Kremer - Interpreting Arnauld (1990)

Interpreting Arnauld
Interpreting
Amauld
Edited by
ELMAR J. KREMER
UNIVERSITY OF TORONTO PRESS

Toronto Buffalo London
© University of Toronto Press Incorporated 1996
Toronto Buffalo London
Printed in Canada
ISBN 0-8020-0841-0
Printed on acid-free paper
Toronto Studies in Philosophy

Editors: James R. Brown and Calvin Normore
Canadian Cataloguing in Publication Data
Main entry under title:
(Toronto studies in philosophy)

All the essays but one were presented at a colloquium held at St. Michael's College,
University of Toronto, Sept. 9-11, 1994.
ISBN 0-8020-0841-0
1. Amauld, Antoine, 1612-1694 - Congresses. I. Kremer, Elmar J. II. Series.
Bl 824.A864l57 1996 194 C95-933240-5
University of Toronto Press acknowledges the financial assistance to its publishing

program of the Canada Council and the Ontario Arts Council.
Contents
Preface v11
Abbreviations x1
1 Amauld on Judging and the Will 3

Jill Vance Buroker
2 The Falsity in Sensory Ideas: Descartes and Amauld 13
Alan Nelson
3 Amauld and the Modem Mind (the Fourth Objections as
Indicative of Both Amauld's Openness to and His Distance
from Descartes) 33
Peter A. Schouls
4 Amauld and Scepticism: Questions de fait and Questions de droit 51
Thomas M Lennon
5 The Status of the Eternal Truths in the Philosophy of
Antoine Amauld 64
Aloyse-Raymond Ndiaye
6 Arnauld's Interpretation of Descartes as a Christian Philosopher 76
Elmar J Kremer
7 Amauld: A Cartesian Theologian? Omnipotence, Freedom of
Indifference, and the Creation of the Eternal Truths 91
Vincent Carraud
Contents v1
8 Arnau Id's Defence of Miracles and Its Context 111

Graeme Hunter
9 Arnau Id versus Nicole: A Medieval Dispute 127
Jean-Luc So/ere
IO 'Tange mantes et fumigabunt': Amauld on the Theodicies of
Malebranche and Leibniz 14 7
Steven Nadler
11 Amauld on Efficacious Grace and Free Choice 164
Robert C. Sleigh, Jr
Notes on Contributors I 77
Index 181
Preface
The essays in this volume, with one exception, were originally presented as
papers at a colloquium held at St Michael's College, University of Toronto, on
9-11 September 1994, commemorating the three-hundredth anniversary of the
death of Antoine Arnauld (1612-1694). 1
Arnauld was an enormously prolific theologian and philosopher who exer
cised considerable influence in both fields during his lifetime. He was the most
important theologian of the Jansenist movement, as well as a talented philoso
pher whose pursuits in that field, in a broadly Cartesian framework, continued
for over half a century and had considerable influence on Malebranche and
Leibniz, as well as on Descartes.
Arnauld was a member of a very prominent French family. He was born on 8
February 1612, the twentieth and last child of Antoine Arnauld, after whom he
was named, and Catherine (nee Marion). His father and his paternal grandfather
were prominent lawyers; both held the post of procureur general to Catherine de
Medici. His father died on 29 December 1619, before the younger Antoine's
eighth birthday.
The life of the Arnauld family was intertwined with the monastery of Port
Royal. Arnauld' s sister Angelique became abbess of Port-Royal in 1602 and
eventually carried out a thoroughgoing reform of the monastery, which became
a centre of intense religious life. Of the ten children in the family who survived
to adulthood, six became nuns at Port-Royal, where they were joined by Arnauld's
widowed mother; his older brother, Robert Arnauld d'Andilly, retired to the life
of a 'solitary' in the neighbourhood of the monastery.
As a young man, Arnauld decided to study law, but was soon convinced by his
Preface vm
mother and her confessor, Jean Duvergier, the Abbe de Saint-Cyran, to devote
himself to theology and the service of the church. He was ordained a priest in
September 1641 and received his doctorate in theology from the Sorbonne in
December of that year. His public career began auspiciously in 1641 with the
Fourth Objections to Descartes's Meditations. But the beginning of his career
was also marked by his undertaking, at the request of Duvergier, to defend Jansen,
who had been accused by various Church officials of heresy. Jansen was the
bishop of Ypres in Belgium, and an old friend of Duvergier's, with whom he had
discussed the question of how to interpret St Augustine's position on original
sin, grace, and predestination. Jansen died in 1638, and his book, Augustinus,
was published posthumously in 1640. It proved at once to be controversial, and
Habert, the theologal of Paris, attacked it in a series of sermons at the cathedral.
Arnauld responded with the first of his Apologies pour Jansenius, published
in 1644. His continued defence of Jansen eventually led to his dismissal from
the faculty of theology at the Sorbonne, after a famous trial that lasted from
1 December 1655 to 30 January 1656. Pascal's Provincial Letters were written
in defence of Arnauld, the first being published on 23 January 1656, as the trial
was coming to an end. During the following twenty-three years, Arnauld was
embroiled, off and on, in a controversy arising from the demand that all priests
and religious in France sign a formulary condemning five propositions found in
Jansen's book. In 1679, Arnauld left France for Belgium, where, except for a
sojourn of two years in Holland, he lived until his death on 7 August 1694. His
years in self-imposed exile were his most productive, and included both his con
troversy with Malebranche and his correspondence with Leibniz.
Arnauld took seriously his calling as a theologian and, as A.-R. Ndiaye points
out, thought he was justified in taking time off for philosophy only because it
made a contribution to theology. Voltaire said of Arnauld that, although 'no one
was ever born with a more philosophical mind,' he wasted his time on theologi
cal disputes and never realized his potential. 2 More recently, Robert McRae re
marked that Arnauld's dedication to theology was a loss to philosophy because
he would otherwise surely have been one of the greatest philosophers of his
century. Yet Arnauld made no small contribution to philosophy and, paradoxi
cally, has had a more enduring influence there than in the field of his chosen
profession. Modem theologians tend to view him as a figure in the (happily)
defunct Jansenist movement. But his work continues to draw fresh responses
from philosophers.
The essays in this volume are presented, for the most part, in the order in
which they were given at the colloquium. They are not connected in any simple,
linear order, but rather in several different, overlapping ways, both historical and
systematic. Most of the essays discuss Arnauld's role both in the development of
Cartesianism and in the continuation of medieval disputes in the seventeenth
Preface 1x
century. Also, most of essays deal either with the relation between philosophy
and theology in Amauld's thought or with theological topics that are of special
interest to philosophers.
Amauld thought it important to keep the distinction between philosophy and
theology clear, as can be seen in this comment at the beginning of his Fourth
Objections:
Although philosophy can claim this entire work [Descartes's Meditations] as its
own, nevertheless, because the author has respectfully and willingly submitted him
self to the tribunal of the theologians, I shall here act in two capacities: I will first put
forward what it seems to me could be objected to by philosophers regarding the
important questions of the nature of our mind and of God; and then I shall set forth
what could be offensive to theologians in the entire work. 3
What emerges in the essays in the volume, however, is the unity of Amauld's
thought. Amauld's skill in logical analysis is discussed in all of the essays, espe
cially in the opening essays, by Buroker and Nelson, and the final essay, by
Sleigh - although he may not always have put this skill to good use. 4 All of the
remaining essays deal with topics that are part of both philosophy and theology.
In particular, Amauld's attitude towards the Cartesian doctrine of the creation of
the eternal truths is discussed in the essays by Ndiaye, Kremer, Carraud, and
Solere. The topics discussed by Hunter (miracles), Nadler (theodicy), and Sleigh
(the compatibility of grace and free choice) also straddle the two disciplines.
Amauld emerges in these essays as a figure who wanted to continue the medi
eval theological tradition and, at the same time, to embrace the new philosophy.
He is, thus, a Cartesian who emphasizes the continuity of Descartes with the
Middle Ages and who understands, and fears, in his contemporaries what Sleigh
has characterized as a certain 'boldness of reason which would, in time, spark
the Enlightenment. '5 His attempt to defend this position is marked by skilful
logical analysis of many propositions and arguments central to seventeenth
century philosophy and theology.
Notes
The exception is the paper by Jean-Luc Solere, which was presented at the
Department of Philosophy, University of Toronto, in December I 994. The
colloquium was funded mainly by a grant from the Social Science and Humanities
Research Council, with assistance from the Department of Philosophy, University
of Toronto, and St Michael's College. The editor would like to thank the Journal
of the History of Ideas for permission to publish the paper by Steven Nadler,
Preface x
an earlier version of which appeared in that journal.

2 Voltaire, Siecle de Louis XIV (Paris: Librairie Hachette 1890), 728.
3 OA 38, 8; AT 7, 197; CSM 2, 138.
4 See the essay by Lennon, in particular, for a negative assessment of one important
logical manoeuvre used by Arnauld in his response to the church's condemnation
of Jansen.
5 Robert C. Sleigh, Jr, Leibniz and Arnauld: A Commentary on Their
Correspondence (New Haven, CT: Yale University Press 1990), 47.
Abbreviations
The following abbreviations are used throughout this volume.
OA Oeuvres de Messire Antoine Arnauld, edited by G. Du Pac de

Bellegarde and J. Hautefauge, with Vie de Messire Antoine Arnauld
by N. De Larriere, 43 vols (Paris-Lausanne: Sigismond D' Amay et
Cie 1775-83).
AT Oeuvres de Descartes, 12 vols, edited by Charles Adam and Paul

Tannery (Paris: Leopold Cerf 1897-1913).
Alquie Oeuvres philosophiques de Descartes, edited by Ferdinand Alquie,

3 vols (Paris: Gamier 1963, 1967, 1973).
CSMK The Philosophical Writings of Descartes, vols 1 and 2 edited and

translated by J. Cottingham, R. Stoothoff, and D. Murdoch; vol 3
edited and translated by J. Cottingham, R. Stoothoff, D. Murdoch,
and A. Kenny (Cambridge: Cambridge University Press 1984-91).
oc Oeuvres completes de Malebranche, under the direction of Andre

Robinet, 20 vols (Paris: J. Vrin 1958----67).
G G. W Leibniz: Die philosophischen Schriften, edited by C. I. Gerhardt,

7 vols (Berlin 1875-90).
1
Arnauld on Judging
and the Will
JILL VANCE BUROKER
The Cartesians of the seventeenth century had very little to say about language.
They were primarily concerned about the nature of mental functions, especially
conceiving and judging. Despite their focus on the mental, their theory of judg
ment incorporates a semantic theory, which invites comparison with current
views in speech-act theory. In this essay I discuss Antoine Arnauld's account of
judgment in the Port-Royal Logic from two perspectives. The first is Descartes's
theory ofjudgment; the second is that of contemporary philosophy of language.
The fust comparison shows how Arnauld and Descartes divide up the functions
of understanding and willing somewhat differently. The contemporary perspec
tive provides a good basis for appreciating the strengths, and ultimately the
weaknesses, of both seventeenth-century theories. I shall reverse chronological
order and start with the contemporary treatment of judgment.
Judgment and Illocutionary Force
Frege was the first philosopher clearly to distinguish the sense expressed
by a sentence, which he calls the thought, from the way a sentence is used
in a given context. This latter aspect of meaning he calls the force of the utter
ance. So he recognizes the difference between merely expressing a thought,
for example, 'five is greater than four,' and the act of asserting that five is
greater than four. 1 In his 1918 essay 'The Thought,' Frege distinguishes think
ing from both judging and asserting. On his view, when one thinks, one merely
apprehends a thought; judging occurs when one mentally acknowledges the
truth of a thought; and, finally, assertion is the linguistic or external expression
of a judgment. Frege was motivated to separate the sense of an utterance from
4 Jill Vance Buroker
its force through recognizing the possibility of using declarative sentences

