o" the "or!er on the &attee or the e-isten#e o" other #a)sa&&y
ne#essary #on*itions o" its o##)rren#e. Neither is it in#onsistent with the o##)rren#e o" per#ept)a& !ista$e or i&&)sion2a point, &i$e
so !any others o" i!portan#e, whi#h is e-p&i#it&y !a*e y Iant. 91 Both Ayer an* 4a#$ie, e-p&i#it&y or i!p&i#it&y, a#$now&e*ge
that the #o!!on2sense s#he!e in#&)*es this ass)!ption o" i!!e*ia#y24a#$ie in a passage I ave ()ote*, Ayer in is *es#ription
o" the #o!!on2sense s#he!e. Dn"ort)nate&y, 4a#$ie's a#$now&e*ge!ent o" the "a#t is e&ie* y is *es#riing #o!!on2sense
rea&is! as representative in #hara#ter an* Ayer's a#$now&e*ge!ent o" it is p)t in *o)t y is *es#riing the #o!!on2sense
s#he!e as having the stat)s o" a theory in re&ation to sensi&e e-perien#e.
999
It is ODe thing to *es#rie the s#he!e o" #o!!on sense% it is another to s).e#t it to #riti#ai e-a!ination. This is the thir* an*
!ost *i""i#)&t paTL o" !y tas$. The !ain ()estion to e #onsi*ere*, as a&rea*y in*i#ate*, is whether we a"e rationa&&y o)n* to
aan*on, or ra*i#a&&y to !o*i"y, the s#he!e in the &ight o" s#ienti"i# $now&e*ge.
Be"ore a**ressing o)rse&ves *ire#t&y to this ()estion, it is worth stressing2in*ee*, it is essentia& to stress2the grip that #o!!on2
91 Iant, 'The Re")tation o" I*ea&is!', in Criti,ue -. Pure /eason! B5;02:.
sense non2representative rea&is! has on o)r or*inary thin$ing. It is a view o" the wor&* whi#h so thoro)gh&y per!eates o)r
#ons*o)sness that even those who a"e inte&&e#t)a&&y #onvin#e* o" its "a&sity re!ain s).e#t to its power. 4a#$ie a*!its as !)#h,
saying that even when we a"e trying to entertain a 3o#$ian or s#ienti"i# rea&is!' 'o)r &ang)age an* o)r nat)ra& ways o" thin$ing
$eep p)&&ing )s a#$' to a !ore pri!itive view. 9< Consi*er the #hara#ter o" those or*inary #on#epts o" o.e#ts on the
e!p&oy!ent o" whi#h o)r &ives, o)r transa#tions with ea#h other an* the wor&*, *epen*C o)r #on#epts o" #aages, roa*s, twee*
#oats, horses, the &ips an* hair o" the e&ove*. In )sing these ter!s we #ertain&y inten* to e ta&$ing o" in*epen*ent e-isten#es
an* we #ertain&y inten* to e ta&$ing o" i!!e*iate&y per#epti&e things, earers o" pheno!ena& 7vis)o2ta#ti&e8 properties. I"
s#ienti"i# or 3o#$ian rea&is! is #orre#t, we #annot e *oing oth at on#e% it is #on")sion or i&&)sion to s)ppose we #an I" the
things we ta&$ o" rea&&y ave pheno!ena& properties, then they #annot, on this view, e physi#a& things #ontin)o)s&y e-isting in
physi#a& spa#e. Nothing per#epti&e2I ere *rop the ()a&i"i#ation 'i!!e*iate&y', "ar !y )se o" it sho)&* now e #&ear2is a
physi#a&&y rea&, in*epen*ent e-isten#e. No two persons #an ever, in this sense, per#eive the sa!e ite!C nothing at a&I is p)&i#&y
per#epti&e.
