/LMAR=0/XLINE=3/FONT#0=BASL30/FONT#1=BASI30/FONT#2=BASB30/FONT#3=STA200/FONT#4=NGB25
␈↓ ↓H␈↓␈↓βS␈↓∧ Artificial Intelligence Laboratory, STANFORD UNIVERSITY, Stanford, California 94305
␈↓ ↓H␈↓∧Telephone 415 497-4430␈↓ 	nFebruary 4, 1977





␈↓ ↓H␈↓Dear Colleagues:

␈↓ ↓H␈↓        Here␈αis␈αanother␈αdraft␈αof␈α␈↓↓Ascribing␈αMental␈αQualities␈αto␈αMachines␈↓.␈α It␈αseems␈αto␈αme␈αthat␈αI␈αdid
␈↓ ↓H␈↓not␈α∩succeed␈α⊃in␈α∩making␈α⊃my␈α∩ideas␈α⊃entirely␈α∩clear␈α⊃in␈α∩the␈α⊃previous␈α∩draft␈α⊃and␈α∩in␈α∩the␈α⊃Philosophy
␈↓ ↓H␈↓Colloquium,␈α∞but␈α∞the␈α∞exchanges␈α∞there␈α∞were␈α
very␈α∞helpful␈α∞in␈α∞pointing␈α∞out␈α∞areas␈α∞of␈α
incompleteness
␈↓ ↓H␈↓and confusion.  I hope the present draft is better.

␈↓ ↓H␈↓        In␈α
particular,␈αI␈α
hope␈αI␈α
have␈α
succeeded␈αin␈α
meeting␈αthe␈α
following␈αpoints␈α
raised␈α
by␈αProfessor
␈↓ ↓H␈↓Moravcsik and others:

␈↓ ↓H␈↓        1.␈α⊂Why␈α⊂doesn't␈α⊂the␈α⊂formalism␈α⊂allow␈α⊂ascribing␈α⊂beliefs␈α⊂to␈α⊂stones?␈α⊂ Ans:␈α⊂Any␈α⊂ascription␈α∂of
␈↓ ↓H␈↓beliefs␈α∞to␈α∞stones␈α∞contains␈α∂a␈α∞subascription␈α∞that␈α∞assigns␈α∞no␈α∂beliefs␈α∞and␈α∞accounts␈α∞for␈α∞just␈α∂as␈α∞much
␈↓ ↓H␈↓internal state, behavior, and is just as good at distinguishing this stone from other stones.

␈↓ ↓H␈↓        2.␈α∂What␈α∂is␈α∂the␈α⊂distinction␈α∂between␈α∂a␈α∂de≡nition␈α∂relative␈α⊂to␈α∂an␈α∂approximate␈α∂theory␈α⊂and␈α∂a
␈↓ ↓H␈↓de≡nition␈α∞based␈α
on␈α∞assumptions?␈α
 Ans:␈α∞The␈α
approximate␈α∞theory␈α
is␈α∞not␈α
assumed␈α∞to␈α
be␈α∞true;␈α∞it␈α
is
␈↓ ↓H␈↓assumed␈αto␈αbe␈αan␈αapproximation,␈αand␈αit␈αis␈αargued␈αthat␈αa␈αde≡nition␈αrelative␈αto␈αsuch␈αa␈αtheory␈αmay
␈↓ ↓H␈↓be useful ␈↓αeven if the concept cannot be de≡ned in the world or in a more re≡ned theory␈↓.

␈↓ ↓H␈↓        I␈α
have␈α
not␈α
yet␈α
formulated␈α
why␈α
I␈α
use␈α
mental␈α
terminology␈α
rather␈α
than␈α
calling␈α
them␈αpseudo-
␈↓ ↓H␈↓beliefs or talking about the presence of information.

␈↓ ↓H␈↓        Let␈αme␈αexpress␈α
my␈αappreciation␈αof␈α
your␈αtolerance␈αof␈αthis␈α
venture␈αinto␈αa␈α
≡eld␈αthat␈αis␈αnew␈α
for
␈↓ ↓H␈↓me.␈α∂ I␈α∂intend␈α∂one␈α∂more␈α∂round␈α∂of␈α∂revision␈α∂before␈α∂I␈α∂submit␈α∂the␈α∂paper␈α∂for␈α∂publication,␈α⊂and␈α∂any
␈↓ ↓H␈↓comments,␈α⊗oral␈α⊗or␈α↔written,␈α⊗will␈α⊗be␈α↔much␈α⊗appreciated.␈α⊗ I␈α↔especially␈α⊗want␈α⊗to␈α↔remove␈α⊗some
␈↓ ↓H␈↓redundancy␈α
that␈α
has␈α
crept␈α
in,␈α
remove␈α
some␈α
extraneous␈α
observations,␈α
and␈α
revise␈α
or␈α
eliminate␈αthe
␈↓ ↓H␈↓polemical section.

␈↓ ↓H␈↓        It␈α⊂now␈α⊂seems␈α⊂to␈α∂me␈α⊂that␈α⊂I␈α⊂should␈α∂eventually␈α⊂try␈α⊂to␈α⊂take␈α∂my␈α⊂own␈α⊂advice␈α⊂and␈α⊂devise␈α∂an
␈↓ ↓H␈↓approximate␈α∃theory␈α∀within␈α∃which␈α∃the␈α∀main␈α∃content␈α∀of␈α∃the␈α∃paper␈α∀would␈α∃be␈α∃technical␈α∀and
␈↓ ↓H␈↓uncontroversial,␈αleaving␈αthe␈αcontroversy␈αto␈αthe␈αoverall␈α≡t␈αof␈αthe␈αideas␈αto␈αreality.␈α
 Unfortunately,␈αI
␈↓ ↓H␈↓don't␈α∪yet␈α∪see␈α∪how␈α∪to␈α∪do␈α∪it.␈α∀ Perhaps␈α∪the␈α∪present␈α∪paper␈α∪should␈α∪be␈α∪regarded␈α∪as␈α∀a␈α∪heuristic
␈↓ ↓H␈↓introduction to the more technical paper that I don't yet know how to write.


␈↓ ↓H␈↓Sincerely,



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