without assertoric force, for example, in making suppositions and in fictional
contexts. Most important, this is the only way to understand the use of embed
ded clauses in complex sentences, such as conditionals and disjunctives. When
I assert the conditional � If you go to Paris, you ought to see the Rodin museum,'
I cannot be taken to be asserting that you are in fact going to Paris. Frege was
concerned about the roles such sentences play in inference, though ironically
his own treatment of valid inferences does not allow for unasserted supposi
tions. In any case, his primary focus on scientific and mathematical contexts
prevented him from giving an adequate general account of force. 2 Despite the
incompleteness of his account, however, the distinction is clearly a step in the
right direction.
Since the 1950s, developments in the philosophy of language have altered the
conception of judgment in at least two major respects. First, the trend has been
towards externalism in philosophy of mind, according to which mental acts are
analysed in terms of linguistic or other behavioural activities. From this per
spective, the unit of cognitive significance is the speech act, and mental acts
such as judging are seen as derivative of speech acts such as asserting. The
second change concerns the increased prominence of force in this theory. Speech
acts are utterances performed in a social context against a background of
conventional rules, including linguistic rules. The most general significance of
the utterance is called its illocutionary force. This is the immediate effect the
speaker intends to produce in the hearer by getting the hearer to recognize the
speaker's intention. For example, your promise to me to visit the Rodin museum
on your next trip to Paris succeeds as an illocutionary act when I recognize it as
that promise. In addition to assertions and promises, other common illocutionary
acts are questions, commands, requests, and so on. What Frege called the sense
of the sentence is here viewed as the propositional content of the act, and speech
acts with different illocutionary force may contain the same propositional con
tent. So my assertion that the window is open contains the same proposition as
the question 'Is the window open?' and the command 'Open the window' (as
suming I am referring to the same window in all three cases). Despite the wide
variety of speech acts, assertion is taken to be primary, since it is essential to
propositions that they be suitable for being asserted. But, contrary to Frege's
tendency to reify the thought as an eternal object, recent thinkers consider the
thought to be one ingredient of the act, one that is really an abstraction from the
entire act.
For my purpose, the key feature is the relation between merely apprehend
ing a thought (in some utterance) and asserting the sentence. In this theory,
whenever one says anything meaningful, the act has some illocutionary force.
But it is possible merely to express a thought without committing oneself to its
Amauld on Judging and the Will 5
truth or to its falsity. This happens if I say 'Suppose you travel to Paris next
year ...' as opposed to predicting that you will in fact travel to Paris next year. Of
course, there is no point to merely entertaining a proposition except in prepara
tion for some other activity. Making assumptions or hypotheses, for example, is
usually significant only in the context of inferring consequences from them. In
any case, in terms of illocutionary force, acts of apprehension are on a par with
judgments or assertions. Rather than being a necessary condition of assertion,
considering a proposition is just one of many different kinds of illocutionary
acts.
Now a word about the relation between assertion and negation. As Frege saw
clearly, the difference between affirming a proposition P and denying it is not in
the nature of the act, but in the proposition. Denying P is equivalent to asserting
not-P. In other words, assertion has no logical poles - there is only one form of
assertion, and negation generally belongs to the content of the proposition. Of
course, it is possible to refuse to assert a proposition, but this is not the same as
denying the proposition. From the modem point of view, then, affirmations and
denials do not differ as assertive acts.
Finally a brief comment about the treatment of truth. While there is some
disagreement among extemalists, the most extemalist position (favoured by
Dummett, for example) takes truth-values to be properties of assertions rather
than of the propositions asserted. When we speak of the truth or falsity of the
thought, we really mean that its assertion would be correct or incorrect, given
the circumstances. (Recall that Frege had to 'decontextualize' the thought by
getting rid of token-reflexives such as pronouns and demonstratives to ensure
that it had an eternal truth-value.) There are items, such as 'eternal truths' of
mathematics, for which conditions of utterance are irrelevant. But it does not
follow from this that propositions are the genuine bearers of truth-values.
Dummett prefers to say that the truth-value of the utterance depends jointly on
the sense of the sentence and the conditions of utterance; in limiting cases, such
as eternal truths, the conditions of utterance play no role at all.3
Descartes on Judgment
Descartes embraces the more familiar theory of internalism. From this intui
tively appealing standpoint, speech is merely the external expression of mental
states, and, so, judging is a prerequisite of assertion. For Descartes, mental states
are produced by the understanding or the will. Whereas the understanding merely
conceives ideas, the will acts or operates in some way on those conceptions.
Judging is an act of willing since it consists in taking a proposition to be true or
false. Now, although Descartes thinks of propositions as composed of ideas, he
does not draw a sharp line between conceiving a non-propositional idea and
thinking a proposition. For him the content of an idea can be very complex, and
this complexity can be expressed propositionally. One famous example is
Descartes's statement in the Fifth Meditation that understanding the idea of a
right triangle entails recognizing that it has 'the properties which license the
inference that its three angles equal no more than two right angles. '4 Similarly,
his version of the ontological argument depends on the view that an adequate
apprehension of the attributes contained in the idea of God logically entails the
proposition that God necessarily exists. Although to my knowledge Descartes
never refers to simple or unanalysable ideas, he does take understanding to be
simpler than willing. He also characterizes acts of understanding as passive, in
contrast with the actions of the will.
For Descartes, willing is more complex than understanding precisely because
acts of will incorporate perceptions of the understanding. In Part I, article 32, of
the Principles, Descartes lists the modes of willing as 'desire, aversion, asser
tion, denial and doubt.'5 In the Meditations he adds emotions to the list. What
these mental states have in common is spelled out in the Third Meditation: In
contrast to ideas, which he says are 'as it were the images of things,' volitions
include 'something more than the likeness of that thing. '6 In general the extra
element is an attitude the subject takes towards the object of thought, such as a
desire for the object, or fear or love of the object, or a judgment concerning the
object.
Descartes's most explicit statement about judgment occurs when he re
ponds to Regius in Comments on a Certain Broadsheet. Noting that Regius
assigns both perception and judgment to the intellect, Descartes makes this
correction:
For I saw that over and above perception, which is a prerequisite of judgement, we
need affirmation and negation to detennine the fonn of the judgement, and also
that we are often free to withhold our assent, even if we perceive the matter in
question. Hence I assigned the act of judging itself, which consists simply in as
senting (i.e. in affirmation or denial) to the detennination of the will rather than to
the perception of the intellect. 7
In short, the view is this: Like other volitions, judgments contain conceptions of
an object or possible state of affairs. In the case of judgment, the conception
takes propositional form, and the volitional attitude is one's commitment to a
truth-value of the proposition. When one affirms, one commits oneself to the
truth of the proposition; when one denies, one commits oneself to its falsity.
Refraining from judging is no less an act of the will: so doubting is among the
modes of willing.
In this model, the conception of the understanding is a necessary ingredient
of the judgment: it represents the object of the judgmental attitude. But since
Arnauld on Judging and the Will 7
one can apprehend a thought without judging it, conceiving is separable from
judging. When we add to this Descartes's view that intellectual intuition is pas
sive, what emerges is a transitive model of judgment. First, the understanding
supplies the propositional content; then, the will acts on it. In this framework,
merely apprehending or entertaining a thought is not only possible, but neces
sary. As opposed to doubting or refraining from judging, however, mere appre
hension is not an act of will.
Although Descartes sometimes describes judging as giving assent, he just as
frequently treats affirmation and denial as different acts of the will. In the pas
sage previously cited from the Principles, for example, he compares affirmation
and denial to desire and aversion. Whether this makes sense for desire and
aversion, it clearly does not work for judgment. As Frege argued in his essay
'Negation, ' this model requires two types of negation: one characterizing the
act of denial, another belonging to the propositional content. 8 On this model, to
affirm 'The Earth is round' and to deny 'The Earth is not round' would be two
different types of acts, operating on different propositions, in spite of their logi
cal equivalence. But propositional negation is also necessary, since it is surely
possible to apprehend a negative proposition. This leads to cases where it is
not clear whether the negation belongs to the proposition or to the act. Is assert
ing a previously apprehended negative proposition an affirmation or a denial?
These are some of the reasons Frege abandoned the view that affirmation and
denial are polar forms of judgment, as well as the traditional classification of
propositions into affirmative and negative.
Finally, a word about Descartes ' s conception of truth . In the Third
Meditation, Descartes officially states that ' formal' truth and falsity reside
in j udgment. But in fact it seems that the propositions being grasped are
the bearers of truth-values, since clear and distinct perception is tantamount
to perceiving the truth of the proposition. Th is is understandable, given
Descartes's preoccupation with eternal truths, such as mathematics, the laws
of physics, and those concerning God's nature. This fits nicely with the transi
tive model, according to which one grasps a proposition which is true or
false independently of its being thought, and then commits oneself to a
truth-value. The problem here is the notion of the commitment to a truth
value, since it is not clear whether the apprehension of the thought ' P is true'
would be a mere conception or a judgment.
Judgment in the Port-Royal Logic
Arnauld's conception ofjudgment in the Port-Royal Logic differs from Descartes's

in some interesting respects. Arnauld agrees with Descartes that thought is prior
to language, and so judging can take place independently of its linguistic
expression. But, whereas Descartes says very little about language, Arnauld
gives it a central focus in the Port-Royal Logic's semantics. What results is a

'constructivist' model of judgment, as contrasted with Descartes's 'transitive'
model. This shift exposes two different ways of dividing up the functions of the
understanding and the will.
In general, Arnauld holds the traditional view of the order of mental opera
tions. As he explains in the introduction, the Port-Royal Logic is organized
according to the four operations of the mind: conceiving, judging, reasoning,
and ordering. Part I concerns ideas, which are the products of conceiving, de
fined as 'the simple view we have of things that present themselves to the mind.'9
The things represented by ideas are substances and their properties, which
Arnauld generally calls 'attributes.' Arnauld says expressly that the act of con
ception is simple in the sense that it does not include an explicit judgment. By
contrast, judging is an action in which the mind, 'joining together different
ideas, affirms of one that it is the other, or denies of one that it is the other'
(p. 37). For example, both the affirmation 'The Earth is round' and the denial
'The Earth is not round' contain the same ideas, 'earth' and 'round'; the differ
ence is in the action the mind takes towards them.
Up to this point, the theory looks Cartesian, but it turns out that, for Arnauld,
joining ideas to form a proposition is an act of the will rather than of the under
standing. This is clear from Chapter 3 of Part II, which begins:
After conceiving things by our ideas, we compare these ideas and, finding that
some belong together and others do not, we unite or separate them, which is called
affirming or denying, and in general judging.
This judgment is also called a proposition, and it is easy to see that it must
have two terms. One term, of which one affirms or denies something, is called the
subject� the other term, which is affirmed or denied, is called the attribute or
Praedicatum. (p. 1 1 3 )
Here Arnauld equates forming the proposition with judging it. 1 0 I n the act
of connecting the subject and the predicate, one necessarily commits oneself
to a truth-value. This reading is reinforced by his treatment of the verb. In
Chapter 2 of Part II, Arnauld explains that what distinguishes a complex
idea, for example, of Peter living, from the affirmation 'Peter is living' is the
presence of the verb. And he defines a verb as
a word whose principal function is to signify an affirmation, that is, to indicate

that the discourse where this word is employed is the discourse of a person who
not only conceives things, but who judges and makes affirmations about them. This
is what distinguishes the verb from several nouns that also signify affirma
tion, such as affirmans, affirmatio [affirming, affirmation], because they signify
affirmation only in so far as it has become an object of thought by mental reflection.

Hence they do not indicate that the people who use these words are making an
affirmation, but only that they conceive of an affirmation . . . Accordingly, we can
say that the verb in itself ought to have no other use than to indicate the connection
the m ind makes between the two terms of a proposition. (p. 1 09)
So, in the Port-Royal Logic, the verb simultaneously performs two functions:
it connects the subject and predicate to form a proposition, and it indicates
the affirmation or denial one makes. When a negative particle such as 'not' is
attached to the verb, the judgment is (apparently) a denial; otherwise one is
affirming the proposition. 1 1
Like Descartes, Amauld treats affirmation and denial as two polar forms of
judgment. But since he cannot separate the proposition from the judgment, the
question of whether the negation affects illocutionary force or the propositional
content cannot even arise. But this is not the worst problem: in explaining how
affinning differs from denying, Amauld makes it impossible for one to deny a
proposition, since he thinks of denying as an action opposite to affirming. In
affirming, one connects the subject and predicate to fonn a propositional unity
(while assenting to the truth of the proposition so formed); thus, in denying, one
must separate the subject from the predicate. Here is what he says in Chapter 3
of Part I I:
As we have already said, this action of the mind [judging] is indicated in discourse
by the verb is, either by itself when we make an affinnation, or with a negative
particle in a denial . Thus when I say, God is just, ... the word is indicates the action
of the mind that affinns, that is, that connects the two ideas God and just as belong
ing together. lf l say, God is not unjust, the word is, when joined to the particle not,
signifies the action contrary to affinning, namely, denying, in which I view these
ideas as repugnant to one another, because the idea unjust contains something con
trary to what is contain in the idea God. (p. 1 1 3)
In his essay 'Negation,' Frege points out the implication of this view: I f, in
denying, the mind separates the subject and predicate ideas, then there is no
propositional unity to be judged. And if there is no proposition, then clearly
there is nothing to take a truth-value. Hence Amauld undercuts the entire idea
of denying as taking a proposition to be false. In one of his brilliant metaphors,
Frege adds that this view of denial makes of double negation a magical sword
'that could heal on again the limbs it had cut off. ' 1 2
Independently of the problem of denial, Amauld's model of judging differs
from Descartes's transitive model in some important ways. I call Amauld's
theory a 'constructivist' model because for him the act of judging creates the
propositional content being judged at the same time. Here the proposition
cannot be passively apprehended by the understanding. So Amauld's view of
conceiving must also differ from Descartes's. For one thing, such acts of the
understanding as forming complex ideas and analysing them into parts cannot
be equivalent to propositional acts for Amauld. Whereas Descartes attempts to
draw a sharp line between forming a proposition and judging it, Amauld at
tempts to draw the line between conceiving an idea and connecting it with oth
ers to form ( and judge) a proposition. I say 'attempts' because neither succeeds.
Descartes had difficulty classifying the state in which one apprehends the truth
of a proposition. And Amauld, in his treatment of relative clauses, maintains
that some complex ideas do contain judgments and others do not. 1 3 In short,
neither version is coherent. Here is how Amauld and Descartes go wrong.
It seems fair to equate the Cartesian notion of acts of the will to the illocutionary
force of speech acts, since both involve taking an attitude to some propositional
content. Recall that, in speech-act theory, merely considering a proposition is
also an illocutionary act, typically preceding some other act. Descartes allows
that one can apprehend a proposition without judging it, which seems right. But
he sees mere apprehension as an act of understanding rather than of will, and
that is where he goes wrong. By contrast, Amauld recognizes that merely ap
prehending a proposition involves some mental activity, which is right, but he
errs in not separating mere apprehension from judging as different types of acts.
The underlying problem for both Descartes and Amauld is the relation be
tween the will and the understanding. Although they divide up the functions of
the understanding differently, each one sees it as a prerequisite to willing, rather
than as a feature of some attitudinal activity. This point is not clarified until
1781, with Kant's full-blown constructivist theory of judgment in Critique of
Pure Reason. 1 4 Whether the synthetic activity involved in judging should be
called an act of will is another question. Kant never refers to it that way in the
first Critique; since ordinary sense perception involves judging, most of the
time we are not even conscious of judging. But we should also recall that part of
Kant's revolutionary treatment is to reject Descartes's notion that human
understanding is capable of a passive, instantaneous, intellectual intuition of
the truth. Although Amauld apparently agrees with this view, he does not see
its far-reaching consequences for his theory of judgment.
Notes
A clear explanation of this aspect of Frege' s theory is available in Michael

Dummett' s Frege: Philosophy of language, 2d ed. (Cambridge, MA: Harvard
University Press 1 98 1 ). See especially ch. 1 0, ' Assertion,' 295-363 .

