/LMAR=0/XLINE=3/FONT#0=BASL30/FONT#1=BASI30/FONT#2=BASB30/FONT#10=BAXM30/FONT#11=ZERO30/FONT#3=STA200/FONT#4=NGB25
␈↓ ↓H␈↓␈↓βS␈↓∧ Artificial Intelligence Laboratory, STANFORD UNIVERSITY, Stanford, California 94305

␈↓ ↓H␈↓∧Telephone 415 497-4430␈↓ 
!April 6, 1979 




␈↓ ↓H␈↓Professor Hilary Putnam
␈↓ ↓H␈↓Department of Philosophy
␈↓ ↓H␈↓Harvard University
␈↓ ↓H␈↓Cambridge, Massachusetts 02138

␈↓ ↓H␈↓Dear Hilary:

␈↓ ↓H␈↓        I␈α∩found␈α∪your␈α∩lecture␈α∩quite␈α∪stimulating,␈α∩although␈α∩I␈α∪have␈α∩to␈α∩confess␈α∪that␈α∩mainly␈α∪I␈α∩was
␈↓ ↓H␈↓stimulated␈α
to␈α
disagree.␈α I␈α
hope␈α
you␈αwon't␈α
think␈α
I␈αalways␈α
disagree;␈α
indeed␈αI␈α
found␈α
much␈α
to␈αagree
␈↓ ↓H␈↓with␈α⊂in␈α⊂some␈α⊂of␈α⊂your␈α⊂earlier␈α⊂papers␈α⊃to␈α⊂which␈α⊂I␈α⊂referred␈α⊂in␈α⊂my␈α⊂␈↓↓Ascribing␈α⊂Mental␈α⊃Qualities␈α⊂to
␈↓ ↓H␈↓↓Machines␈↓.

␈↓ ↓H␈↓        First,␈αand␈αthis␈αmay␈αre∨ect␈αonly␈αmy␈αphilosophical␈αinexperience,␈αI␈αwasn't␈αconvinced␈αthat␈αthere
␈↓ ↓H␈↓was␈α
a␈α
real␈α
problem␈α
about␈αreference.␈α
 There␈α
is␈α
a␈α
relation␈α
between␈αthe␈α
outline␈α
drawn␈α
by␈α
the␈αant␈α
and
␈↓ ↓H␈↓President␈α
Carter,␈α
and␈α
it␈α
is␈α
legitimate␈α
to␈α
study␈α
the␈α
relation␈α
between␈α
pictures␈α
and␈α
the␈α
objects␈α
they
␈↓ ↓H␈↓look␈αlike␈α
independent␈αof␈αanyone's␈α
intention.␈α Such␈αa␈α
theory␈αwould␈αtell␈α
us␈αwhen␈αa␈α
cloud␈αor␈αink␈α
blot
␈↓ ↓H␈↓or rock formation or pattern of chicken entrails would be found to be a picture.

␈↓ ↓H␈↓        Likewise,␈α∪there␈α∀is␈α∪another␈α∀theory␈α∪of␈α∪systems␈α∀of␈α∪representation␈α∀that␈α∪would␈α∀cover␈α∪both
␈↓ ↓H␈↓linguistic␈αand␈αpictorial␈αrepresentations.␈α
 Such␈αsystems␈αmight␈αbe␈α
used␈αby␈αeither␈αpeople␈αor␈α
machines.
␈↓ ↓H␈↓Many␈α∞questions␈α∞about␈α∂whether␈α∞an␈α∞alleged␈α∂symbol␈α∞represents␈α∞something␈α∂may␈α∞be␈α∞better␈α∂asked␈α∞in
␈↓ ↓H␈↓connection with a speci≡c system of representation rather than as intrinsic questions.