B)t how *eep the #on")sion or the i&&)sion !)st goJ @ow ra*i#a&&y it in"e#ts o)r #on#eptsJ S)re&y we !ean y a #aage a
$in* o" thing o" whi#h !ost o" the spe#i!ens we ave en#o)ntere* ave a #hara#teristi# range o" #o&o)rs an* vis)a& shapes an*
"e&t te-t)res% an* not so!ething )noserva&e, !enta&&y represente* y a #o!p&e- o" sensi&e e-perien#es whi#h it #a)ses. The
#o!!on #ons#io)sness is not to e "oe* o"" with the #on#ession that, a"ter a&&, the physi#a& thing has2in a way2a shape. The
way in whi#h s#ienti"i# rea&is! #on#e*es a shape is a&together the wrong way "or the #o!!on #ons#io)sness. The &over who
a*!ires the #)rve o" is !istress's &ips or the &over o" ar#hite#t)re who a*!ires the &ines o" a )i&*ing ta$es hi!se&" to e
a*!iring "eat)res o" those very o.e#ts the!se&ves% )t it is the vis)a& shape, the vis)a&&y *e"ine* shape, that he a*!ires. 4a#$ie
s)ggests that there is a gen)ine resemblance etween s).e#tive representation an* o.e#tive rea&ity as "ar as shape is #on#erne*%
9; )t this s)ggestion is ()ite )na##epta&e. It !a$es no sense to spea$ o" a pheno!ena& property as
9< 4a#$ie, Problems! <?.
9; Ii*., #hs. 9 an* 5, passim.
esemb*in) a non2pheno!ena&, astra#t property s)#h as physi#a& shape is #on#eive* to e y s#ienti"i# rea&is!o The property o"
&oo$ing s()are or ro)n* #an no !ore rese!&e the property, so #on#eive*, o" eing physi#a&Iy s()are or ro)n* that the property
o" &oo$ing inte&&igent or &oo$ing i&& #an rese!&e the property o" eing inte&&igent or eing i&&. I" it see!s to !a$e sense to spea$
o" a rese!&an#e etween pheno!ena& properties an* physi#a& properties, so #on#eive*, it is on&y e#a)se we give o)rse&ves
pi#t)res 2pheno!ena& pi#tHres2o" the &atter. The rese!&an#e is with the pi#t)re, not the p&#t)re*.
So, then, the #o!!on #ons#io)sness &ives, or has the i&&)sion o" &iving, in a pheno!ena&Iy propertie* wor&* o" per#epti&e
things in spa#e. We !ight #a&I it the &ive* wor&*. It is a&so the p)&i# wor&*, a##essi&e to oservation y a&IC the wor&* in whi#h
ODe !an, "o&&owing another's pointing "inger, #an see the very thing that the other sees. 7Even in o)r phi&osophi#a& !o!ents we
hait)a&Iy #ontrast the #o&o)rs an* vis)a& shapes o" things, as eing p)&i#&y oserva&e, with the s).e#tive #ontents o"
#ons#io)sness, private to ea#h o" )s, tho)gh not therey )n$nowa&e to others.8
S)#h a re!in*er o" the *epth an* rea&ity o" o)r hait)a& #o!!it!ent to the #o!!on2sense s#he!e *oes not, y itse&", a!o)nt
to a *e!onstration o" that s#he!e's i!!)nity "ro! phi&osophi#a& #riti#is!. The s#ienti"i# rea&ist, tho)gh no Iantian, !ay e
rea*y, y way o" !a$ing is !a-i!)! #on#ession, with a rep&y !o*e&&e* on Iant's #o!ination o" e!piri#a& rea&is! with
trans#en*enta& i*ea&is!o @e !ay *isting)ish etween the )n#riti#a& stan*point o" or*inary &iving an* the #riti#aI stan*point o"
phi&osophy in"or!e* y s#ien#e. We a"e h)!an&y, or nat)ra&&y2he !ay say2#onstraine* to 'see the wor&*' in ODe way 7that is, to
thin$ o" it as we see! to per#eive it8 an* rationa&&y, or #riti#a&&y, #onstraine* to thin$ o" it in ()ite another. The "irst way 7eing
itse&" a #a)saI pro*)#t o" physi#a& rea&ity8 has a $in* o" va&i*ity at its own &eve&% )t it is, #riti#a&Iy an* rationa&&y spea$ing, an
in"erior &eveJ. The se#on* way rea&Iy is a #orre#tion o" the "irst.