2 In fact, in ' On Sense and Reference,' he claims that what distinguishes assertions
from interrogatives, imperatives, and optatives is the sense rather than the force.
And in the Begriffschrift, he introduces only two signs for force: the judgment
stroke indicating assertion (' I-'), and the sign for definition, indicating stipulation
C'II-').
3 See Dummett, Frege, 4 1 8-20.
4 AT 7, 68.
5 AT 8, 1 7.
6 AT 7, 37.
7 AT 8B, 3 63 .
8 Reprinted i n Translations from the Philosophical Writings of Gottlob Frege,
Peter Geach and Max Black, eds (Oxford: Basil Blackwel l 1 966), 1 1 7-35 .
9 Antoine Amauld and Pierre Nicole, La Logique ou I 'art de penser, critical edition
by Pierre Clair and Fran�ois Girbal (Paris: J. Vrin 1 98 1 ), 37. All references to the
Logic are to this volume. All translations are my own.
1 0 It is noteworthy that in the first ( 1 662) edition of the Logic, Amauld apparently
wavers between this and Descartes's conception of judgment, that is, as to whether
the act of judging also constitutes the proposition. Here is what he says at what
later becomes Chapter 1 7 of Part II: ' It is not easy to make clear nor even to
understand what takes place in the mind whenever we affirm something, and to
decide whether this is done by the simple view of the mind accompanied by
consent, by which it represents something as containing a certain attribute by a
single idea, or whether there really are two ideas, one for the subject and the other
for the attribute, with a certain act of the mind which connects one with the other'
( 1 68). Here he expresses uncertainty about whether affirming consists in assenting
to a single propositional idea, or in forming a complex idea by connecting the
subject and attribute.
1 1 On this conception, verbs signify quite differently from the way the subject-term
and the predicate-term signify. The latter words, typically nouns and adjectives,
express ideas of substances and attributes. In medieval parlance, such words
are called ' categorematic' expressions. The verb, by contrast, signifies neither
a substance nor an attribute, but the action of the mind in forming and judging
the proposition. Other examples of syncategorematic expressions are quantifiers
such as ' al l ' and 'some. ' In this same discussion, Arnauld recognizes illocutionary
acts other than assertion, since he points out that verbs can also signify 'other
actions of the soul, such as desiring, requesting, commanding, etc. But this
happens only by changing the inflection and the mood ' ( 1 09).
1 2 Translations from the Philosophical Writings of Gottlob Frege, 1 24.
1 3 I discuss this in ' Judgment and Predication in the Port-Royal Logic, ' in The Great
Arnauld and Some of His Philosophical Correspondents, Elmar J. Kremer, ed.

(Toronto : University of Toronto Press 1 994 ), 3-27.
14 Ironically, in so far as Frege treats thoughts as objects that are passively grasped,
he backslides towards a more Cartesian view.
2
The Falsity in Sensory Ideas:
Descartes and Arnauld
A L AN N E L SON
Do Ideas Deceive?
Since Arnauld's set of Objections to the Meditations, many commentators have

found parts of Descartes's theory of sensory ideas particularly difficult. 1 Some
are inclined to attribute changes of position, confusion, and worse to the texts
dealing with material falsity. 2 This essay develops an interpretation of Descartes's
doctrines of materially false ideas and of sensory error that strives to be wholly
Cartesian in spirit. It also makes aspects of Arnauld's own positive theory rel
evant to understanding Descartes. Finally, there is a new theory of ' obscurity
and confusion' (and hence of clarity and distinctness) as it applies to sensory
ideas. The drawback of this interpretive strategy is that Descartes's interests
and goals do not tum out to be as continuous with contemporary philosophical
concerns as some have hoped. In particular, Descartes does not have (or need)
an impressive theory of what is now called ' intentionality'; worse still, his fre
quent, and in some ways seminal, talk of representation turns out to be driven
by concerns in natural science and not in semantics.
This essay's interpretation relies on a picture of the relation between false
ideas and false judgments that departs from a widely accepted one. The widely
accepted picture begins with Descartes's distinguishing a class of ideas as
'materially false.' Descartes plainly thought that many people have an inclina
tion to make judgments on the basis of these materially false ideas, particularly
sensory ideas, that tum out to be false judgments. It is important to ask what
Descartes thought explained this inclination to make false judgments. The widely
accepted picture goes wrong in answering this question. Its answer is that
14 Alan Nelson
materially false ideas 'present themselves,' in a phenomenological sense, as

representing or exhibiting objects as having properties that those objects don't
really have. So, for example, when I see an apple, I have a sensation of redness
- the sensory idea of red is occasioned in my mind. This sensory idea of red
presents itself to me as though its content, the redness, is really in the apple. If
I am not on guard, or if I am a young child, and nai'vely judge that 'what the idea
tells me' is true, I will mistakenly judge that the redness is a property of the
apple, that it is 'in' the apple. My inclination to judge falsely in this case is
explained by my uncritically acceding to 'what the idea presents to me' or 'seems
to tell me.' In general, the widely accepted picture explains this class of false
judgment by appealing to an intrinsic deceptiveness in the phenomenological
contents of false ideas.
The picture advanced in this essay offers a different interpretation of what
Descartes thought explained false judgments. I argue that the inclination to
judge falsely is not explained by the faculty of judgment (or will) gullibly going
for what is held up to it by deceptive ideas. The inclinations are instead ex
plained as habits of judgment. They are, from the perspective of philosophy, bad
habits fostered partly by poor techniques of reasoning that serve us well in child
hood, and even in everyday life, but not when doing philosophy. In scholarly
circles, however, the bad habits are reinforced by the practice of bad science.
According to Descartes, bad science includes the sort of 'folk physics,' or 'folk
physiology of perception' that insufficiently thoughtful people engage in from
time to time, and also organized, institutional, bad science of the kind that he
was concerned to overthrow. One upshot of this picture is the shifting of much
of the blame for the mistakes in question from the intrinsic deficiency of our
ideas (which tum out not to be so bad after all) to correctable habits of judg
ment. Moreover, any intrinsic deficiency in our ideas turns out to be traceable to
prior bad judgments.
Preliminaries: Material Falsity and Objective Reality
Let us begin with the difficult Third Meditation treatment of sensory ideas that
drew a sharp response from Amauld. Amauld focused on Descartes's charac
terization of some ideas as 'materially false.' In the Fourth Replies, Descartes
elaborates on this terminology:
As I interpret this claim, it means that the ideas are such as to provide subject
matter for error. (AT 7, 23 1 ; CSMK 2. 1 62)3
It is this idea [of cold] which, I claim, can provide subject-matter for error if it is in
fact true that cold is an absence and does not have as much reality as heat; for if I
The Falsity in Sensory Ideas 15
consider the ideas of cold and heat just as I received them from my senses, / am
unable to tell that one idea represents more reality to me than the other. (AT 7,
232-3 ; CSMK 2, 1 63 ; emphasis added)
So an idea is materially false if it provides 'subject-matter' for error, for false

judgment. This plain fact about Descartes's use of the term 'materially false'
has some straightforward, but interesting consequences. 4
First, it is not only ideas occasioned by motions in the sensory organs and
brain that can be materially false; any idea that provides subject-matter for false
judgment is materially false. Since we very easily make false judgments, what
ideas could fall outside the range of the materially false? They would have to be
ideas that are true, and whose truth is invariably accompanied by a judgment
that they are true. Descartes consistently cal ls these 'clearly and distinctly
perceived ideas. ' Clear and distinct perceptions are invariably accompanied by
assent to the truth of the perceived ideas:
I could not but judge that something which I understood so clearly was true; but
this was not because I was compelled so to judge by any external force, but because
a great light in the intellect was followed by a great inclination in the will . . . (AT 7,
5 8-9; CSMK 2, 4 1 )
Admittedly my nature is such that so long as I perceive something very clearly and
distinctly I cannot but believe it to be true. (AT 7, 69)
The will of a thinking thing is drawn voluntarily and freely (for that is the essence
of will), but nevertheless inevitably, towards a clearly known good. (AT 7, 1 66;
CSMK 2, 1 1 7; emphasis added)
These texts plainly say that clear and distinct perceptions are invariably accom
panied by assent. 5 The only way to avoid the 'compulsion' of a clear and distinct
idea is to distract the attention so that the idea is no longer clearly present to the
intellect. If we allow the attention to wander in this way, then we no longer
'continue in the same thought' so that particular thought no longer compels
assent.
The second consequence of Descartes's notion of material falsity to consider
here is that it is a matter of degree. All ideas that are not clear and distinct are
materially false to the degree that one is likely to make false judgments when
considering them.
Yet ideas which give the judgment little or no scope for error do not seem as much
entitled to be called materially false as those which give great scope for error. It is
16 Alan Nelson
easy to show by means of examples that some ideas provide much greater scope for
error than others. (AT 7, 233; CSMK 2, 163).
The ideas with the least amount of material falsity (setting aside the clear and
distinct ideas that give no material whatsoever for false judgment) are the facti
tious ideas. Since they are made up at will, we typically have some recognition
of their origin, and that can suggest that the composites produced need not
correspond closely to things outside me. Those labouring under the prejudices
of childhood or bad science, however, will often go wrong in these cases. The
greatest scope for false judgments is provided by those ideas concerning which
our composite nature inclines us to judge falsely: sensations, and especially
appetites. 6 In the Sixth Meditation, Descartes says that God has endowed us
with the faculty of sensation to provide for the preservation of lives as embodied
humans. When we follow our natural inclinations to act on the basis of these
ideas, the ordinary result will be something beneficial to the union. The situa
tion is completely different when our immediate goal is, not to preserve the
union, but to discover and contemplate the truth. In this case, we should employ
our faculty of understanding (the literal translation of intellectio, 'intellection, '
might be preferable) and assent to only those ideas which it clearly and dis
tinctly perceives.
For the proper purpose of the sensory perceptions given me by nature is simply to
inform the mind of what is beneficial or harmful for the composite of which the
mind is a part ... But I misuse them by treating them as reliable touchstones for
immediate judgments about the essential nature of the bodies located outside us ...
(AT 7, 83; CSMK 2, 57-8; see also Principles II, 3)
If we misuse our faculties by uncritically attempting to gain the truth about the
extramental world through our senses, we almost always fail.
It is tempting to characterize material falsity differently. The Third Medita
tion seems to suggest that some materially false ideas 'represent non-things
as things' (AT 7, 43 ; CSMK 2, 30), and this seems to lead to a famous puzzle
about these ideas' objective reality. Objective reality is characterized in the Third
Meditation in terms of representation: 'Undoubtedly, the ideas which represent
substances to me amount to something more and, so to speak, contain within
themselves more objective reality than ideas which represent merely modes or
accidents' (AT 7, 40; CSMK 2, 28). Now, if the idea of cold represents a non
thing to me, then, since the objective reality of an idea has as its efficient and
total cause the formal reality of what is represented, ideas of this kind appear to
have no objective reality. That would be a bad result because all ideas are like
images in so far as they have objects - in so far as they are ' as if of things' (AT
7, 44; CSMK 2, 30).
We should, however, strive to find an interpretation in which there can be no
special difficulty about the degree of objective reality of any materially false
ideas, including the sensory ideas. Let us consider some further texts concerned
with what Descartes calls ' objective reality.' It is important first to mark that
Descartes conceives objective reality as quantifiable into exactly three discrete
degrees or levels. In Axiom VII, from the ' Geometrical Exposition' in Second
Replies, he wrote:
There are various degrees of reality or being: a substance has more reality than an
accident or a mode; an infinite substance has more reality than a finite substance.
Hence, there is more objective reality in the idea of an infinite substance than in
the idea of a finite substance. (AT 7, 1 65---{i; CSMK 2, 1 1 7).
When Hobbes explicitly challenges him to 'consider afresh what "more reality"
means,' Descartes replies in exactly the same tenns. 7 Both these passages sim
ply echo what he originally wrote in the Third Meditation, which now appears
in a different light:
Undoubtedly, the ideas which represent substances to me amount to something