␈↓ ↓H␈↓        It␈α
seems␈α
to␈α∞me␈α
that␈α
the␈α
possibility␈α∞of␈α
asking␈α
what␈α
a␈α∞symbol␈α
represents␈α
in␈α
isolation␈α∞from␈α
a
␈↓ ↓H␈↓particular␈α
system␈α
of␈α∞reference␈α
is␈α
that␈α
it␈α∞almost␈α
never␈α
turns␈α
out␈α∞that␈α
a␈α
large␈α
sequence␈α∞of␈α
symbols
␈↓ ↓H␈↓admits␈αmore␈αthan␈αone␈αinterpretation.␈α This␈αis␈αsuggested␈αby␈αexperience␈αwith␈αcryptography.␈α So␈αfar
␈↓ ↓H␈↓as␈α⊃I␈α⊃know,␈α⊃there␈α∩are␈α⊃no␈α⊃known␈α⊃examples␈α⊃of␈α∩cryptograms␈α⊃that␈α⊃admit␈α⊃more␈α⊃than␈α∩one␈α⊃lengthy
␈↓ ↓H␈↓interpretation␈α∞in␈α∞ciphers␈α∞with␈α∞key␈α∞length␈α∞substantially␈α∞shorter␈α∞than␈α∞the␈α∞message.␈α∂ Although␈α∞you
␈↓ ↓H␈↓didn't␈αspell␈αout␈αyour␈αexample␈αof␈αsomething␈αthat␈αcould␈αbe␈αinterpreted␈αas␈αrepresenting␈αeither␈αprime
␈↓ ↓H␈↓numbers␈αor␈αapple␈αtrees,␈αI␈αdoubt␈αthat␈αthere␈αis␈αsuch␈αan␈αexample␈αin␈αany␈αnon-trivial␈αsense.␈α If␈αthis␈αis
␈↓ ↓H␈↓right,␈αhowever,␈αthere␈αwon't␈α
be␈αany␈αgood␈αsimple␈αexamples,␈α
because␈αthe␈αuniqueness␈αof␈αreference␈α
will
␈↓ ↓H␈↓depend␈αon␈αhaving␈αa␈αlarge␈αamount␈αof␈αdata.␈α Therefore,␈αit␈αseems␈αto␈αme␈αthat␈αthe␈αtheory␈αof␈αreference
␈↓ ↓H␈↓systems will be more fundamental and much more tractable.

␈↓ ↓H␈↓        Maybe␈αthere␈αare␈αeven␈αmore␈αways␈αof␈αlooking␈αat␈αrepresentation.␈α It␈αseems␈αunlikely␈α
that␈αthere
␈↓ ↓H␈↓will␈α
be␈α
a␈α
single␈α∞true␈α
concept␈α
of␈α
representation.␈α
 Moreover,␈α∞it␈α
seems␈α
unlikely␈α
that␈α∞examination␈α
of
␈↓ ↓H␈↓ordinary␈α∃language␈α∃will␈α∃dictate␈α∃preferring␈α∃one␈α⊗of␈α∃them,␈α∃especially␈α∃if␈α∃it␈α∃turns␈α∃out␈α⊗that␈α∃the
␈↓ ↓H␈↓di≥erentiating examples are outside of ordinary experience.

␈↓ ↓H␈↓        The␈αproblem␈αof␈αthe␈α␈↓↓vat␈αmen␈↓␈αbrings␈αup␈αmatters␈αto␈αwhich␈αI␈αhave␈αgiven␈αconsiderable␈αthought
␈↓ ↓H␈↓αProfessor Hilary Putnam␈↓ ¬uApril 6, 1979␈↓ 
nPage 2␈↓ 

␈↓ ↓H␈↓though␈α∞mainly␈α∞of␈α∞a␈α∞science␈α∞≡ction␈α∞character.␈α
 Might␈α∞we␈α∞be␈α∞simulations␈α∞in␈α∞some␈α∞giant␈α
computer,
␈↓ ↓H␈↓and could we investigate the question?

