1

Richard Kennington. "Finitude of Descartes' Evil Genius". Journal of the History of

Ideas, vol. 32, No. 3, pp. 441-46, 1971.
2
THE FI N I TU DE OK DESCARTES
1
EVI L GENI US
Ev Ri chard Kknni not on
At the present st age of Cart s ti)n studies, Inquiry into the nature of Carte-
sian doubt remains impeded by the familiar ticw that it is intended by Descartes
to he "universal dyubl.
1
' The conclusive evidence that duubt is intended to be
universal is eu s tu mari! y found in the omnipotence of the being who is the
ultimate ratio dubiltatdi. In the well-known argument of Cartesian doubt in
Meditations I -I , Cud, who may possibly M evil, is replaced by the supposition
of mi Evil Genius. Y et God and the Evil Genius are both omnipotent; this
view is unanimously held, so far as we have been able to ascertain, by the
distinguished Cartesian scholars of recent decades. No issue has been raised
about the equivalence of the powers of God and the Evil Genius, and hence no
distinction drawn concerning what is rendered dubitable by the one and the
other. The equivalence thesis is found in the two rtcent English studies of
L. J . Heck and Anthony Kenny; it is held by Gouhier, Al^uic and Gueroult;
il is that of leading German and American i nterpretati ons, including the mosl
recent one by Harry G. Frankfurt.
1
It is essential not only for the prevalent
interpretations of the cogito and the doctri ne of consciousness, but for the
entire litter course of he Mditations as well. On this view, it be com c s inevi-
table lo hold that Cartesi an doubt seeks, as its general intention, to establish
by universal doubt a presupposi!ion!ess first principle.
In this note we seek to establish three points. First, mathemati cs is not
doubted in Cartesian doubt. Some evidence for this exemption is that no form
of mathemati cs is even mentioned after the introduction of the Evil Genius in
Medii, I and prior to the cogito in Medit. LI. Second, the texts of the
Meditations by no means establish that the Evil Genius is omnipotent, but
imply rather his linitudc. The parallel account of doubt in Principles 1 corrob-
orates these points. Finally, the iinitude of the power of the Evil Genius is
demanded by the argument: if the Evil Genius were omnipotent, the law of
'The equivalence thesis is usually held explicitly: it is maintained implicitly by
those whu fail to discuss the issue, but say thai mat h cm a lies is doghu-d pri^r iq the
cogitai See H, Guuhic, 1-o pan.te mtaphysique de Descartes (Paris, 1962). 113; F.
Alt|ui. La dcouvert? mwphytique de J'tomme chez Descarten (Paris, I95Q), 176;
M, Guerault, Descanes selon l'ordre des raisons (Paris., 19 5 31, 4 Iff.; E. Gilson, Pisctiurt
de la rthotie, texte et commentaire (Parii, 19621, 290; L. J . Beck. The Metaphysics
o) Descartes (Onford, 1965), 72, 77; Anthony Kenny, Descartes (New York, 196$) 35-
6; Harry fi. Frankfurt, Demoni, Dreamers, and M admen. The Defense of Reason in
Descartes' Meditations (New York, 1970). 86, 92. W. H is contained In I he influential
study of J. Hinlikka.
-L
C'of'to. ergo sum'. Inferente or Performance'.'", The Philosoph-
ical Review. LXXI (1962), 9 CI Krifger identifies God and Uni Evil Genius ("Goti
ali ein 'genius malignas' , , , ."). Die Hetkurtft des phiiosophischen Seibstbe Wits s titi tis
(Darmstadt, 1962). 23. 32, Sec. however. G. W, Leibniz, Thfodice, Para. IS6. and
lelter (tu- Malebranche?) in Malebranche et Leibniz, relations personnelles, ed. A.
Robinet ( Paris, 1937), 246.
3
noncont radi ci ion would be suspended, arid all further reasoni ng would have
to cease. Hence Lhe concl usi on cap be establ i shed thai Cartesi an doubt was
never meant to be uni versal , atid the endl ess cavils aboui Descartes' proce-
dure which arc based on that assumpti on can te laid aside, The way is opened
for the presupposi ti ons of Cartesi an doubt to be com e a possible and necessary
region of inquiry, preci sel y from Descartes' own standpoi nt.