more and, so to speak, contain within themselves more objective reality than the
ideas which merely represent modes or accidents. Again, the idea that gives me my
understanding of a supreme God, ... certainly has in it more objective reality than
the ideas that represent finite substances. (AT 7, 40; CSMK 2, 28)
This should be read as saying that some ideas represent to me modes of sub
stances and contain the modal level of objective reality; some represent sub
stances and contain the (finite) substantial level of objective reality. Both kinds
of ideas contain finite objective reality; the idea of God contains the infinite
level of objective reality. 8 This interpretation carries with it a reading of 'repre
sent' (or ' exhibit') that is applicable in the current context. In this sense of
' represent,' ideas represent either God or created substances or modes, without
further differentiation. In other words, in this sense of 'represent,' ideas do not
specifically represent such particular objects as the sun, dogs, heat, etc. They
simply represent, in virtue of their objective reality, the amount of formal reality
their causes have. 9 The general point is familiar from the Third Meditation' s
'causal proof' o f God's existence. We might say that ideas present this reality,
but let us instead coin the term reality-representation. 1 0 Reality-representation
is intrinsic to the idea and does not depend on any judgment. We might err in
18 Alan Nelson
judging how much reality an idea represents if the idea is not clear and distinct,
but that does not affect how much reality it actually, intrinsically represents. Of
course, I shall have to explain in what follows Descartes's other use of 'repre
sent,' the sense in which the sun, a dog, heat, etc. , could be individually repre
sented. It also remains to be explained how 'representing a non-thing as a thing'
is connected with material falsity.
Let us summarize what the texts reviewed in this section establish: ( I ) the
materially false ideas are, by definition, exactly those that are not clear and
distinct - these are two complementary sets of ideas; (2) material falsity is ana
lysed in terms of the philosophically prior notions of clarity and distinctness -
ideas are materially false to the degree that they fall short of clarity and
distinctness; and (3) objective reality falls into three discrete levels: infinite,
substantial, and modal. What ideas reality-represent (or present, or exhibit) in
virtue of their objective reality is some degree of (formal) reality. If, however, an
idea is not clear and distinct, I might 'not be able to tell' how much reality is
represented to me.
'Representation' and the Idea of Cold
Why has this natural interpretation of the texts been overlooked? Since the quoted
texts are mostly from Descartes's responses to objections to the Third Medita
tion, the obvious answer is that Descartes does not make his intentions clear in
that particular text. Arnauld (and some recent commentators) are misled espe
cially by Descartes's treatment of the idea of cold, which is regarded as a kind of
test-case in the Third Meditation. Let us review it. As an example, Descartes
considers this possibility: 'if it is true that cold is nothing but the absence of
heat, the idea which represents it to me as something real and positive deserves
to be called false; and the same goes for other ideas of this kind' (AT 7, 44;
CSMK 2, 30). It is important to remember that, in the context of the Third
Meditation, the meditator, who might start as a Scholastic, has achieved hardly
any Cartesian knowledge concerning corporeal nature - he does not even know
for certain whether it exists. He might go so far as to wonder at this point whether
heat and cold are themselves substances instead of modes. Might they be
substances (or perhaps 'real qualities') emitted by ordinary bodies or associated
with them in some other way? Another alternative is that both heat and cold
are positive modes of bodies. Yet another alternative is that heat is such a posi
tive thing, but that cold is 'nothing but the absence of heat.' Supposing this
last alternative, an apposite example of a false judgment would be to judge
that coldness is a substance associated with a cold body when, as a matter of
(hypothetical) fact, coldness is the absence of heat, a privation. 1 1 In this case, we

could be judging that ' a non-thing was a thing. '
For although, as I have noted before, falsity in the strict sense, or formal falsity, can
occur only in judgements, there is another kind of falsity, material falsity, which
occurs in ideas, when they represent non-things as things. For example, the ideas
which I have of heat and cold contain so little clarity and distinctness that they do
not enable me to tell whether cold is merely the absence of heat or vice versa, or
whether both of them are real qualities, or neither is. And since there can be no
ideas which are not as it were of things, if it is true that cold is nothing but the
absence of heat, the idea which represents it to me as something real and positive
deserves to be called false . . . (AT 7, 43--4; CSMK 2, 30)
The example is doubly hypothetical. For all we know at this point in the Medi
tations, cold m ight be a positive mode of bodies, and therefore a 'thing' (res),
even if it is not a substance in its own right. Furthermore, for all the meditator
knows, the idea of cold m ight even resemble such a positive mode. The point is
that the idea, being obscure and confused, is materially false and does not com
pel us to assent to anything. To remove the obscurity and confusion from the
idea, we need to direct our attention to it, and somehow manage to perceive
simply that we have the idea. This is expressed in Principles I, 68:
In order to distinguish what is clear in this connection from what is obscure, we

must be very careful to note that pain and colour and so on are clearly and dis
tinctly perceived when they are regarded merely as sensations or thoughts. But
when they are judged to be real things existing outside our mind, there is no way of
understanding what sort of things they are. (AT 8A, 33; CSMK 1, 2 17; emphasis
added)
We shall likely fall into error if we use a sensory idea as the basis for any
judgment about what is outside our m inds. If we did go ahead and judge that
the idea of cold was caused or occasioned by an extramental thing, then (given
the hypothesis that cold is a privation) we would falsely judge that the idea
'represents' a thing to me.
Th is is the second way Descartes uses 'represent' (repraesentare). In the first
sense explained above, ideas represent some amount of reality. In the second
sense, ideas represent what we 'refer' them to as causing them. 1 2 Let us call this
sign-representing. Sensory ideas ' actual occasioning causes are generally
very complex because they are always proximately occasioned by characteristic
20 Alan Nelson
brain states, but only remotely and partially caused by such things as the sun,
a dog, heat, etc. There are, therefore, many candidates to choose among when
someone 'refers' a sensory idea to a cause or, in our new terminology, assigns a
sign-representation to it. Sign-representation is always a result of an active
judging and assigning. No idea comes with an intrinsic sign-representation. 1 3
This is an important difference between reality-representation and sign
representation.
One might incidentally get the impression from the way Descartes handles
the example of the idea of cold that sign-representing a non-thing as a thing is
the sine qua non of material falsity. But this is only an example of the way that
materially false ideas fit the correct criterion - providing the material for error.
After completing the Meditations and studying the Principles, Descartes's readers
should know that, in fact, the ideas of both heat and cold are occasioned by
motions in the brain that are themselves caused by motions in the sensed object
and the intervening medium. 1 4 Anyone who does not know this might make
some very bad judgments when assigning sign-representations.
There is a further matter to consider regarding Descartes's use of 'represent.'
It is necessary to distinguish the notion of objective reality already discussed
from another notion for which I shall use the term objective being. Descartes
himself sometimes seems to use these terms interchangeably, but there are in
deed two notions at work. In the First Replies, Descartes writes:
' Objective being in the intellect' ... will signify the object's being in the intellect in
the way in which its objects are normally there. By this I mean that the idea of the
sun is the sun itself existing in the intellect - not of course formally existing, as it
does in the heavens, but objectively existing, i.e., in the way in which object nor
mally are in the intellect. (AT 7, 1 02; CSMK 2, 75)
This might be taken to suggest that the idea of the sun reality-represents the sun
as it exists formally; instead, we should understand this as saying that, when we
use an idea to sign-represent the sun, we can say that the sun has objective being
in the intellect. We might even say that the objective being of this idea is the
sun. This is, however, just a way of talking about how ideas represent the things
that they sign-represent. This text and others like it have no bearing on what we
are calling 'reality-representation.' 1 5
Let us return to the problematic passages from the Third Meditation. Arnauld' s
objections bring the troubles to the fore. He brings out two problems with
the idea of cold. The first concerns what the idea is an idea of, its intentionality.
The second problem concerns the formal source of the idea' s objective reality.
Since objective reality is a kind of reality, it cannot derive from nothing; there
ultimately must be a causal source of the objective reality which itself has
sufficient formal reality to account for that objective reality. Suppose that we
conflate the cause of the idea's objective reality with what the idea is 'of. ' Since
the idea of cold should have objective reality (as all ideas do), if cold is (hypo
thetically) a privation, a non-thing, it seems strange that this idea could, after
all, be the idea of cold. We shall see that Descartes has ready replies, given the
interpretation being developed.
Here is Arnauld' s formulation of the trouble:
But if cold is merely an absence, then there cannot be an idea of cold which repre
sents it to me as a positive thing, and so our author is here confusing a judgement
with an idea.
What is the idea of cold? It is coldness itself in so far as it exists objectively in
the intellect. But if cold is an absence, it cannot exist objectively in the intellect by
means of an idea whose objective existence is a positive entity. Therefore, if cold is
merely an absence, there cannot ever be a positive idea of it, and hence there
cannot be an idea which is materially false. (AT 7, 206� CSMK 2, 145)
Lastly, what does the idea of cold, which you say is materially false, represent to
your mind? An absence? But in that case it is true. A positive entity? But in that
case it is not the idea of cold. (AT 7, 207; CSMK 2, 146)
Descartes' s response is going to be that the idea we call 'the idea of cold, ' as
suming that cold is a privation, is a sensory idea that reality-represents the
modal level because it is a mode of mind. In what sense does it 'represent' a
non-thing as a thing? The answer is that the victim of all this confusion will be
falsely judging that the idea in question is occasioned by a privation. The victim
then attempts to refer the idea to what is not there; he uses it to sign-represent a
privation. So something that reality-represents a thing (the sensory idea) is used
to sign-represent a hypothetical non-thing (the privation). Let us see how
Descartes arrives at this response to Amauld's problem.
Bad Science and the Idea of Cold
Commentators have taken it for granted that there is no question about exactly
which ideas Descartes is talking about in the Third Meditation when he men
tions the ideas of heat and cold. This leaves an important gap in our story be
cause the answers to Arnauld's questions depend on how 'the idea of cold' is
understood. I think it important to bring out three ways to understand 'the idea
of cold' because Amauld and Descartes seem to understand it differently. This
22 Alan Nelson
will also illustrate how bad science can lead to habits of judgment that confuse
our sensory ideas. Descartes's preferred way of taking 'the idea of cold' as the
particular, conscious, sensory experience is saved for last. Suppose, first, that a
particular cold thing affects one's sensory organs; for example, one touches a
piece of ice. It would be natural to say that, in normal conditions, one then has
an idea of that piece of ice, but suppose that attention is concentrated on the
ice 's coldness. In other words, one forms an idea of whatever it is in the ice or
about the ice that causes or occasions one's conscious experience of coldness.
Whatever that turns out to be (a scientific matter), there can be no problem
about the objective reality of this idea. It would reality-represent to me either
modal or substantial reality, depending on which scientific account of the ice's
coldness is adopted. But this much reality could derive from the ice itself, or the
internal motions of the ice, or the intervening media, or the brain, etc. Does this
mean that, if science tells us the ice's coldness is a privation, there is a problem
about the idea's intentionality? If there is no thing answering to the description
'the ice's coldness,' then the idea must actually be occasioned by something
else: a rash judgment, the desire to think about that idea, or even the ice, or
internal motions of the ice, etc. That might lead us to judge that the idea sign
represents one of those things. That, in tum, should lead us to reconsider our
verbally describing this idea as 'the idea of the ice's coldness. ' A scientific
discovery that what we call 'the ice's coldness' is not a thing would make evi
dent the fatuity of the idea we had thought up for ourselves. The verbal descrip
tion 'the ice's coldness' would be misleading in the extreme, and an invitation
to a host of bad judgments - judging a non-thing to be a thing, for instance.
Descartes warns against just this kind of error in Principles 1, 74. whose title
reads: 'The fourth cause of error is that we attach our concepts to words which
do not precisely correspond to real things. '
In short, the highly obscure and confused idea of the ice's coldness need
not be occasioned by any particular thing, though there are facts of the matter
about its actual occasioning cause and about its reality-representation. There is,
therefore, no salient candidate for its sign-representation, and there are various
assignments one might make. And, to repeat, its failure to intrinsically sign
represent does not mean that it fails to reality-represent, although this idea might
be too obscure and confused to enable us to tell whether it reality-represents the
modal or the substantial level. Part of Amauld's difficulty seems to come from
his thinking that, given the hypothesis that cold is a privation, the idea of 'the
ice's coldness' must nevertheless (a) be 'of' a privation, and consequently, (b)
fail to reality-represent. Descartes, however, is committed to neither (a) nor (b).
Let us begin in a second way. If we now begin instead with an idea of coldness
that has been abstracted from ideas of a particular piece of ice, or from ideas of
various cold things, then we would again be judging a non-thing to be a thing if
we supposed that this abstraction had an extramental existence. This possibility