␈↓ ↓H␈↓        It␈αseems␈αto␈αme␈αthat␈αthe␈αperfect␈αsimulation␈αis,␈αas␈αyou␈αmentioned,␈αa␈αlimiting␈αcase␈αof␈αimperfect
␈↓ ↓H␈↓simulation.␈α⊃ Moreover,␈α⊃while␈α⊃investigating␈α⊃the␈α∩matter,␈α⊃the␈α⊃alleged␈α⊃simulee␈α⊃might␈α⊃say␈α∩that␈α⊃the
␈↓ ↓H␈↓simulation,␈α
if␈α
such,␈α
seems␈αperfect␈α
so␈α
far,␈α
but␈α
he␈αstill␈α
hopes␈α
for␈α
a␈α∨aw.␈α
 One␈α
∨aw␈α
I␈α
imagined␈αwas
␈↓ ↓H␈↓that␈α
a␈αdigital␈α
simulation␈αof␈α
di≥erential␈αequations␈α
might␈α
show␈αround-o≥␈α
error,␈αand␈α
that␈αsuch␈α
errors
␈↓ ↓H␈↓would␈α⊂seem␈α⊂random␈α⊂when␈α⊃looked␈α⊂at␈α⊂super≡cially,␈α⊂but␈α⊃would␈α⊂betray␈α⊂their␈α⊂digital␈α⊃nature␈α⊂when
␈↓ ↓H␈↓examined␈α
more␈α
closely,␈α
i.e.␈α
the␈α
di≥erent␈α
errors␈α
would␈α
show␈α
correlations␈α
and␈α
other␈α
forms␈α∞of␈α
non-
␈↓ ↓H␈↓independence.␈α∞ My␈α
fantasy␈α∞went␈α∞on␈α
to␈α∞suppose␈α∞that␈α
the␈α∞program␈α∞was␈α
huge,␈α∞i.e.␈α∞corresponded␈α
to
␈↓ ↓H␈↓millions␈α∞of␈α∞years␈α∞of␈α∞human␈α∞work,␈α∂and␈α∞embodied␈α∞billions␈α∞of␈α∞arbitrary␈α∞decisions␈α∂among␈α∞di≥erent
␈↓ ↓H␈↓ways␈α∂of␈α∂accomplishing␈α⊂the␈α∂same␈α∂purpose,␈α∂each␈α⊂of␈α∂which␈α∂gave␈α∂a␈α⊂clue␈α∂to␈α∂the␈α∂psychology␈α⊂of␈α∂the
␈↓ ↓H␈↓Programmer.␈α∞ From␈α∞this,␈α∂again␈α∞using␈α∞cryptography,␈α∞we␈α∂might␈α∞conjecture␈α∞properties␈α∞of␈α∂the␈α∞"real
␈↓ ↓H␈↓world" and how we might act in the "real world".

␈↓ ↓H␈↓        While␈α∞listening␈α
to␈α∞your␈α∞lecture,␈α
another␈α∞way␈α∞of␈α
detecting␈α∞simulation␈α∞occurred␈α
to␈α∞me.␈α∞ It␈α
is
␈↓ ↓H␈↓problematical␈αwhether␈αit␈αcould␈αbe␈αcalled␈α
looking␈αfor␈αan␈αimperfection.␈α To␈αput␈αit␈α
in␈αold-fashioned
␈↓ ↓H␈↓terms,␈α
we␈α
could␈α
look␈α
for␈α
␈↓↓≡nal␈α
causes␈↓,␈α
i.e.␈α
we␈α
could␈α
attempt␈α
to␈α
make␈α
a␈α
theory␈α
that␈α
would␈α
explain
␈↓ ↓H␈↓apparently␈α∀random␈α∀connections␈α∀among␈α∀phenomena␈α∃as␈α∀consequences␈α∀of␈α∀the␈α∀purposes␈α∃of␈α∀the
␈↓ ↓H␈↓Programmer.␈α∂ Since␈α⊂the␈α∂time␈α⊂of␈α∂Galileo,␈α⊂≡nal␈α∂causes␈α∂have␈α⊂quite␈α∂properly␈α⊂gone␈α∂out␈α⊂of␈α∂fashion.
␈↓ ↓H␈↓However, they still have a small domain of legitimacy.