The a)thenti#a&&y Iantian #o!ination is open to o.e#tion in !any ways% )t, y reason o" its very e-travagan#e, it es#apes
ODe spe#i"i# Cor! o" *i""i#)&ty to whi#h the s#ienti"i# rea&ist's soerer variant re!ains e-pose*. Iant )n#o!pro!ising&y
*e#&ares that spa#e is in )s% that it is 'so&e&y "ro! the h)!an stan*point that we #an spea$ o" spa#e, o" e-ten*e* things et#.', 9?
that things as they
9? Iant, 'Re")tation o" I*ea&is!', in Criti,ue! B05.
a"e in the!se&ves a"e not spatia& at a&I. This wi&& not *o "ar the s#ienti"i# rea&ist. The pheno!ena&&y propertie* ite!s whi#h we
ta$e o)rse&ves to per#eive an* the apparent re&ations etween whi#h yie&* 7or #ontri)te vita&&y to yie&*ing8 o)r notion o" spa#e,
a"e in*ee* *e#&are* to ave no in*epen*ent rea&ity% )t, when they a"e anishe* "ro! the rea&! o" the rea&, they a"e s)ppose* to
&eave ehin* the!2as o##)pants, so to spea$, o" the eva#)ate* territory 2those spatia&&y re&ate* ite!s whi#h, tho)gh ne#essari&y
)noserva&e, neverthe&ess #onstit)te the who&e o" physi#a& rea&ity. Ayer re"ers in severa& p&a#es to this #onse()en#e% an*
()estions its #oheren#e.9: @e writes, "ar e-a!p&e, 'I *o)t whether the notion o" a spatia& syste! o" whi#h none o" the e&e!ents
#an e oserve* is even inte&&igi&e.'
It is not #&ear that this *i""i#)&ty is ins)pera&e. The s#ienti"i# rea&ist wi&& #&ai! to e a&e to astra#t the notion o" a position in
physi#a& spa#e "ro! the pheno!ena& integ)!ents with whi#h it is origina&&y an* *e#eptive&y asso#iate*% an* it is har* to thin$ o"
a #on#&)sive reason "ar *enying hi! this power. @e wi&& say that the p&a#es where the pheno!ena&&y propertie* things we see! to
per#eive see! to e a"e, o"ten eno)gh, p&a#es at whi#h the #orre&ate* physi#a&&y rea& ite!s rea&&y a"e. S)#h a #&ai! !ay !a$e )s
)neasy% )t it is not ovio)s nonsense.
Sti&I, to say that a *i""i#)&ty is not #&ear&y ins)pera&e is not to say that it is #&ear&y not ins)pera&e. It wo)&* e etter to avoi*
it i" we #an We #annot avoi* it i" we e!ra#e )na*)&terate* s#ienti"i# rea&is! an* in#i*enta&&y anno)n#e o)rse&ves therey as the
s)""erers "ro! persistent i&&)sion, however nat)ra&. We #an avoi* it, perhaps, i" we #an s)##ee* in #o!ining e&e!ents o" the
s#ienti"i# story with o)r #o!!on2sense s#he!e witho)t *owngra*ing the &atter. This is the #o)rse that Ayer re#o!!en*s,5= an*,
9 s)spe#t, the #o)rse that !ost o" )s se!i2re"&e#tive&y "o&&ow. The ()estion is whether it is a #onsistent or #oherent #o)rse. An* at
otto! this ()estion is ODe o" i*entity. Can we #oherent&y i*enti"y the pheno!ena&&y propertie*, i!!e*iate&y per#epti&e things
whi#h #o!!on sense s)pposes to o##)py physi#a& spa#e with the #on"ig)rations o" )noserva&e )&ti!ate parti#)&ars y whi#h
an )n()a&i"ie* s#ienti"i# rea&is! p)rports to rep&a#e the!B
9 approa#h the ()estion in*ire#t&y, y #onsi*ering on#e again 4a#$ie's version o" #o!!on2sense rea&is!o A##or*ing to this
9: Ayer, Centrai Questions! ?0, ?<2;, 99=. 5= Ii*. 99=299.
version, it wi&& e re!e!ere*, physi#a& things, tho)gh not *ire#t&y per#eive*, rea&Iy possess vis)a& an* ta#ti&e ()a&ities whi#h
rese!&e thoSe we see! to per#eive the! as possessing% so that i", per impossibile! the vei& o" per#eption were *rawn asi*e an*
we saw things in their tr)e #o&o)rs, these wo)&* t)ro o)t to e #o&o)rs in*ee* an*, on the who&e, .)st the #o&o)rs with whi#h we
were naive&y in#&ine* to #re*it the!. 4a#$ie *oes not represent this view as as)r* or in#oherent. @e .)st thin$s that it is, as a
!atter o" "a#t, "a&se. Things could rea&Iy e #o&o)re*% )t, sin#e there is no s#ienti"i# reason "or s)pposing they a"e, it is
grat)ito)s to !a$e any s)#h s)pposition.