might seem to reinstate Amauld's worries, if it was not indeed part of what was
worrying him. Descartes has a theory of abstraction; is it adequate for handling
the problem at hand? For Descartes, all abstractions are abstract ideas - mental
entities. Many of them are materially false since we are liable to judge incor
rectly (as Platonists do) that the abstraction has extramental being. When we
judge correctly, Descartes thinks we realize that abstractions simply are modes
of thought. 1 6 So the abstraction would still reality-represent. The objective real
ity would correspond to the formal reality of the mode of thought, namely, the
modal level of objective reality. What is the abstract idea of cold an idea of ?
One correct judgment to make would be to judge it a mode of our mind. If we
have made this correct judgment, then the idea can sign-represent the appropri
ate mode of mind. If one has erroneously reified the abstraction with a bad
judgment, then there is no telling what sign-representation the deluded one will
assign to the resulting obscure and confused idea. We now tum to the question
of how sensory ideas become obscure and confused and in what their obscurity
and confusion consists.
Obscure and Confused Ideas
There is yet another interpretation of 'the idea of cold'; it is the one most natural
to twentieth-century readers and, I think, to Descartes himself. Suppose that we
neither derive the idea in question from a scientific hypothesis about a particu
lar object nor produce it by a process of abstraction. Suppose we are experienc
ing the sensory idea itself, the brute 'quale' of cold. We are simply feeling cold
very vividly. In this case we cannot pin the intentionality on anything in a body
that resembles the idea because bodies do not actually have sensory qualities.
Does that mean trouble for reality-representation? No; if I perceive clearly and
distinctly regarding this quale (as described in the above quotation from Princi
ples I , 68), then I will realize that it is a mode of my mind. Therefore, it reality
represents the modal level. If the idea is obscured and confused as a result of bad
judgments, then I might be unable to tell how much reality it represents, but that
does not mean that it fails intrinsically to represent reality.
It might be objected here that it is pointless to insist that ideas intrinsically
reality-represent, even if we cannot tell how much. Why not say instead that, if
we cannot tell how much reality is represented, as we typically cannot when
dealing with obscure and confused ideas, the idea altogether fails to reality
represent? In more conventional terminology, why not say that sensations or
sensory ideas have no objective reality? This question is easily answered if one
is prepared to take Principles I, 68, and similar texts seriously. 1 7 A sensory idea
properly attended to will be clearly and distinctly perceived, provided that no
24 Alan Nelson
false judgments are made concerning it. It is, therefore, a metaphysical fact that
these ideas reality-represent the modal level. A troubling question remains: If
sensory ideas can be clear and distinct in this way, how do they become the
Cartesian paradigm of obscure and confused ideas?
The answer is that the potentially clear and distinct sensory idea is typically
obscured by, and confused with, judgments that are made in connection with it.
The underlying clear and distinct sensory idea remains there, as it were, but it is
overlain by other ideas - usually the results of bad judgments concerning the
underlying sensory idea. The contents of these bad judgments, which are them
selves ideas, augment the bare sensory idea, thereby creating obscurity and con
fusion. In other words, a confused idea is quite literally the result of the confu
sion of more than one idea. And one (or more) of the components of a confused
idea might be a sensory idea. The following important text says as much:
For example, when someone feels an intense pain, the perception he has of it is
indeed very clear, but is not always distinct. For people commonly confuse this
perception with an obscure judgment they make concerning the nature of some
thing which they think exists in the painful spot and which they suppose to resem
ble the sensation of pain; but in fact it is the sensation alone which they perceive
clearly. (Principles I, 46; AT 8A, 22; CSMK 1 , 208)
To understand how this works, consider first that the terms 'obscure' and
'confused' are always to be understood in their technical senses as the comple
ments of 'clear' and 'distinct. ' Descartes makes use of a visual metaphor to
explain that clear perceptions are 'present and accessible to the attentive mind. ' 1 8
Similarly, we see clearly when our eyes are strongly stimulated, even if we can't
tell what is stimulating them. Very clear ideas are attention-grabbers. A clear
idea is obscured if it has to compete with other elements for the attention of the
mind' s eye. The visual metaphor is extended to say that a distinct perception is
'so sharply separated from all other perceptions that it contains within itself
only what is clear' (Principles I, 45). A confused idea, therefore, is a complex of
simples which can be mistakenly regarded as itself simple.
The quotation from Principles I, 46 confirms that sensations can be clear and
distinct, it says they are 'not always distinct, ' not that they cannot be distinct.
This brings us to the principal point I want to make about this text. The expla
nation of confusion given here is that the sensory idea itself is somehow juxta
posed with, or overlain by, or augmented with, a bad, 'obscure' judgment.
We might say that the judgment about the sensation' s resembling something
outside the mind is an obscuring judgment. It diverts our attention from the
sensory idea itself so that it is no longer clear, and hence no longer distinct, in
the technical senses of those terms. This obscuring judgment is separate from,
and additional to, any further false judgments we make on the basis of the now
confused, materially false idea. For example, once a clear and distinct sensory
idea is obscured, one is liable to make bad scientific judgments about its occa
s1omng cause.
We have seen that a sensory idea might be perceived distinctly in Descartes's
technical sense: ' it is so sharply separated from all [others] that it contains
within itself only what is clear' (Principles I, 45). If this idea is then made the
basis of a bad judgment, and the bad judgment augments the original, distinct
sensory idea, then the original idea is now obscured by, and confused with, the
results of the judgment. What kind of obscuring and confusing judgment might
augment otherwise clearly and distinctly perceived sensory ideas? Descartes is
very explicit about this at almost every opportunity.
In the example of pain quoted above from Principles I, 46, the judgment is
' concerning the nature of something which they think exists in the painful spot
and which they suppose to resemble the sensation of pain. ' The general problem
is that ' we easily fall into the error ofjudging [ e.g.,] that what is called colour in
objects is something exactly like the colour of which we have sensory aware
ness. (Principles I, 70; AT 8A, 35-6; CSMK I , 2 1 8). 1 9 Descartes says that we
' easily fall' into this kind of error. We indeed fall into it so easily that almost
every sensory idea we happen to attend to in our thinking will have been ob
scured and confused by having annexed to it this kind of erroneous judgment.
A reader of Descartes might want to object that she does not remember mak
ing a raft of false judgments of this kind. Descartes's response to this reaction
follows immediately in various texts; the judgments that obscure and confuse
our sensory ideas are a residue from a time when we rarely or never exercised
our pure intellects.
In our childhood the mind was so immersed in the body that although there was
much that it perceived clearly, it never perceived anything distinctly. But in spite of
this the mind made judgements about many things, and this is the origin of the
many preconceived opinions which most of us never subsequently abandon. (Prin
ciples I, 47; AT 8A, 22; CSMK 1, 208)
Children make these obscuring and confusing judgments, not because they
are particularly gullible, but because they are exclusively concerned (if young
enough) with pursuing pleasure and avoiding pain. 20 They become cognizant of
how bodily effects (e.g., damage) are correlated with sensations, emotions, and
appetites (e.g. , pain, sadness, and avoidance). Useful judgments of correlation
eventually slip into the paradigmatic mistaken judgments that pleasures and
pains are in the affected parts of the body, and that colours, light, etc., are in the
external bodies that occasion the sensations. 2 1 These mistaken judgments, in
26 Alan Nelson
tum, become so frequent that they are performed very swiftly and can become
so habitual that they cannot be remembered and go unnoticed. Or instead of
performing a new judgment of this kind, we swiftly remember an old one and
reaffirm it.22 The final step in this epistemically sorry (but biologically impor
tant) process is that the potentially clear and distinct sensory ideas become
almost automatically obscured and confused by false judgments about the loca
tion of their occasioning causes or about their degree of resemblance to those
causes.
One result of our cognitive development from infancy to potential systematic
users of pure intellect is, therefore, that our sensory ideas become obscure and
confused. This seldom entails dire consequences for the survival of substan
tially united human beings; on the contrary, judgments involving sensations are
typically efficacious in preserving life. The obscurity and confusion in sensory
ideas can be disastrous, however, for anyone embarking on a philosophical or
scientific project. One of Descartes' s principal goals was to prevent these disas
ters in the search for knowledge. His writings contain countless references to
how his philosophical method enables one to detach from the senses and re
cover what can be clear and distinct. Descartes thought that most bad science of
his time resulted from judgments concerning obscured and confused sensory
ideas. This coheres very naturally with the Fourth Meditation injunction that
when seeking the truth one should never make judgments when not perceiving
clearly and distinctly. Those who have not repaired their cognitive habits left
over from childhood through the process of Cartesian meditation have materi
ally false, obscured and confused, sensory ideas. By definition, materially false
ideas provide scope for error - for what sorts of error do these particular
materially false ideas provide scope? We have already seen two examples. One
might reify a sensory idea: 'Heat is a substance in the fire that bums me. ' One
might also reify an abstraction: 'There is some positive quality, heat, which all
hot things have. '
In this section we have seen wherein the source of obscurity and confusion in
sensory ideas lies. It is our failure to separate the sensory idea as a mode of mind
from ill-informed hypotheses about the occasioning cause of that mode. It is not
something intrinsic to the offending ideas. Our underlying, potentially clear
and distinct, sensory ideas are obscured and confused by the results of overlying
bad judgments. Armed with this theory of how sensory ideas are corrupted, we
can resume our analysis of the exchange between Descartes and Amauld.
The Idea of Cold Again
An important test of any interpretation of these matters must make good sense
of the Third Meditation and Descartes's notorious reply to the problems posed
by Arnauld. More can now be said, for example, about the characterization of
materially false ideas as representing non-things as things. Let us consider again
the passage already quoted which begins one line earlier by characterizing
sensory ideas as confused and obscure:
But as for all the rest, including light and colours, sounds, smells, tastes, heat and
cold and the other tactile qualities, I think of these only in a very confused and
obscure way, to the extent that I do not even know whether the ideas I have of them
are ideas of real things or of non-things. For although, as I have noted before,
falsity in the strict sense, or formal falsity, can occur only in judgements, there is
another kind of falsity, material falsity, which occurs in ideas, when they represent
non-things as things. (AT 7, 43; CSMK 2, 30)
Descartes calls these ideas obscure and confused� we now know this means
that they have been obscured and confused by bad judgments. The bad judg
ments concern the occasioning causes, and therefore the sign-representations
one assigns to these ideas. We can now also see how to interpret the rest of this
passage:
For example, the ideas which I have of heat and cold contain so little clarity and
distinctness that they do not enable me to tell whether cold is merely the absence of
heat or vice versa, or whether both of them are real qualities, or neither is. And
since there can be no ideas which are not as it were of things, if it is true that cold
is nothing but the absence of heat, the idea which represents it to me as something
real and positive deserves to be called false and the same goes for other ideas of
this kind. (AT 7, 44; CSMK 2, 30)
To see how this kind of trouble arises, suppose someone under the influence of
bad science entertains the palpably false hypothesis that cold is an absence of
some positive thing called 'heat.' Let us also suppose this person tries to use the
sensory idea of cold to sign-represent the supposed privation, and proceeds to
make judgments concerning this idea and concerning what it supposedly sign
represents. He might, for example, judge that privations can have causal pow
ers. He would be treating what he first thinks of as a non-thing as a thing. This
train of thought will, of course, result in his sensory idea of cold being obscured
and confused. Since the idea reality-represents a mode (although it is too con
fused for me to realize that at this point), it 'deserves to be called false.' It is, of
course, materially false simply by virtue of its containing ' so little clarity and
distinctness.'
Suppose that someone listening to a different bad scientist decides to have his
sensory idea of cold sign-represent whatever it is that all cold things have in
28 Alan Nelson
common. If he makes the further judgment that it is this abstraction that causes
his sensory idea, the sensory idea will become obscured and confused. Observ
ing this, we might describe his plight by saying that his obscured and confused
idea came to sign-represent to him a non-thing as a thing. 23
Amauld posed another problem:
What is the idea of cold? It is coldness itself in so far as it exists objectively in the
intellect. But if cold is an absence, it cannot exist objectively in the intellect by
means of an idea whose objective existence is a positive entity. Therefore, if cold is
merely an absence, there cannot ever be a positive idea of it, and hence there
cannot be an idea which is materially false. (AT 7, 206; CSMK 2, 145 )
Descartes might well have replied that, if cold is an absence, there is really no
idea of cold after all, there is only the idea that we call 'the idea of cold.' He
instead goes back to basics: '[material falsity] arises solely from the obscurity
of the idea - although this does have something positive as its underlying sub
ject, namely, the sensation involved' (AT 7, 234; CSMK 2, 164). This says that
an idea is materially false just in case it is obscured and confused, that is, not
clear and not distinct. When a sensory idea is clear and distinct, we inevitably
form a true judgment; namely, that the idea is a mode of our mind. The clear
and distinct sensory idea, therefore, is what Descartes here calls the ' underlying
subject. '24 This is one sense in which materially false ideas are positive regard
less of what we judge them to sign-represent. This much positivity is required,
of course, since God allows us to have materially false ideas.
Descartes also writes:
Thus if cold is simply an absence, the idea of cold is not coldness itself as it exists
objectively in the intellect, but something else, which I erroneously mistake for
this absence, namely, a sensation which in fact has no existence outside the intel
lect. (AT 7, 233; CSMK 2, 163; emphasis added)
This can now be interpreted as saying that on the (palpably false) hypothesis
that cold is an absence, the idea that we call the idea of cold will typically be the
obscured and confused sensory idea. Returning to the question of intentionality,
what should we now say this idea is an idea of? The most plausible way of
talking is to say that the idea is of itself. If it were a clear and distinct sensory
idea - that is, if one realized it was simply a mode of one's mind - then it would
be true. And the truth expressed would be simply that one had that sensory idea.
It is interesting in this regard that Amauld himself came to think that every
perception involves a reflective awareness of what the perception is 'of. '2 5 Since
the materially false idea also contains other obscuring and confusing elements,
there might be contexts in which we would choose to say the idea is 'of' one or
more of those other elements. It is not easy, however, to think of many contexts
for which anything other than the underlying sensory idea would be a plausible
choice.
Arnauld and Descartes
Many think that Arnauld got the better of Descartes in their exchange concern
ing sensory ideas. This is hard to square with the fact that Arnauld himself
seems to have been well satisfied with what Descartes wrote! He seems to have
been satisfied, moreover, with something very close to the interpretation of
Descartes developed above, as shown by the official statement on sensory ideas
from the Logic.
Our ideas of colors, sounds, smells, tastes, cold, heat, heaviness, and other sensible
qualities, as well as our ideas of hunger, thirst, bodily pain, and so on, are confused
and obscure ideas. And the reason these ideas are confused can be explained as
follows:
. . . the mind was unsatisfied with j udging merely that there was something external
to it causing these sensations, although in this judgment there would have been no
error. The mind went on to judge that there was something in the external object
exactly like the sensations or ideas which the object occasioned ... Such a transfer
ence results in the confused and obscure ideas we possess ofsensible qualities, for
the mind has added falsejudgments to what nature reveals to us. (emphasis added)
The doctrine here is that26 obscured and confused sensory ideas are 'results' of
false judgments. What of sensory ideas that we do not obscure and confuse
in this way? Arnauld has again accepted Descartes's account. In On True and
False Ideas, he cites with approval the assertion in Principles I, 68 that sensory
ideas can be clear and distinct, and concludes: 'pain, color and other similar
things are known obscurely and confusedly only when we refer them to bodies
as if they were there modifications. ' 27 One might question whether Arnauld
thought, against Descartes, that particular extramental objects like dogs
literally have a kind of being in our thought,28 but there seems little room for
doubt that he came to understand Descartes, and to agree with him, concerning
the falsity of sensory ideas. 29 I take it to be a virtue of the interpretation
presented in this essay that it shows how this meeting of the two great minds
was effected.
30 Alan Nelson
Notes
This essay is based on ideas presented at the 1 994 conference on Amauld at the
University of Toronto. I benefited from the discussion on that occasion. I have also
received valuable help from Vere Chappell, Keith DeRose, Paul Hoffman, Jeremy
Hyman, Elmar Kremer, Lex Newman, Lawrence Nolan, Calvin Nonnore, and Kurt
Smith.
Standardized terminology is essential here, but neither Descartes nor subsequent