␈↓ ↓H␈↓        When␈α
we␈α≡nd␈α
that␈α
an␈αindividual␈α
or␈αgroup␈α
in␈α
history␈αperformed␈α
some␈αaction␈α
or␈α
had␈αsome
␈↓ ↓H␈↓custom,␈αit␈αis␈αoften␈αhelpful␈αto␈αask␈αwhy␈αthey␈αdid␈αit␈αthat␈αway.␈α When␈αwe␈αsee␈αan␈αorgan␈αon␈αan␈αanimal,
␈↓ ↓H␈↓such␈α
as␈α
the␈α
pit␈α∞of␈α
a␈α
pit␈α
viper␈α∞or␈α
the␈α
comb␈α
of␈α
a␈α∞cock,␈α
it␈α
is␈α
often␈α∞helpful␈α
to␈α
ask␈α
what␈α∞function␈α
it
␈↓ ↓H␈↓might␈αperform.␈α We␈αmay␈αor␈αmay␈αnot␈αget␈αa␈αuseful␈αanswer.␈α On␈αthe␈αother␈αhand,␈αit␈αseems␈αunfruitful
␈↓ ↓H␈↓to␈α
ask␈α∞the␈α
purpose␈α
of␈α∞a␈α
range␈α
of␈α∞mountains␈α
along␈α
the␈α∞west␈α
coast␈α
of␈α∞the␈α
American␈α∞continent.␈α
 It
␈↓ ↓H␈↓seems␈α∂to␈α∂me␈α∂that␈α∂the␈α∂inutility␈α∂of␈α∂asking␈α∂for␈α∂≡nal␈α∂causes␈α∂outside␈α∂the␈α∂limited␈α∂domain␈α∂of␈α∂human
␈↓ ↓H␈↓purpose␈αand␈α
natural␈αselection␈α
is␈αan␈α
empirical␈αfact;␈α
it␈αwould␈α
not␈αhave␈α
been␈αlegitimate␈α
to␈αmake␈αit␈α
an
␈↓ ↓H␈↓␈↓↓a priori␈↓ philosophical principle.

␈↓ ↓H␈↓        Indeed,␈α⊃it␈α⊃is␈α⊃common␈α⊃for␈α⊃people␈α⊃to␈α⊃ascribe␈α⊃the␈α⊃course␈α⊃of␈α⊃political␈α⊃events␈α⊃to␈α⊂purposeful
␈↓ ↓H␈↓behind-the-scenes␈α
action␈α
whether␈α
it␈α
be␈α
the␈αunfolding␈α
of␈α
God's␈α
grand␈α
plan␈α
or␈α
the␈αevil␈α
machinations
␈↓ ↓H␈↓of␈αthe␈α
Vatican␈αor␈αBritish␈α
Intelligence␈αand␈α
the␈αRockefellers␈α(the␈α
latter␈αbeing␈α
the␈αcurrent␈αdoctrine␈α
of
␈↓ ↓H␈↓the U.S. Labor Party and its maximal leader).

␈↓ ↓H␈↓        From␈α
an␈αaesthetic␈α
point␈αof␈α
view,␈αit␈α
may␈α
be␈αunfortunate␈α
that␈αit␈α
is␈αunfruitful␈α
to␈α
explain␈αthe
␈↓ ↓H␈↓ant␈α∞as␈α∞an␈α∞admonishment␈α∞to␈α∞be␈α∞diligent␈α∞or␈α∞the␈α∞rainbow␈α∞as␈α∞a␈α∞sign␈α∞that␈α∞God␈α∞will␈α∞destroy␈α∞the␈α
the
␈↓ ↓H␈↓world with ≡re next time, i.e. that

␈↓ ↓H␈↓␈↓↓God gave Noah that rainbow sign.
␈↓ ↓H␈↓↓God gave Noah that rainbow sign
␈↓ ↓H␈↓↓That it won't be water
␈↓ ↓H␈↓↓But ≡re next time␈↓.

␈↓ ↓H␈↓does␈αnot␈α≡t␈αinto␈αa␈αlarger␈αtheory␈αthat␈αallows␈αsuccessful␈αprediction␈αand␈αcontrol␈αof␈αphenomena␈αother
␈↓ ↓H␈↓than the behavior of people who believe the theory.