4a#$ie is s)re&y too &enient to is version o" #o!!on2sense rea&is!o That version e""e#ts a #o!p&ete &ogi#a& *ivor#e etween
a thing's eing re* an* its eing re*2Ioo$ing. A&tho)gh it is a part o" the theory that a thing whi#h is, in itse&", re* has the power
to #a)se )s to see! to see a re* thing, the &ogi#a& *ivor#e etween these two properties is aso&)te. An*, as "ar as I #an see, that
*ivor#e rea&Iy pro*)#es nonsense. The as#ription o" #o&o)rs to things e#o!es not !ere&y grat)ito)s, )t sense&ess. Whatever
!ay e the #ase with shape an* position, #o&o)rs a"e visii&ia or they a"e nothing. I ave a&rea*y pointe* o)t that this version o"
#o!!on2sense rea&is! is not the rea& rea&is! o" #o!!on senseC that rea&is! e""e#ts no &ogi#a& *ivor#e etween eing re* an*
eing re*2Ioo$ing% "or it is a per#ept)a&Iy *ire#t an* not a per#ept)a&Iy representative rea&is!o The things seen as #o&o)re* a"e
the things the!se&ves. There is no 'vei& past whi#h we #annot see'% "or there is no vei&.
B)t this *oes not !ean that a thing whi#h is re*, that is, re*&oo$ing, has to &oo$ re* a&I the &i!e an* in a&I #ir#)!stan#es an* to
a&I oservers. There is an irre*)#i&e re&ativity, a re&ativity to what in the roa*est sense !ay e #a&&e* the per#ept)a& point o"
view, )i&t in to o)r as#riptions o" parti#)&ar vis)a& properties to things. The !o)ntains a"e re*2Ioo$ing at this *istan#e in this
Jight% &)e&oo$ing at that *istan#e at that &ight% an*, when we a"e #&a!ering )p the!, perhaps neither. S)#h2an*2s)#h a s)r"a#e
&oo$s pin$ an* s!ooth "ro! a *istan#e% !ott&e* an* grainy when #&ose&y e-a!ine*% *i""erent again, perhaps, )n*er the
!i#ros#ope.
We asor this re&ativity easi&y eno)gh "or or*inary p)rposes in o)r or*inary ta&$, ta#it&y ta$ing so!e range o" per#ept)a& #on2
*itions, so!e per#ept)a& point o" view 7in the roa* sense8 as stan*ar* or nor!a&, an* intro*)#ing an e-p&i#it a#$now&e*ge!ent
o" re&ativity on&y in #ases whi#h *eviate "ro! the stan*ar*. 'It &oo$s p)rp&e in this &ight,' we say, ')t ta$e it to the *oor an* yo)
wi&& see that it's rea&&y green.' B)t so!eti!es we *o so!ething e&se. We shi"t the stan*ar*. 4agni"ie*, the "ari# appears as
printe*' with tiny &)e an* ye&&ow *ots. So those a"e the #o&o)rs it rea&&y is. Does this as#ription #ontra*i#t 'it's rea&&y green'B No%
"ar the stan*ar* has shi"te*. 3oo$ing at photographs, in .o)rna&s o" pop)&ar s#ien#e, o" pat#hes o" h)!an s$in, vast&y !agni"ie*,
we say, '@ow "antasti#a&&y )neven an* ri*gy it rea&&y is.' We st)*y a sa!pie o" &oo* thro)gh a !i#ros#ope an* say, 'It's !ost&y
#o&o)r&ess.' B)t s$in #an sti&I e s!ooth an* &oo* e re*% "ar in another #onte-t We shi"t o)r stan*ar* a#$. S)#h shi"ts *o not
#onvi#t )s o" vo&ati&ity or #on*e!n )s to internaI #on"&i#t. The appearan#e o" oth vo&ati&ity an* #on"&i#t vanishes when we
a#$now&e*ge the re&ativity o" O)r 'rea&&ys' .