commentators have settled on one. I use 'sensory idea' to refer to the mode of
mind present when a human being is sensing or having a sensation. I use
'sensation' to refer to what happens to a human being when a motion of the pineal
gland occasions a sensory idea.
2 An excellent introduction to some of the problems can be found in M. Wilson,
Descartes (London: Routledge 1 978), 1 0 1 -20.
3 English translations of Descartes are taken from the three volumes of Cottingham,
Stoothoff, Murdoch, and Kenny, which give cross-references to volume and page
numbers in AT.
4 Descartes's Third Meditation characterization of materially false ideas as
representing non-things as things is discussed below.
5 There are many similar texts. See for example the Fifth Meditation (AT 7, 433)
and the 2 May 1 944 letter, probably to Mesland (AT 4, 1 1 5- 1 6 and 1 1 7).
6 On both of these points see AT 7, 234. Judgments of this type are explained later
in this essay.
7 ' I have also made it quite clear how reality admits of more and less. A substance is
more of a thing than a mode; if there are real qualities or incomplete substances,
they are things to a greater extend than modes, but to a lesser extent than complete
substances; and, finally, if there is an infinite and independent substance, it is
more of a thing than a finite and dependent substance. All this is completely self
evident': AT 7, 1 30.
8 Objective reality is something that is intrinsic to ideas. If an idea is not clear and
distinct, we can easily be mistaken about how much objective reality it has. This is
important in what follows.
9 But isn't the objective reality of my idea of the sun caused by the sun itself? No;
the Third Meditation specifies that the cause in question is the 'efficient and total'
cause. The 'total' cause obviously includes much more than the sun itself.
Descartes emphatically reiterates the point in the 3 1 December 1 640 letter to
Mersenne: ' It is certain that there is nothing in an effect which is not contained
formally or eminently in its EFFICIENT AND TOTAL cause. I added these two
words on purpose. The sun and the rain are not the total cause of the animals they
generate' : AT 3, 274; caps in original. We can add than the sun is not the total
cause of the objective reality belonging to the idea of the sun. The efficient causal
connection between the two is exceedingly complex, even when psycho-physical
connections are ignored.
10 The apt term 'present' is best eschewed here to avoid confusion with Wilson 's
related, but quite different, distinction between 'presentational representation' and
' referential representation, ' in M. Wilson, 'Descartes on the Representationality
of Sensation, ' in Central Themes in Early Modern Philosophy, J. Cover and
M. Kulstad, eds. (Indianapolis: Hackett 1990).
11 Descartes's real position, of course, is that heat is a sensation occasioned by
certain motions. Although the sensation of heat tends to be occasioned by external
objects with violent internal motions, and cold by less violent motions, it would be
a mistake to conclude that cold is in any sense an absence of heat.
12 The Latin is referantur. See for example AT 7, 233.
13 In the case of clear and distinct ideas, however, the relevant judgment is always
correct, and the assignment that gets made is inevitably the true one. This is the
case because of the special, divinely guaranteed, connection between clear and
distinct ideas and truth. One might want to use the term 'intrinsic' for this kind of
sign-representation, but this question is merely terminological.
14 See for example Principles I, 69-7 1, and IV, 196-8.
15 It is possible that Descartes and Arnauld did not see eye to eye on this matter. See
note 28, below.
16 See Principles I, 57-9 and 62, which make Descartes's view explicit.
17 See also Principles I, 66, which is similarly explicit. Also relevant is the famous
claim from the Notae: 'The ideas of pain, colours, sound and the like must be all
the more innate if ... our mind is to be capable of representing them to itself ... '
(AT 8B, 359). In so far as these ideas are innate, they are bestowed upon us by
God and subject to clear and distinct perception. And, in the Sixth Meditation, we
find: ' And since the ideas perceived by the senses were much more lively and
vivid and even, in their own way, more distinct than any of those which I
deliberately formed . .. ' (AT 7, 75; emphasis added).
18 Principles I, 45. I am going to ignore the difference between applying the notions
of clarity and distinctness to perceptions, on the one hand, and to ideas, on the
other.
19 Descartes makes this point throughout his career. See also The World, ch. 1 and 2,
and the Regulae (AT 10, 423).
20 Descartes held this view as early as 164 1 (see AT 3, 424) and later incorporated it
into his theory of the passions (see AT 4, 604-6).
21 But doesn't a toothache 'present itself' as if it is in the tooth? No: we judge that it
is in the tooth, though this kind of judgment can be so habitual that we do not
remember having made it a moment afterward. Part of the illusion that pains
present themselves as if in the painful body part might come from such purely
32 Alan Nelson
mechanical reflex actions as reaching to soothe the offended part.

22 Descartes explains these psychological mechanisms in a slightly different context
at AT 7, 439. Even though we might not notice that we are consciously making
these judgments, they must remain conscious, Wilson, in 'Descartes on the
Representationality of Sensation, ' raises some good questions about the degree of
consciousness of such habitual judgments.
23 It is also possible to mistakenly judge that the confused and obscure idea, which
actually really reality-represents the modal level, instead reality-represents the
substantial level, for example, of some imagined fluid-like, subtle body that makes
cold things cold. That error could also be called judging a non-thing to be a thing
since the shift from the modal level to the substantial level is analogous to the
shift from nothingness to the modal level. This, however, does not seem to be
Descartes's point in the Third Meditation.
24 ' Underlying [fondement],' added in the French (AT 9A, 18 1), supports my
interpretation as it suggests that an obscuring judgment 'overlays' the potentially
clear and distinct sensory idea.
25 S. Nadler, in Arnauld and the Cartesian Philosophy of Ideas (Princeton, NJ :
Princeton University Press 1989), 1 18-22, has textual citations and a helpful
account of Amauld's doctrine of virtual reflection.
26 The A rt of Thinking, by Antoine Arnau Id, translated and edited by J. Dickhoff and
P. James (Indianapolis: Bobbs-Merrill 1964), 66--7. Amauld has absorbed
Descartes's doctrine of material falsity so thoroughly that he can even call
confused and obscured sensory ideas 'false ideas.' He writes: 'These confused
ideas may also be called false ideas for the reason we stated [in the passage quoted
in the text] ' : Dickhoff and James, 22.
27 E. Kremer, trans. and ed., On True and False Ideas, by Antoine A rnauld
(Lewiston: Edwin Mellen 1990), 139.
28 For an interpretation of Amauld as a 'direct realist' see Nadler, A rnauld and the
Cartesian Philosophy of Ideas. For an examination of how Arnauld thought that
the objects of knowledge are 'immanent' to the knower, see Kremer, trans. and
ed., On True and False Ideas, xxiii-xxvii. See P. Hoffman, ' Descartes on
Misrepresentation, ' Journal of the History of Philosophy (forthcoming), for an
interpretation of Descartes that makes him rather like Kremer's Amauld in this
respect.
29 More evidence that Amauld was satisfied on this score comes from an
examination of the subsequent letters he wrote to Descartes (3 June 1648,
AT 5, 184-9 1, and July 1648, AT 5, 2 1 1- 15). These contain substantial and
wide-ranging queries about the various issues, but nothing about the theory of
ideas.
3
Arnauld and the Modern Mind
(the Fourth Objections as Indicative
of Both Arnauld' s Openness to and
His Distance from Descartes)
P E T E R A. S C HO U L S
There are important issues underlying the questions arising from Amauld' s
interaction with Descartes on the latter's Meditations, an interaction known as
the Fourth Objections and Replies. Here are two of the seventeenth century's
brightest people who respected and liked each other, one of them (Descartes)
understanding the other well, the other (Amauld) doing his very best to under
stand but, on crucial points, falling short. The ultimately interesting matter is
that Amauld falls short in his understanding of Descartes because he is still
situated in a tradition whose very rejection forms an integral part of Descartes' s
position.
It is no doubt true that, in his self-conscious attempt to reject tradition,
Descartes was responding to a shift in point of view already well in progress. By
the time Descartes started to write, Galileo had provided firm support for the
view of Copernicus, which placed human beings in a more peripheral position
in the universe than that to which they had grown accustomed, thus at least
relativizing an important aspect of the tradition and, in the process, attracting
the wrath of its guardians, the church's Inquisition. And Thomas Hobbes had
for several years been an intimate of Mersenne's intellectual circle, where, with
out anything like the Cartesian fanfare called Discourse on the Method, he had
accepted something very like this method and was beginning to apply it in his
work on optics as well as in civil philosophy. 1 Descartes, however, makes it very
clear that, unlike his predecessors and contemporaries, he aims to reject, not
this or that aspect of tradition, but all of it. Unlike his contemporaries, he
34 Peter A. Schouls
universalizes method, making it apply to all objects of thought - a method which

dictates initial rejection of that to which it is applied. His attempt announces
modem philosophy's intention 'to rise above tradition, above the particularity
of any historical location' and so, 'purified of historical contingency' by reject
ing it as prejudice, to speak only the absolutely trustworthy and universally
valid language of reason. Today, we tend to recognize this language for what it
is: 'local custom masquerading as universal speech. '2
In this vacuum left by a rejected tradition, Descartes comes to his Archimedean
point. It is to this Archimedean point, and to the 'rational scheme' built on it,
that Mersenne asks Amauld to respond. Amauld does so in a language some
times different in sound, often incommensurable in signification, from that used
by Descartes. But Amauld's is as well a language of local custom, of longer
standing and wider spread than Descartes's, equally masquerading as universal
speech. As he looks at Descartes's scheme, Amauld withdraws - perhaps recoil
ing at the terror of the vacuum of perfect freedom? - and tries to fit the Cartesian
position within the confines of the established tradition. It is on this unsuccess
ful attempt, thus not on the ultimate but on the penultimate matter of interest,
that I focus in this essay on Amauld and the modem mind.
II
It is no small feat to publish, at the youthful age of twenty-nine, what turns out
to be a document one of whose chief statements remains the subject of vigorous
debate today: Arnauld's Objections to Descartes's Meditations contains the clas
sical formulation of the charge that the attempt to validate reason through rea
son fails because of circularity. 3 This, however, is hardly sufficient to warrant
the judgment that Amauld ' is one of the most important intellectual figures of
the seventeenth century,' that he is 'the great Amauld.'4 If such an evaluation
were to rest on the character of his work as revealed in these no doubt ' able,'
even 'astute,'5 Objections, it might well be inappropriate. Those of our contem
poraries who do make this judgment have broader grounds for it, not the least of
which are La Logique ou / 'art de penser (1662) and the Nouveaux Essays de
geometrie (1667). Primarily on the basis of works such as these and half a
year's personal acquaintance, the young Leibniz wrote to the Duke of Hanover
about the sixty-two-year-old Amauld, describing him as 'the world-famous
M. Amauld.'6 Whether Leibniz's as well as our contemporaries' judgments are
warranted on these grounds is not my concern in this essay, for I restrict my
attention here almost entirely to the Objections.
It seems reasonable to say that (possibly among other criteria) a person
deserves the accolade of being a century's ' most important intellectual figure,'
Amauld and the Modem Mind 35
first, through insightfully bringing to bear traditional wisdom on the most