␈↓ ↓H␈↓        Indeed␈α∞there␈α∞is␈α∞another␈α
science␈α∞≡ction␈α∞fantasy␈α∞that␈α
suggests␈α∞a␈α∞form␈α∞of␈α∞astrology.␈α
 Suppose
␈↓ ↓H␈↓that␈αthere␈α
is␈αenough␈α
matter␈αin␈αthe␈α
universe␈αso␈α
that␈αit␈αwill␈α
contract␈αagain,␈α
and␈αeverything␈α
will␈αbe
␈↓ ↓H␈↓burned␈αup␈αin␈αthe␈αcompression␈αwhich␈αwill␈αbe␈αfollowed␈αby␈αanother␈αbig␈αbang.␈α If␈αthe␈αcompression␈αis
␈↓ ↓H␈↓merely␈α~to␈α≠a␈α~high␈α≠density␈α~and␈α≠not␈α~to␈α≠an␈α~informationless␈α≠singularity,␈α~then␈α≠large␈α~scale
␈↓ ↓H␈↓inhomogeneities␈αin␈α
the␈αpresent␈α
universe␈αwill␈α
lead␈αto␈α
inhomogeneities␈αin␈α
the␈αlarge␈αscale␈α
distribution
␈↓ ↓H␈↓αProfessor Hilary Putnam␈↓ ¬uApril 6, 1979␈↓ 
nPage 3␈↓ 

␈↓ ↓H␈↓of␈α∞matter␈α∞after␈α∞the␈α∞next␈α∂big␈α∞bang.␈α∞ If␈α∞we␈α∞want␈α∂to␈α∞"survive"␈α∞the␈α∞compression,␈α∞we␈α∂might,␈α∞having
␈↓ ↓H␈↓nothing␈αbetter␈αto␈α
do␈αfor␈αa␈αfew␈α
tens␈αof␈αbillions␈α
of␈αyears,␈αencode␈αour␈α
own␈αdescriptions␈αin␈α
the␈αlarge
␈↓ ↓H␈↓scale distribution of matter in the universe, i.e. by rearranging the galaxies.

␈↓ ↓H␈↓        When␈α
life␈α
evolves␈α
after␈α
the␈α
next␈α
big␈αbang,␈α
and␈α
scientists␈α
study␈α
the␈α
large␈α
scale␈αdistribution␈α
of
␈↓ ↓H␈↓matter,␈α⊃they␈α∩will␈α⊃notice␈α∩that␈α⊃information␈α⊃is␈α∩encoded␈α⊃in␈α∩this␈α⊃distribution.␈α⊃ If␈α∩we␈α⊃make␈α∩a␈α⊃plot
␈↓ ↓H␈↓analogous␈α
to␈α
that␈α
of␈αFred␈α
Hoyle's␈α
␈↓↓A␈α
for␈α
Andromeda␈↓,␈αthis␈α
information␈α
would␈α
contain␈αinstructions␈α
on
␈↓ ↓H␈↓how␈αto␈αbuild␈αa␈αcomputer.␈α If␈αthey␈αfoolishly␈αbuilt␈αthis␈αcomputer␈αwithout␈αprecautions,␈αit␈αwould␈αtake
␈↓ ↓H␈↓over␈α∞their␈α∞society␈α∞and␈α∞use␈α∂its␈α∞resources␈α∞to␈α∞read␈α∞the␈α∞rest␈α∂of␈α∞the␈α∞message␈α∞and␈α∞reconstruct␈α∂us.␈α∞ Of
␈↓ ↓H␈↓course,␈α
we␈α
can␈α
ask␈α
whether␈α
there␈αis␈α
a␈α
message␈α
in␈α
the␈α
distribution␈αof␈α
galaxies␈α
or␈α
in␈α
the␈α
digits␈αof␈α
the
␈↓ ↓H␈↓dimensonless␈αconstants␈αof␈αphysics␈αleft␈αover␈αfrom␈αbefore␈αthe␈αbig␈αbang.␈α I␈αsuppose␈αthat␈αin␈α
the␈αnext
␈↓ ↓H␈↓few billion years, even such improbable possibilities as this will be checked out.