. God, "who can du everythi ng," enters M edit. 1 as a reason for doubti ng
ilE things, especially ari thmeti c and geometry; towards the end of Mdit. 1 he
is wi thdrawn as a ratio dubitandi and repl aced by the Evil Geni us, Si nce God
s omni potent, it is pl ausi bl e and customary to assume that his repl acement is
al so omni potent yet preci sel y thi s assumpti on is at issue, [f the range of
what is made dubi tabl e by the Evil Geni us is not uni versal , his omni potence is
que s lion a ble. Thi s range is i ndi cated by Descartes in two ways: 1) by the
specific list of dubi tabl es i temi zed i mmedi atel y af ter the entry of the Evii
Geni us in M edit. I; and 2) by the general review of the dubi tabl e that precedes
Lhecogd'fi) in M edit. H. The specific list divides into two parts. Fi rst, Descartes
asserts that "the heaven, air, earth, colors, figures, sounds, and all external
thi ngs" are only "the mockeri es of dreams." Secondl y, he says that he will
consi der hi msel f as havi ng "no hands, no eyes, no flesh, no blood, no senses."
1
Mai hem at ics is absent; the dubi tabl e is exclusively the external , or more
preci sel y, the bodily, i ncl udi ng one
1
; own body. The summary review 3) of the
dubi tabl e in M edit. 1 is also silent regardi ng mathemati cs, ari thmeti c, or
geometry; the Evil Geni us is again menti oned, but not mathemati cs. Descartes
does not say that the Evil Geni us renders all thi ngs dubi tabl e, but only that
"he empl oys all his effort to deceive me al ways "*
[L Our first point, that the doubt exempts mathemati cs, would remai n
probl emati c wi thout the second, that the Evil Geni us is not omni potent. For
if evil and omni potent, the Geni us would cast everythi ng, i ncl udi ng mathe-
mati cs, into dubi ety. Si nce God is unmi stakabl y omni potent, it is necessary to
compare the several descri pti ons of the power of God and the Evil Geni us, in
the Latin and French versi ons, to establ i sh their sameness or di fference, Ln
the Lati n original, l he power of God is descri bed in three ways: I) as a being
qui potest omnia; 2) as omnipejtens; 3) as tutnme patens-
1
' The power of the
Evil Geni us, thri ce menti oned in the Meditations, is rendered only by the
thi rd of these way$, by summe patens on two occasi ons, and potentissimus on
the thi rd/ Thus I) and 2], which lack evi dent ambi gui ty, arc used only to
desi gnate God and never the Evil Geni us. On the other hand, empl oys the
superl ati ve of the adverb or adj ecti ve which is notably ambi guous: it may mean
*Oettvra de etcarta, eds. C. Adam and P, Tannery (Paris, 1697-1913), V[I,
22-23; The Phtioaophitti Works of Descartes, trans, E. S. H aidant and G. R. T. Ross
(London, 193t>, L 14. The former will he referred to as AT; the latter hy MR, and U
corrected when required.
*T VI I , 24-25; HR 1,149-50.
*AT VII, 21, 40, 45, With 3) should be classed haec praeconeepia de tumma Dei
poten tia opinio. M edit. [It; AT VII, 36.
^T VII, 23,25,36,
4
either a very high degree of power, or lhe highest. IF Descartes had sought to
present the Lvii Genius as omnipotent, he could have employed one of the
unambiguous Locutions of I ) ur 2), The limitili ion of the vil Genius becomes
clearer ir lhe French version. We may legitimately appeal lo the first f-'rench
translation of the Meditations as an authoritative version si net it was read,
approved, arid, or occasi on "re touch e par Desear es."" For ih is version ir
ito case liSCS the s uni e phrases to describe lhe power of the Lvil Gcmus ai d of
God, and in no i ristante does il imply the omnipotence of the Evil Genius- The
Evi] Genius is "non moins rus ci trompeur que pui ssant," "trs pui ssant,"
and "extrmement puissant";
7
while God, "qui peuL tout." is "l i souveraine
pui ssance/' and "lout puissant
11
;* these phrases exhaust the cases in which
The non un versali ly of Cartesian do ubi may be est a Wished in twu e lose! y
related wayseither by specific exemptions, e.g., mathematics, or the
linilude of the power of the ultimate ratio dubitandi. We now consider the
compressed treatment of doubt in Principles I, 5-7", parallel to that f Mdit.
[-I I . In Princ. 1, 5 mathemati cs may be doubted because "God who created us
can do all that he desi res," In Princ. I, 6 "he who l us created us" may be
"al i -powerfu|" and seek lo deceive us
r
This being must be understood to be
God, for the Evil Genius is never mentioned in the Principles* moreover, the
power of creation is never impulcd lo the Evil Genius in the descriptions of
the Mditations, in Pnne. I, 7 God is removed from the argument of doubt,
j ust as in M edit. 1, but he is noi replaced and the Evil Genius is wholly absent.