crucial contemporary concerns or, second, through taking distance from
traditional wisdom and presenting one's contemporaries with a point of view so
fundamentally different from traditional wisdom that it changes the very con
cerns which are now thought to be of major importance. In the Meditations,
Descartes accomplished the second of these. In his Objections, Amauld attempted
the first but because - in contrast, for example, to his later Port-Royal collaborater,
Pascal - he did not at this time discern the profound shift in perspective and
ambition which Descartes introduced, he achieved neither. The traditional wis
dom he drew upon, Amauld, no doubt ably and astutely, made relevant to mat
ters no longer among the most crucial for his world. If the latter judgment ap
pears excessively harsh, it has at least the virtue that it can serve as a heuristic
device to bring into clear focus the young Amauld's relationship to 'the modem
mind' as expressed in the Meditations.
By 'the modem mind' I refer to the stance in which a thinker believes it
possible and necessary to start de nova, in a situation where at least historical
and cultural and, at specific stages, even physical contexts are ruled to be non
essential, or even hindrances, to the enterprise of developing scientia and the
moral and political practices built thereon. It is a stance in which neither thought
nor action is necessarily related to the Good (as in Plato) or to God (as in Augus
tine), 7 in which the validity of both thought and action is ultimately founded on
immanent subjective consciousness rather than grounded in an objective order.
If the pre-modem ( classical or medieval) mind is characterized by relation, the
modem mind is distinguished by isolation. Whereas Augustine's attitude is typi
cally pre-modem, Descartes's is distinctively modem. It is true that the projects
of both Augustine and Descartes are radically reflexive. For Augustine, how
ever, this reflexivity is encapsulated in given relationships with God, other hu
man beings, and the physical world. For Descartes, in contrast, reflexivity is an
exercise in disengagement from all of these and is to establish the thinker's
thoroughgoing autonomy. 8 When, believing that there was an intrinsic affinity
between (on the one hand) Descartes's sum res cogitans and its immanent idea
of God and (on the other) Augustine's Deum et animam scire volo, Port-Royal
Jansenists accepted Descartes's methodology as the correct way of thinking,
they made more than a single error - errors which were foreshadowed in
Amauld's Objections to Descartes's Meditations. These errors do not preclude
astute observations about, for example, material falsity or the Eucharist; on
matters such as these Amauld excelled in relating traditional wisdom to con
temporary concerns. But, though not without interest to Descartes, these are not
the sort of items of central concern to him - and they are certainly not the items
of his position which embody the thrust of the Cartesian revolutionary, modem
36 Peter A. Schouls
stance. Items of the latter kind, though they do not escape scrutiny in the Fourth
Objections, there remain beyond Amauld's grasp.
Already in these Objections, Amauld's lifelong preoccupation with issues of
a moral and theological nature predominates,9 but neither their morality nor
their theology here reveals innovation, whether in statement or in application.
Their statement is hallmarked by a brand of orthodox safety; their application is
an attempt to assimilate Descartes to the orthodoxy of the tradition as shaped
initially by Augustine - an enterprise Amauld would hardly have embarked
upon had he discerned the revolutionary drive of Descartes's work.
Amauld' s objections are of four kinds: (i) He raises points of ostensible or
possible agreement with Descartes which coincide with real agreement. Among
these are (a) the insistence that an act of will (in the form of an act of attentive
ness) is required if one is to achieve knowledge, 10 (b) that time is discontinuous
or atomic, 1 1 and, as a consequence of (b), (c) that in reality there is no distinc
tion between creation and conservation. 1 2 (ii) There is ostensible or possible
agreement which does not coincide with real agreement. Here, there are matters
such as the (a) indicators of epistemic individualistic autonomy, 1 3 (b) insistence
on separating understanding, belief, and opinion, 14 which for Amauld is impor
tant especially because it allows for separation of epistemological error and moral
wrong, 1 5 and (c) identification of Augustine's and Descartes's cogito. (iii) There
is ostensible disagreement coinciding with real disagreement, as on (a) the cir
cularity of Descartes's argument 1 6 and (b) the extent of Descartes's doubt. 1 7 (iv)
Finally, there is ostensible disagreement not (at least according to Descartes)
coinciding with real disagreement on issues such as (a) material falsity 1 8 and (b)
the Eucharist. 1 9
This incomplete20 list is sufficient to indicate Amauld's thoroughness: he had
read the Meditations - and, as is clear from passing comments, also the Dis
course on the Method - with extreme care. Agreements as well as disagree
ments, ostensible or real, can therefore tell us much about the young Amauld' s
position with respect to the modem mind. As I now develop this theme, I shall
restrict myself to two of the four kinds of objections raised: I shall disregard the
first and last kinds and focus on the second and third, on some of the ostensible
points of agreement which do not coincide with real agreement, and on some of
the ostensible points of disagreement which really are disagreements.2 1
III
What ostensible agreements tell us. (a) The second paragraph of Amauld's
objections can hardly be less than a deliberate echo of the opening paragraph
of the Meditations and, likely, of the opening paragraph of Part II of the
Discourse. In the latter, Descartes relates the circumstances under which this
crucial part of the Discourse had come into being: not subject to clamours of
war and politics, diversions of conversation, cares about personal well-being,
and troubles of passion, he could say he was ' completely free to converse with
myself about my own thoughts' (AT 6, 11). In the former, he set out to accom
plish the task of rightly founding the sciences, once he could say that 'I have
expressly rid my mind of all worries and arranged for myself a clear stretch of
free time,' that he was 'quite alone' and so could devote himself to the task
'sincerely and without reservation (/ibere).' Amauld, likewise, set himself to
his task of commenting on the Meditations with 'a calm mind' which had to be
' free from the hurly-burly of all external things' because it needed 'the leisure to
consider itself - something which . .. can happen only if the mind meditates
attentively and keeps its gaze fixed upon itself.' In the case of Descartes, these
phrases indicate an epistemology and methodology characterized by individual
istic autonomy. The Discourse paragraph which begins by stating the need and
achievement of solitary self-centred reflection ends with an intimation of dis
trust of one's cultural situation and natural condition - a distrust given concrete
resolution in the immediately following paragraph: " regarding the opinions to
which I had hitherto given credence, I thought that I could not do better than
undertake to get rid of them, all at one go.' The task to be accomplished during
the free and detached time of the First Meditation is (in words from its first
paragraph) 'the general demolition of my opinions' which (the second para
graph adds) can be accomplished by 'undermining' 'the foundations' or 'the
basic principles' 'on which all my former beliefs rested.' Amauld's phrases, in
contrast, are meant to convey anything but individualistic epistemic autonomy.
Instead of distrust, there is trust: trust in the (more than eucharistic) presence of
God, in the value of philosophical and theological traditions, in the genuine
willingness of Descartes to submit himself to the discipline of these traditions,
in the responsibilities which attend friendship ('you have done me a kindness'
and, now, 'since you command, I must obey' are words from the two opening
paragraphs which set their tone as much as any others). In short, it is the duties
imposed by Mersenne's friendship and by scholarly responsibility which make
Amauld take Descartes's work utterly seriously, a fact which he tries to make
clear from the outset by adopting some of Descartes's language, which is there
fore used here to indicate that the task of responding is accepted with the kind of
wholeheartedness which responsibly shuts out distracting influences. Where
Descartes's stance will soon tum out to be founded on the potentially solipsistic
cogito ergo sum, Amauld's position - in spite of these echoes from Discourse
and Meditations - is from the very beginning one that can perhaps be best
captured by the necessarily other-involving respondeo ergo sum. 22 For Descartes,
38 Peter A. Schouls
truth and goodness are to be discovered and developed by the individual on

self-determined foundations; for Amauld, they are to be given by, or discovered
and developed in relation to, a Platonic-Aristotelean reality translated by
Augustine and Thomas into the being of or emanations from the Christian God.
For Descartes, change demands the revolutionary act of wholesale uprooting,
which requires that the old - if it is to survive at all - is to stand the test of newly
articulated criteria discovered within the individual' s reason. For Amauld. change
is the reformer's act in which the new - if it is to stand at all - is to survive the
test of traditional criteria, a test passed when the new can be seen harmoniously
to fit (or even be a retelling of) long-sanctioned positions.23
The contrast I have drawn is that between the modem and pre-modem mind.
Is it warranted by the material so far adduced? I believe it is fair to say that this
material at least indicates that direction. Through cumulatively relating the other
points on which I will focus, the grounds for this contrast will become stronger,
hence its validity more assured.
(b) Both Amauld and Descartes insist on separating understanding, belief,
and opinion; Amauld takes there to be no disagreement here between himself
and Descartes or, more broadly, between Descartes and the tradition. To bring
home the latter point, Amauld quotes at length from Augustine and then con
cludes that 'Descartes, prudent man that he is, will readily judge how important
it is to make the distinctions just outlined' (AT 7, 217; CSMK 2, 217). When
Descartes responds, he gives no indication that he disagrees with the particulars
of Amauld's and Augustine' s distinctions, and he ostensibly makes a conces
sion as a consequence of apparent agreement with respect to what is for Amauld
a crucial distinction, that between epistemological error and moral wrong. There
is, however, implicit disagreement which, as we shall see later, becomes in part
explicit in their conflict about the circularity of the argument for the validation
of reason.
The details of Amauld' s distinctions in Augustine's words, and Descartes's
(at this juncture unstated) points of disagreement with them, are as follows. 24
First, Amauld holds that one understands if one ' grasps something by means of
a reliable reason [qui certa ratione aliquid comprehendit].' For Descartes, but
apparently not for Amauld at this stage, this is only half the story of what it is to
understand. It omits the crucially important grasping of foundations: these are
grasped per se not per aliud; hence, there is in their case no grasping 'by means'
of ' a reliable reason,' for they are understood immediately rather than mediately.25
Second, one believes, says Amauld, if one is 'influenced by weighty authority to
accept a truth even though he does not grasp it by means of a reliable reason.'
For Amauld, following Augustine, believing can be a permanent state in which
legitimately to rest.26 For Descartes, believing can be legitimate so long as it is
for the time being, because authority, though it can be weighty, is at best second
best; hence, to rest in belief is to abrogate the basic human responsibility for
autonomy. Third, for Amauld, one commits the 'very grave fault' 'of being
opinionated if he thinks he knows something of which he is ignorant.' Descartes
agrees, but would impute this fault to all who are content to rest in belief on the
grounds of authority, for all such persons believe authorities because they think
they know that these authorities know - and (unless the authority is known to be
God) that they cannot know.
I n the response (AT 7, 218; CSMK 2, 172) to this lengthy invocation of Au
gustine, Descartes points Amauld to his reply to the Second Objections, where,
he claims, the maxim ' we should assent only to what we clearly know' is 'ex
plained quite explicitly' as 'always subject to the exception of "matters which
belong to faith and the conduct of life".' In addition, he claims that he has 'also
given advance warning of [this restriction] in the Synopsis.' Both claims are
disingenuous. As to the first, he advances as his position in the Second Replies
that, whether with respect to faith or life, ' we commit a sin by not using ...
reason correctly' (AT 7, 148; CSMK 2, 106) and the correct use of reason with
respect to obscurities of faith or unresolved complexities of life is: to accept
doctrine or pursue action only on the ground of good reasons for such accept
ance or pursuit. He there refers to Part III of the Discourse, where it is made
very clear, on the one hand, that life is such that we cannot always know while
yet we must act, and that, therefore, until we know, prudence dictates that, for
the time being, we conform to the most moderate manifestations of the status
quo in politics, morality, religion, and everyday concerns. In the meantime, on
the other hand, we may never relinquish the freedom to review this conformity
(AT 6, 24-27; CSMK 2, 1 23-4): 'For since God has given each of us a light
to distinguish truth from falsehood, I should not have thought myself obliged to
rest content with the opinions of others for a single moment if I had not in
tended in due course to examine them using my own judgment; and I could not
have avoided having scruples about following these opinions, if I had not hoped
to lose no opportunity to discover better ones, in case there were any.' As to the
second, the advance warning in the Synopsis: 'in order to show how much I
respect M. Amauld's judgement' Descartes entered a disclaimer in the Synopsis
to the effect that, in the Fourth Meditation, he did 'not deal at all with sin,
i. e. the error which is committed in the pursuit of good and evil.' But in the
Fourth Meditation itself he quite emphatically leaves in place the explicit
references to error in extra-intellectual pursuits when he writes about the 'true
and good/verum & bonum' (p. 40), 'error and sin/fallor & pecco' (p. 41), and
'falsity and wrong/falsitates & culpae' (p. 42). The Meditations, as Descartes
insisted time and again, are themselves of instrumental value only. They are to
40 Peter A. Schouls
be the foundation for the intellectual pursuits which result in the sciences, which
themselves are to determine the quality of life as reason applied through me
chanics (power over nature), medicine (power over the body), and morals (power
over the passions). The Meditations are to be the foundation for activity which
is to result in paradise regained: freedom from labour, pain, and evil - not by
faith in and grace from God, but by trust in the efficacy of human reason and
actuality of human generosity.2 7 Amauld, in sharp contrast, is at this time pre
occupied with the kind of Augustinian view about the human state in which
human beings unassisted by divine grace are pretty well incapable of good and
prone to all evil, so much so that, if God were to bestow grace but were to leave
it up to individuals to make use of it, the grace bestowed would not be effica
cious - a view which underlies the first of Amauld' s controversial writings
which was to be published two years after the Objections to the Meditations, the
anti-Jesuit De lafrequente Communion. On this score of human rational ability
assisted only by human generosity to produce moral good and so, in principle,
overcome sin and obviate culpability, Amauld and Descartes are miles apart, as
far apart as is the Augustinian mind from the modem. 28
(c) From this perspective it is not difficult to show that the ostensi ble
agreement on the nature of the foundation of philosophy is in fact Amauld' s
misjudgment of the spirit of Descartes's project. Amauld' s very first comment
on the Meditations is that he finds it 'remarkable' that Descartes 'has laid down
as the basis for his entire philosophy exactly the same principle as that laid
down by St Augustine,' namely, that 'you yourself exist' (AT 7, 197-8; CSMK
2, 139).29 Descartes responds only by saying: 'I shall not waste time here by
thanking my distinguished critic for bringing in the authority of St Augustine to
support me' (AT 7, 154 ; CSMK 2, 219). This curtness may be indicative of
Descartes' s characteristic irritation when critics complement him on (or charge
him with) stating what others have already said; or it may be his way to pass
over a point whose greater explication will bring out his departure from tradi
tion in colours too stark for safety let alone the comfort of respect.30 Where, for
Augustinian Amauld, the Archimedean point of his philosophizing would be
found in belief which relates the embodied thinker to the transcendent God, for
Descartes this point lies in the subjectivity of the independent and isolated im
manent cogito - a point so firm that not even God can shake it. In other words,
for Amauld, philosophy is carried out in dependence on God, while for Descartes
we have in the cogito the unilateral declaration of philosophy' s independence
from God. In Gareth Matthews's words, Augustine's (and, through his identifi
cation with him, Amauld's) 'inquiry is conducted from the standpoint of faith'
so that 'even in his most rational and systematic moods, Augustine never tries
to undermine all appeal to outside authority,' while Descartes 'makes himself
Amauld and the Modem Mind 4 1
his own authority'; and where, for Augustine, 'some things cannot even be
understood unless they are first believed,' for Descartes knowledge is strictly a
' "do-it-yourself" achievement' in which 'systematic doubt' serves the isolated
individual's total 'reconstruction of knowledge.' In Augustine, says Matthews,
there is no such 'rational reconstruction of knowledge,' let alone a cogito which
is 'the independent foundation for reconstructing knowledge.'3 1
Now for my last two points, which will produce greater relief in the picture so
far sketched.
IV
What stated disagreements tell us. The assertions concerning (d) the circularity
of Descartes's argument and (e) the extent of Descartes's doubt express real
disagreements.
(d) Amauld' s formulation of 'the circle' (at AT 7, 2 1 4; CSMK 2, 1 50) is
probably the best-known part of his Objections:
I have one further worry, namely how the author avoids reasoning in a circle when
he says that we are sure that what we clearly and distinctly perceive is true only
because God exists.
But we can be sure that God exists only because we clearly and distinctly per
ceive this. Hence, before we can be sure that God exists, we ought to be able to be
sure that whatever we perceive clearly and evidently is true.
Amauld's objection is inevitable, coming as it does from a person who at this