4y e-a!p&es a"e anaI. B)t perhaps they s)ggest a way o" reso&ving the apparent #on"&i#t etween s#ienti"i# an* #o!!on2
sense rea&is!o We #an shi"t o)r point o" view within the generaI "ra!ewor$ o" per#eption, whether ai*e* or )nai*e* y arti"i#iai
!eans% an* the *i""erent sensi&e2()a&ity as#riptions we then !a$e to the sa!e o.e#t a"e not seen as #on"&i#ting on#e their
re&ativity is re#ogniAe*. Can we not see the a*option o" the viewpoint o" s#ienti"i# rea&is! as si!p&y a !ore ra*i#a& shi"t2a shi"t to
a viewpoint "ro! whi#h no #hara#teristi#s a"e to e as#rie* to things e-#ept those whi#h "ig)re in the physi#a& theories o"
s#ien#e an* in 't# e-p&anation o" what goes on in the physi#a& wor&* in the pro#esses whi#h &ea* to o)r having the sensations an*
per#eptions that we have'B59 We #an say that this is how things rea&&y a"e so &ong as the re&ativity o" this 'rea&&y' is re#ogniAe* as
we&&% an*, when it is re#ogniAe*, the s#ienti"i# a##o)nt wi&I no !ore #on"&i#t with the as#ription to things o" vis)a& an* ta#ti&e
()a&ities than the assertion that io o* is rea&&y a !ain&y #o&o)r&ess "&)i* #on"&i#ts with the assertion that it is right re* in #o&o)r.
O" #o)rse, the s#ienti"i# point o" view is not, in ODe sense, a point o" #ie" at a&I. It is an inte&&e#t)a&, not a per#ept)aI, stan*point.
We #o)&* not o##)py it at a&&, *i* we not "irst o##)py the other. B)t we #an per"e#t&y we&& o##)py oth at on#e, so &ong as we
rea&iAe what we a"e *oing.
This !etho* o" re#on#i&ing s#ienti"i# an* #o!!on2sense rea&is! re()ires )s to re#ogniAe a #ertain re&ativity in o)r #on#eption
o" the rea& properties o" physi#a& o.e#ts. Re&ative to the h)!an per#ept)a& stan*point the grosser physi#a& o.e#ts a"e vis)o2
ta#ti&e
59 4a#$ie, Probiems! 18.
#ontin)ants 7an* within that stan*point the pheno!ena& properties they possess a"e re&ative to parti#)&ar per#ept)a& viewpoints,
ta$en as stan*ar*8. Re&ative to the s#ienti"i# stan*point, they ave no properties )t those whi#h "ig)re in the physi#a& theories o"
s#ien#e.
S)#h a re&ativisti# #on#eption wi&& not p&ease the aso&)te2!in*e*. Ayer re#o!!en*s a *i""erent pro#e*)re. @e s)ggests that we
sho)&* #on#eive o" per#epti&e o.e#ts 7that is, o.e#ts per#epti&e in the sense o" the #o!!on2sense s#he!e8 as eing &itera&&y
#o!pose* o" the )&ti!ate parti#&es o" physi#a& theory, the &atter eing i!per#epti&e, not in prin#ip&e, )t on&y e!piri#a&&y, as a
#onse()en#e o" their eing so !in)te. 55 I *o)t, however, whether this proposa&, whi#h Ayer right&y *es#ries as an atte!pt to
blend the two s#he!es, #an e regar*e* as satis"a#tory. I" the i!possii&ity o" per#eiving the )&ti!ate #o!ponents is to e viewe*
as !ere&y e!piri#a&, we #an sensi&y as$ what the #on#ept)a& #onse()en#es wo)&* e o" s)pposing that i!possii&ity not to e-ist.
The answer is #&ear. Even i" there were so!ething whi#h we #o)nte* as per#eiving the )&ti!ate parti#&es, this wo)&* sti&I not,
"ro! the point o" view o" s#ienti"i# rea&is!, #o)nt as per#eiving the! as they rea&&y a"e. An* nothing #o)&* so #o)nt% "ar no
pheno!ena& properties we see!e* to per#eive the! as having wo)&* "ig)re in the physi#a& e-p&anation o" the #a)saI !e#hanis!s
o" o)r s)##esso B)t, so &ong as we stay at this point o" view, what goes "ar the parts goes "ar any who&es they #o!pose. @owever
gross those who&es, they re!ain, "ro! this point o" view, i!per#epti&e in the sense o" #o!!on sense.