In contrast to the Meditations, the doubt proceeds on explicitly atheist prem-
ises: "we easily suppose ihul there is no God," In Princ, I, 7 we lin d a brief
summary of lhe total range of the dubitable, but it does not mention or imply
mathemati cs as dubitable. By no sentence or phrase does Prtni. L 7 suggest
thai doubt is universal- As in M edit, [-11 the sphere of lhe dubitable is restric-
ted to the external in general, and in particular the bodily, including the body
of the doubter It may be objected, however, that Descartes in Princ. h 7
says "I h is conclusion. / think, there/ore I am, is the first and most certain of
all that occurs to one who philosophizes in an orderly way." But in Princ, 1,
10 he repeats this same sentence and qualifies il: "I did not for ail [hat deny
that we must first of all know what is knowledge, what is existente, and what
i,s certainty, and thai in order io think we must he, and such like." He did
not think these exemptions from the doubt "worthy of being put on record"
because "these are notions of the simplest possible kind, which of themselves
give us no knowledge (if anything thai exists."
LD
The cogito, then, is the
n
G. Rwdis-Lewift. ed-, hUdittftOtU da prima phiOSOphB, Mi-dillion.t mta'
"AT ]X, 16, 32. 36. Ref erences to an "author of my beirtg" vt'ho may decisive,
as in M edit. VI (AT v i . 77; I X 61), can hot refer lit rhe Evil Geni us si nce thi s phrase i i
never applied to him when he is explicitly menti oned, in. ei ther Lati n or French versi ons.
5
first and most certain amone propositions that assert existence. We may there-
fore include among the exempti ons from Cartesi an doubt all other proposi-
tions that do not assert existence, a class which must include the '"common
notions" of Fritte. T, 13, and what Descartes calls "eternal truths," "common
notions," ut "axi oms" in Princ. [ . 49. This procedure i$ also legitimated by the
fact that the one proposition explicitly exempted from doubt in Fri n . I, 10
is listed as an "eterna! truth" in Princ. I, 49'
1
: that **in order to thinfc we
must be
1
' is the same assertion as "that he who thinks must exist while he
thi nks." Moreover, Pritte
t
I, 49 lists the law of noncontradiction ("it is im-
possible that the same thing can be and not be at the same ti me"}as an eternal
truth; its exemption from the dubi tabl e is confirmed by the 2d Replies.
1
"
On the basis of these more authori tati ve versions of the Cartesi an meta-
physics in the Meditations and Principies, we may safely discount Discourse
TV, in which mathemati cs is apparentl y doubted. Whereas in the former
writings, only an omni potent being can be a ratio dubitandi for mathemati cs,
in the l atter Descartes arguts. with evident irony, that if some men someti mes
malte mi stakes even in the simplest matters of geometry, he, being also a man,
should reject all that he had ever accepted as demonstrati on,
11
In summary,
wc find this general result in the Principies and the Meditations (I ncluding the
Replies of Descartes to Objections i and even all Lhe Objections of Descartes'
critics except those of Bo urdir
1
*): if the omni potent God is a ratio dubitandi,
mathemati cs is dubitable; and if mathemati cs is dubitable, the Evil Genius is
not the ratio dubitandi- God, however, is removed from both the Meditations
and Principles asa ratio dubitandi prior to the cogito. Hence mathemati cs i
not doubl ed because only an omni potent being could make its certai nti es
dubitable. Moreover, as we have seen, what is exempt from doubt is not
limited to mathemati cs, but includes the enti re cl ass of eternal truths and
therefore the Lav of noncontradi cti on.
L
AT VU], 9-10, B, 23-24.
"AT VU, 45-46. Descartei here assigns the same status to both the cogito and
lhe law f noncontradiction, Of t ego. dum cagiio. exiftam" and "quae semel
facta imi!. Infecta esse non possimi See M, Heidegger, Die F rage nach d e m Ding
(Tilhngen, 1962), S3,
"AT VL 33. In 2d Replies (AT Vt[, 146) "no difficulty arises by the objection
that we have often found that others have been deceived in matters in which they
believed they had ltnowledge as plain as daylight." For such deception occurs only if
knowledge is drawn from the senses and not "the intellect alone."
|Jl
H the 4th Objections AauLd quotes Descartes
1
description of the Evil
Genius as surnme potens (AT VU, 19S) which is rendered in the translation as "tres
puissant" (AT IX, 154), although ill the English mistranslated as "ali powerful" (Hit
I I . BO). In 4 th Replies Descartes mentions the hyperbolic doubts of M edit, I as inclusive
of mathematics, but refers the basis of the doubts to "the author of my origin," know!-
edge of which is as yet not available. Bourdin in 7th Objections refers the doubt of
mathematics to the Evil enius{AT VII, 471), but Descartes docs not acknowledge this
basis (AT VII, 474-76), In Recherche de ta vrit, prior to the cogito there is no doubt
of mathematics, confirming ogr general thesis, although Ood is mentioned as ratio
dubitandi: the Evil Genius is absent.