␈↓ ↓H␈↓        The␈α∪third,␈α∪and␈α∀admittedly␈α∪weak,␈α∪reason␈α∪for␈α∀considering␈α∪the␈α∪vat␈α∪hypothesis␈α∀might␈α∪be
␈↓ ↓H␈↓philosophical␈αprejudice,␈αi.e.␈αa␈α
person␈αmight␈αdislike␈αthe␈α
big␈αbang␈αhypothesis␈αenough␈αto␈α
hypothesize
␈↓ ↓H␈↓that␈α∩the␈α∪apparently␈α∩big␈α∩bang␈α∪universe␈α∩is␈α∩a␈α∪simulation␈α∩within␈α∩a␈α∪universe␈α∩that␈α∩has␈α∪gone␈α∩on
␈↓ ↓H␈↓inde≡nitely.␈α The␈αidea␈αthat␈αthe␈αreal␈αuniverse␈αis␈αa␈αHilbert␈αspace␈αor␈αother␈αin≡nite␈αdimensional␈αspace
␈↓ ↓H␈↓has many attractions for the wishful thinker.

␈↓ ↓H␈↓        Incidentally,␈α∂when␈α∂you␈α∞referred␈α∂to␈α∂me␈α∞in␈α∂connection␈α∂with␈α∞"procedural␈α∂semantics"␈α∂which␈α∞I
␈↓ ↓H␈↓took␈α∩to␈α∪correspond␈α∩to␈α∪empiricism,␈α∩I␈α∪wondered␈α∩whether␈α∪you␈α∩thought␈α∪that␈α∩e≥orts␈α∪to␈α∩program
␈↓ ↓H␈↓machines␈α∞to␈α∞behave␈α∞intelligently␈α∞were␈α∞somehow␈α∂committed␈α∞to␈α∞giving␈α∞them␈α∞an␈α∞empiricist␈α∂way␈α∞of
␈↓ ↓H␈↓looking␈α∂at␈α∂the␈α∂world,␈α⊂i.e.␈α∂having␈α∂them␈α∂attempt␈α∂to␈α⊂"construct"␈α∂their␈α∂world␈α∂as␈α⊂the␈α∂"input-output"
␈↓ ↓H␈↓relationship␈α
of␈αthe␈α
machine's␈αenvironment.␈α
 It␈α
seems␈αto␈α
me␈αthat␈α
this␈α
is␈αnot␈α
the␈αonly␈α
way␈α
or␈αeven
␈↓ ↓H␈↓the␈α∂best␈α∂way␈α∞to␈α∂program␈α∂machines.␈α∂ My␈α∞thoughts␈α∂have␈α∂always␈α∂run␈α∞towards␈α∂trying␈α∂to␈α∂specify␈α∞a
␈↓ ↓H␈↓"naive␈αrealist"␈αpoint␈αof␈αview␈αwell␈αenough␈αto␈αprogram␈αit.␈α The␈αintelligent␈αprogram␈αshould␈αbe␈αready
␈↓ ↓H␈↓to␈α
accept␈αnames␈α
as␈αidentifying␈α
unknown␈αobjects␈α
and␈αhope␈α
later␈αto␈α
learn␈αmore␈α
about␈αthem.␈α
 Ideally,
␈↓ ↓H␈↓it␈αshould␈αhave␈αthe␈αability␈αof␈αa␈αchild␈αto␈αlisten␈αto␈αa␈αpoem␈αbeginning␈α␈↓↓"The␈αlion␈αand␈αthe␈αunicorn␈αwere
␈↓ ↓H␈↓↓≡ghting␈α
for␈αthe␈α
crown"␈↓␈αand␈α
only␈αlater␈α
come␈αto␈α
realize␈α
that␈αa␈α
lion␈αis␈α
a␈αreal␈α
beast␈αand␈α
a␈α
unicorn␈αa
␈↓ ↓H␈↓mythical beast.