Ayer atte!pts to "or! one viewpoint o)t o" two *is#repant viewpointsC to "or! a sing&e, )ni"ie* *es#ription o" physi#a& rea&ity
y &en*ing "eat)res o" two *is#repant *es#riptions, ea#h va&i* "ro! its own viewpoint. @e #an see! to s)##ee* on&y y *oing
vio&en#e to one o" the two viewpoints, the s#ienti"i#o I a#$now&e*ge the *is#repan#y o" the two *es#riptions, )t #&ai! that, on#e
we re#ogniAe the re&ativity in o)r #on#eption o" the rea&, they nee* not e seen as in #ontra*i#tion with ea#h other. Those very
things whi#h "ro! one stan*point we #on#eive as pheno!ena&&y propertie* we #on#eive "ro! another as #onstit)te* in a way
whi#h #Hn on&y e *es#rie* in what a"e, "ro! the pheno!ena& point o" M&ew, astra#t ter!s. 'This s!ooth, green, &eather ta&e2
top', we say, 'is, #onsi*ere* s#ienti"i#a&&y, nothing )t a #ongeries o" e&e#tri# #harges wi*e&y separate* an* in rapi* !otion.' Th)s
we #o!ine the two stan*points in a sing&e senten#e. The stan*point o"
55 Ayer, CentraI Questions! 99=.
#o!!on2sense rea&is!, noi e-p&i#it9y signa&&e* as s)#h, is re"&e#te* in the senten#e's gra!!ati#a& s).e#t phrase, o" whi#h the
wor*s a"e e!p&oye* in no esoteri# sense. The stan*point o" physi#a& s#ien#e, e-p&i#it9y signa&&e* as s)#h, is re"&e#te* in the
pre*i#ate. On#e re&ativity o" *es#ription to stan*point is re#ogniAe*, the senten#e is seen to #ontain no #ontra*i#tion% an*, i" it
#ontains no #ontra*i#tion, the pro&e! o" i*enti"i#ation is so&ve*.
I re#ogniAe that this position is )n&i$e&y to satis"y the *eter!ine* s#ienti"i# rea&ist. I" he is on&y !o*erate&y *eter!ine*, he !ay
e partia&Iy satis"ie*, an* !ay #ontent hi!se&" with saying that the s#ienti"i# viewpoint is superior to that o" #o!!on sense. Re
wi&& then si!p&y e e-pressing a pre"eren#e, whi#h he wi&& noi e-pe#t the artist, "or e-a!p&e, to share. B)t, i" he is a har*2&iner, he
wi&& insist that the #o!!on2sense view is who&Iy )n*er!ine* y s#ien#e% that it is shown to e "a&se% that the vis)a& an* ta#ti&e
properties we as#rie to things a"e nowhere )i in o)r !in*s% that we *o noi &ive in a wor&* o" per#epti&e o.e#ts, as )n*erstoo*
y #o!!on sense, at a&I. @e !)si then a##ept the #onse()en#e that ea#h o" )s is a s)""erer "ro! a persistent an* ines#apa&e
i&&)sion an* that it is "ort)nate that this is so, sin#e, i" it were noi, we sho)&* e )na&e to p)rs)e the s#ienti"i# enterprise itse&".
Witho)t the i&&)sion o" per#eiv!g o.e#ts as earers o" sensi&e ()a&ities, we sho)&* noi ave the i&&)sion o" per#eiving the! as
spa#e2o##)piers at a&I% an* witho)t that we sho)&* ave no #on#epL o" spa#e an* no power to p)rs)e o)r resear#hes into the
nat)re o" its o##)pants. S#ien#e is noi on&y the o""spring o" #o!!on sense% it re!ains its *epen*ant. For this reason, an* "or
others to)#he* on ear&ier, the s#ienti"i# rea&ist !)si, however r)e")&Iy, a*!it that the as#ription to o.e#ts o" sensi&e ()a&ities,
the stan*ar* o" #orre#tness o" s)#h as#ription eing 7what we ta$e to e8 inters).e#tive agree!ent, is so!ething ()ite se#)re&y
roote* in o)r #on#ept)a& s#he!e. I" this !eans, as he !)si !aintain it *oes, that o)r tho)ght is #on*e!ne* to in#oheren#e, then
we #an on&y #on#&)*e that in#oheren#e is so!ething we #an per"e#t9y we&I &ive with an* #o)&* noi per"e#t9y we&I&ive witho)t.