6
I l L Finally, if the Evil Genius were omnipotent, tie would presumably
have the same capacity as the omnipotent God. Ur if this is alleged not to be
the case, some Cartesian text or consideration would have to be adduced to
establish thai ihc meaning of omnipotence differs i n the two cases. What then
would be the consequence of an omnipotent Evil Genius as a ratio dubitarteli
within the structure of Cartesian doubt? The om ni polen ce of God implies his
capacity to dcccivc as regards the sum of two plus three, in M edit. I
14
; the
sum could he other than five. Omni potence thus implies the power to make
true what is self-contradictory to human reason, or to suspend the law of
noncontradiction, as Descartes makes evident in At edit. I l I .'* In the Mdita-
tions. as distinct from other Cartesian texts, this extraordi nary capacity
belongs to God in virtue of one of his attri butes taken by itself, i.e., the divine
power, will, or freedom; it does not belong to the nature of God in its entirety,
or as regards all his attri butes taken together. For En the MedittiOtiS, in con-
trast with the 6th Replies and other texts,
17
the omnipotent power or freedom
of God is exercised only in accordance with the divine goodness. Because
Descartes claims to demonstrate the goodness of God in M edit. I l l IV he
can assert, at the beginning of M edit, VI, "I have never judged that anything
was impossible for God to do, unless 1 found a contradiction in being able to
conceive it clearly."
1
* In sum. whereas an omni potent being, in virtue of his
power alone, could suspend the law of noncontradiction, only an omnipotent
being w h o i s evi I would do so. H en cc. as rega rds th e co n tx I o f C a rl esian doub t,
God may be re placed, as a ratio d ubi tondi, by a Genius who is evil, but noi by
one who is omnipotent. Otherwise, the further course of Cartesian doubt,
as a reasoning inquiry, would be impossible. Only if the Evil Genius is of finite
power can the doubt or the cogito be any sort of argument whatever, ll is
therefore inconsistent for interpreters to maintain as all do explicitly or
implicitly- -that the Evil Genius is omni potent, but also to proceed to
inquire what species of argument the cogit might be, whether syllogism,
immediate inference, or whatever as most indeed do. The fmitudc of the
Evil Genius is thus Descartes" acknowledgement that the founding argument
of Cartesian doubt prior to the ro/ro is not meant to be universal. It is the
premise on which the set of presuppositions of doubt, as Descartes understood
it, becomes a region for inquiry.
The finitudc of the Evil Genius enables us to remove a problem that has
frequently clouded the structure of the Meditations. I f, on the common as-
sumption, the COgitO triumphs over an omnipotent ratio dubitandi in Medir.
11. why does Descartes reassert an omnipotent ratio dubitandi near the
beginning of M edit. I I I ? Neither the first, the Evil Genius, nor ihc second,
God, is known to be good: is not the j eopardy to knowledge the same
1
* This
^AT Vii. 2]; MR i, 147.
10
AT VII, Si; HR J, 158 59.
"6th Replies {AT V[l, 431-33, HR 1I
L
248 49}; letters to M^rsenne. 15 April, 6
and 27 May, 1630 {AT t, 145-46; 149-50; 151-53; letter Lo M est and, 2 May 1644 (AT
IV,LIE 191
"A t VII TI; HR 85.
7
difficulty has led i nterpreters lu attempt distinctions
14
that reduce the scope
of what the Evil Genius makes dubious, as contrasted with what God makes
dubious, when God reenters the argument as a raiio dubitarteli in M edit. I l l ,
But if the cogito had vanquished an omnipotent and actually Evil Genius in
M edit, I I , it surely need fear no threat from an omnipotent and possibly evil
God in M edit. I ll,
In conclusion we may remark that the above is only the Itsi preliminary
to an analysis of the profound weaning of the Evjj Genius who begins his
career in the Ofympica dream manuscript.
30
In the Meditations he reappears
almost immediately after the cogito, ari event often neglected, but explicable
on our vie* that he renders dubitable only the bodily sphere.
Pennsylvania Stal e University.
Kenny maintains that both God and the Evil Genius arc omnipotent and
hence "the two hypotheses do not dilTcr in any respect of epistemologica! significance."
yet attempts a distinction between a ''lirsHjrder doubt" in M edit. 1 II and a
11
second-
tirdcr doubt" in Medit. HI (o/?, cit,, 35, 13-4), Similarly, A. Gewiriti's distinc-
tion between a "methodological doubt
1
" and a "metaphysical doubt
1
' has the difficulty
that both are based on an omnipotent rado dubitndi that is not known to be goud and
henee both must include deception in every possible respect (fhitos. Rrvttw, L (I 9J I ),
368 95, csp. 37 J); cf. L- . Btcfc, I6>.
"Se my "Descartes' OiympiC," Sodai ReiWifl (Summer I96IJ , 176-77,
NT 204.