␈↓ ↓H␈↓        I␈α∞have␈α∞done␈α
some␈α∞thinking␈α∞about␈α∞what␈α
I␈α∞used␈α∞to␈α∞call␈α
metaphilosophy␈α∞and␈α∞which␈α∞may␈α
be
␈↓ ↓H␈↓more␈α⊃precisely␈α∩called␈α⊃metaepistemology.␈α⊃ Specify␈α∩a␈α⊃world␈α⊃as␈α∩a␈α⊃causal␈α⊃system,␈α∩e.g.␈α⊃a␈α∩system␈α⊃of
␈↓ ↓H␈↓interacting␈α∞≡nite␈α∞automata,␈α∞and␈α∞suppose␈α∞that␈α∞it␈α∞contains␈α∞a␈α∞computer␈α∞that␈α∞can␈α∞execute␈α
programs,
␈↓ ↓H␈↓and␈α⊃consider␈α∩programs␈α⊃that␈α∩"attempt␈α⊃to␈α∩discover␈α⊃the␈α∩structure␈α⊃of␈α∩their␈α⊃world".␈α∩ We␈α⊃consider
␈↓ ↓H␈↓building␈α⊂into␈α∂the␈α⊂program␈α∂various␈α⊂epistemological␈α∂methodologies,␈α⊂e.g.␈α∂empiricism.␈α⊂ It␈α∂is␈α⊂then␈α∂a
␈↓ ↓H␈↓technical␈α∪question␈α∪whether␈α∪the␈α∪program␈α∪will␈α∀indeed␈α∪discover␈α∪the␈α∪structure␈α∪of␈α∪its␈α∀world.␈α∪ If
␈↓ ↓H␈↓metaepistemology␈α⊂could␈α⊂be␈α⊂developed␈α∂as␈α⊂a␈α⊂technical␈α⊂subject,␈α⊂it␈α∂might␈α⊂have␈α⊂the␈α⊂same␈α⊂e≥ect␈α∂on
␈↓ ↓H␈↓epistemological␈αcontroversies␈αas␈αmetamathematics␈αhad␈αon␈αthe␈αcontroversies␈αabout␈α
the␈αfoundations
␈↓ ↓H␈↓of␈αmathematics.␈α It␈α
didn't␈αactually␈αdestroy␈α
the␈αcontroversies,␈αbut␈αit␈α
limited␈αthem.␈α It␈α
is␈αhard␈αto␈αbe␈α
a
␈↓ ↓H␈↓fanatical␈α∀intuitionist␈α∀on␈α∃discovering␈α∀that␈α∀intuitionist␈α∀and␈α∃classical␈α∀number␈α∀theory␈α∃are␈α∀equi-
␈↓ ↓H␈↓consistent.

␈↓ ↓H␈↓        Enclosed␈αare␈α
some␈αpapers,␈αsome␈α
of␈αwhich␈αI␈α
may␈αhave␈α
sent␈αyou␈αbefore.␈α
 They␈αare␈αnot␈α
mainly
␈↓ ↓H␈↓a␈α
computer␈α∞scientist's␈α
desire␈α
to␈α∞poach␈α
on␈α
philosophical␈α∞territory,␈α
even␈α
if␈α∞they␈α
may␈α
be␈α∞taken␈α
that
␈↓ ↓H␈↓way.␈α Rather␈α
they␈αare␈α
an␈αattempt␈α
to␈αtie␈αdown␈α
some␈αnaive␈α
philosophical␈αideas␈α
to␈αthe␈α
point␈αwhere
␈↓ ↓H␈↓there␈α∞is␈α∞a␈α∞chance␈α∞of␈α∞building␈α∞them␈α∞into␈α∞an␈α∞intelligent␈α∞computer␈α∞program.␈α∞ I␈α∞am␈α∞convinced␈α∞that
␈↓ ↓H␈↓such a program has to be provided with some kind of philosophical system, however naive.
␈↓ ↓H␈↓αProfessor Hilary Putnam␈↓ ¬uApril 6, 1979␈↓ 
nPage 4␈↓ 


␈↓ ↓H␈↓        I␈α⊂hope␈α∂you␈α⊂≡nd␈α∂the␈α⊂above␈α∂amusing,␈α⊂and␈α⊂I␈α∂will␈α⊂surely␈α∂call␈α⊂you␈α∂the␈α⊂next␈α∂time␈α⊂I␈α⊂come␈α∂to
␈↓ ↓H␈↓Boston.


␈↓ ↓H␈↓Best regards,



␈↓ ↓H␈↓John McCarthy
␈↓ ↓H␈↓Director
␈↓ ↓H␈↓Professor of Computer Science