point considers understanding to be always a matter of ' grasping something by
means of a reliable reason.' In that case, if 'whatever we perceive clearly and
evidently' is true 'only because we clearly and distinctly perceive' that a trust
worthy God exists (here is the reliable reason), and if the existence of a non
trustworthy God has been posited (thus removing the reliable reason) and has
thus cast doubt on the very trustworthiness of the criteria of clarity and distinct
ness in all cases of purported knowledge, then circularity is inescapable.
In his brief rebuttal (AT 7, 245-6; CSMK 2, 1 7 1 ) Descartes again refers
Amauld to the much longer response in his Second Set of Replies (AT 7, 1 40-
6; CSMK 2, 1 00-5). In this response it is clear that Descartes works with a
distinction which makes illegitimate Amauld's use of 'whatever' in 'whatever
we perceive clearly and evidently,' namely, the distinction between notitia
(' awareness of first principles' - which, once one is aware of them at all, are
always fully known and hence never remembered), and scientia ( which is sys
tematic knowledge ultimately founded on such principles, 'knowledge of those
42 Peter A. Schouls
conclusions which can be recalled when we are no longer attending to the

arguments by means of which we deduced them'). I n other words, Descartes
distinguishes between things known immediately through intuition and things
known mediately through deduction. Once, through focusing on the immedi
ately known cogito, Descartes has shown (first) that, with respect to what is
immediately known, the criteria of clarity and distinctness remain intact whether
or not there exists a God and whether or not this God is a deceiver, and (second)
that a trustworthy God exists, he can reinstate trust "in the criteria of clarity and
distinctness with respect to whatever is mediately known, for it is only the
hypothesis of the existence of a deceiving God which called into question
the efficacy of these criteria in the context of mediation. But first showing the
absolute trustworthiness of some things32 known immediately is a necessary
prerequisite for the second move.
Two decades later, Amauld himself articulates this doctrine of the two kinds
of knowledge and their corresponding human capacities in the terminology of
' intellection' and 'understanding. '33 But if we were to import this later distinc
tion into the debate we are now considering, I suspect Amauld would have had
a different objection to the legitimacy if not validity of Descartes' s argument.
The crucial foundational item intuitively known is that if I think, then I am. It is
knowledge with which God cannot tinker. Thus for the foundation of philoso
phizing it holds that it is irrelevant whether or not God exists as good or evil.
Showing this absolute trustworthiness of foundational thinking, showing that
the foundation for philosophy lies in pure subjectivity and does not require au
thentication from anything transcending the individual thinker' s act of con
sciousness, is an emancipation of this act from all that transcends it - an eman
cipation which firmly launches Western philosophy on the adventure later named
' the Enlightenment.' From all we have so far seen, it would seem safe to con
clude that Amauld would refuse to be a passenger, let alone a member of the
crew, aboard a ship with this destination. This conclusion stands even when we
realize that Descartes's act is one of partial rather than full emancipation.
For it is emancipation of thinking, not of being. The very existence of
the person qua thinking and extended substance (if only because of the discon
tinuous nature of time) is, for both Amauld and Descartes, dependent on the
providence of God. But autonomous thought, first in metaphysics, then in the
pure, and finally through the applied sciences, determines the quality of this
existence. Autonomous thought holds out the promise of paradise regained
in this world, through human, not divine activity. Descartes's radicalism is a
watershed: before him, paradise was lost in a distant past to be regained by
select humans in a non-earthly context through the merit of Christ; after him,
paradise lies in an earthly future, to be regained by the work of human reason
coupled with hum an generosity. That to which in his Objections Amauld was
blind, his younger contemporary, and later defender and collaborator, PascaP 4
eventual ly saw clearly : 'I cannot forgive Descartes. In all his philosophy he
would h ave been quite willing to dispense with God. But he could not help
granting him a flick of the forefinger to start the world in motion; beyond th is
he has no further need of God. ' 3 5
Amauld' s early disregard of the distinction between knowledge per se and
per aliud may have been the screen that prevented him from seeing the radicality
of the Cartesian cogito. His use of Augustine' s distinctions among understand
ing, belief, and opinion led him to impute circularity to Descartes ' s argument;
this charge of circularity possibly kept him from more carefully examining the
foundation of Descartes' s system, thus misjudging it to the extent that he pro
nounced it Augustinian.
( e) There now remains the matter of doubt. It is perhaps telling that Amauld' s
worry about the extent of Descartes' s doubt is the first point raised under those
' which m ay cause difficulties to theologians' (AT 7, 2 1 4- 1 5 ; CSMK 2, 1 5 1 ): 'I
am afraid that the author' s somewhat free style of philosophizing, which calls
everyth ing into doubt, may cause offence' ; for example, 'where we find the
clause "since I did not know the author of my being", ' Amauld suggests as a
substitution ' the clause "'since I was pretending that I did not know . . . " ' The
Meditations, says Amauld, are 'dangerous for those of only moderate intelli
gence ' - a judgment which, given its locus, would seem to refer to theologians.
Descartes ' s response? 'I completely agree' (AT 7, 247; CSMK 2, 1 72). The
activity of pretending, wh ich is the use o f non-corpore a l imagination in
philosophy, 3 6 is absolutely crucial for Descartes. It is that which allows for the
generation of hypotheses, hence is necessary if any progress is to be made in
philosophy as in any systematic thinking. Descartes already shared Amauld' s
language before the latter raised the point. As we read in the penultimate para
graph of the First Meditation, 'I think it will be a good plan to tum my will in
completely the opposite direction and deceive myself, by pretending [/ingam]
for a time th at these former opinions are utterly false and imaginary. I shall do
this until the weight of preconceived opinion [ praejudiciorum] is counter
balanced and the distorting influence of h abit no longer prevents my judgment
from perceiving things correctly. ' The agreement here between Descartes and
Amauld is strictly superficial, at best covering the judgment that the average
theologian is none too bright, and that therefore it is best to use the language of
make-believe.
Amauld seems oblivious of what this pretending in fact is meant to accomplish.
It allows for efficacious universa l doubt. And although, for both Descartes
and Augustine, doubt leads to awareness of the existence of the doubter, the
44 Peter A . Schouls
resemblance is again superfici al. F or Augustine, th is doubt underm ines

scepticism in that it leads it into self-contradiction. For Descartes, the conse
quences of th is doubt are intended to be both deeper and more extensive: it
is to authorize the cogito as the immovable foundation for the autonomous con
struction of scientia; 37 it is to free the mind from all h abit or custom, from
the ' habitual opinions' which 'keep coming back, and, despite my wishes . . .
capture m y belief, which i s a s it were bound over to them a s a result o f long
occupation and the law of custom ' (as the opening sentence of the penultim ate
paragraph of the First Meditation has it); it is to free the th inker from any
influence of culture and location; it is to throw the thinker back on strictly
individual resources, and as such it is the first sounding of Kant' s Enlighten
ment call for each to be self-em ancipated from self-incurred tutelage. 3 8 Counter
balancing a ' supremely good' God with a ' m alicious demon of the utmost power
and cunning (summe potentem & callidum)' has the effect of freeing the thinker
from both . 39 Here is the typically modem mind.
Conclusion. In a recent essay, 40 N icholas Jolley writes th at, ' for A m auld,
Descartes' philosophy offered the best philosophical support of the Christian
faith; it provided the best arguments for the existence of God and the spirituality
of the mind. ' This accurate statement about Amauld indicates the point I h ave
argued: Amauld' s is not the modem mind. To the contrary, he is so enveloped
in pre-modem Augustinianism that he fails to recognize modernity ' s character
when it faces him. He believes it sufficiently innocuous to allow its incorpora
tion into Christian traditional thought with few adjustments in doctrine and
tone. In this respect Amauld differed from many of his contemporaries, both
Protestant and Catholic.
It did not take long for Descartes to become persona non grata at m aj or
Protestant Dutch universities. At Utrecht, Regius was forbidden to te ach
Cartesianism within two years of the publication of the Meditations and its
Objections and, two years after that, Utrecht' s city council pronounced a formal
ban on all discussion of Descartes in print. Within a few years, the University of
Leiden followed suit. In both places Cartesianism was seen as anti-Christian, in
part because of its assertion of the limitlessness of human freedom. In Rome,
Descartes ' s works were placed on the Index in 1 663 , and in 1 67 1 French
universities were forbidden to teach Descartes by royal decree. Both Roman
church and French king acted under the strong influence of Jesuit thought, which
held that Descartes ' s position undermined central tenets of the Catholic faith.
Throughout, Am auld rem ained a steadfast champion of Descartes as defender
of the faith. It would seem that Dutch Protestants as well as Italian and French
Jesui ts had more insight than Arnauld into Descartes the revolutionary
challenger of tradition (although it would be anachronistic by a century to say
that they consciously were this revolutionary's anti-revolutionary opponents).
In the subsequent century, Amauld's enervated orthodox type of Cartesian ism
was alive and well in the staid confines of the French Academy. In the mean
time, Cartesianism of the radical revolutionary kind - that promulgated by
Descartes when he did not feel constrained to say less than he would through
fear for loss of personal safety or respect - returned to French soil, in part through
the influence of the writings of Locke as these were embraced by the equally
revolutionary philosophes such as d'Alembert, Voltaire, Diderot, and Condoret. 4 1
Notes
I am indebted to my colleagues Vivien Bosley and Martin Tweedale for helpful

discussions of various aspects of this essay.
This is not to say that Hobbes consistently and exclusively appl ied a single
scientific method to political and physical science. The former was at least in part
structured by his negative reaction to the rhetorical tradition of classical and
Renaissance humanism. On this point see Quentin Skinner's "'Scientia civilis' in
Classical Rhetoric and in the Early Hobbes, ' in Political Discourse in Early
Modern Britain, Nicholas Phillipson and Quentin Skinner, eds (Cambridge:
Cambridge University Press 1 993), 67-93 .
2 The phrases in quotation marks in this paragraph are from Jeffrey Stout, Ethics
after Babel: The Languages of Morals and Their Discontents (Boston: Beacon
Press 1 988), 74.
3 Amauld ' s was not the first charge of circularity to which Descartes replied. His
first reply came in the Second Set of Replies, in response to objections Mersenne
supposedly collected from various philosophers and theologians but which were
most likely largely Mersenne's own. As we shall see, in his response to Amauld,
Descartes refers to the Second Set as containing grounds to counter Amauld's
charge. Neither of these repl ies have been convincing to all phi losophers.
Beginning with Gassendi in 1 644 (see The Selected Works of Pierre Gassendi,
Craig Brush, ed. and trans. [New York: 1 972], 204 ), the debate has remained
continuous until our day - as evidenced, for example, by Willis Doney 's selection
of articles on the topic published in Eternal Truths and the Cartesian Circle (New
York : Garland 1 987). The most recent interesting discussion of 'the circle' is that
of Georges Dicker in his Descartes: An Analytical and Historical Introduction
46 Peter A. Schouls
(New York and Oxford: Oxford University Press I 993 ), ch. 3.

4 The first of the quoted phrase is from the opening sentence of Steven Nadler's
review of A.-R. Ndiaye's la Phi/osophie d 'Antoine Arnauld (Paris: Vrin 199 1) in
the British Journal for the History of Philosophy I / I ( 1993), 138-4 1; the second
is from Genevieve Rodis-Lewis's 'Descartes Life and the Development of His
Philosophy, ' in The Cambridge Companion to Descartes, John Cottingham, ed.
(Cambridge: Cambridge University Press 1992), 2 1-57, 47.
5 John Cottingham writes of Amauld as 'Descartes' astute critic' who 'ingeniously
proceeds to show . . . ': Descartes (Oxford: Blackwell 1986), I I 2. His comments
echo those of Bernard Williams, who speaks of Descartes's " able critic Amauld' :
Descartes: The Project of Pure Enquiry, (Sussex: Hassocks I 978), I 60.
Cottingham praises Arnauld for showing the difficulty with Descartes's argument
as it moves from (i) I can doubt that I have a body, through (ii) I cannot doubt that
I exist, to (iii) therefore I who doubt am not a body. (Williams has said that
Arnauld's formulation reveals this argument's 'embarrassing resemblance to the
masked man fallacy' - (p. 1 12). Williams's accolade comes in the context of
Amauld's objection to Descartes' proposal that God can be the efficient cause of
his own existence.
6 Gottfried Wilhelm Leibniz: Samtliche Schriften und Briefe (Darmstadt and Berlin,
1923- ), ii. 1, 230.
7 I contrast the (neo-)Platonic/Augustinian rather than the Aristotelian/Thomistic
mind with the modem mind because Amauld approached Descartes from an
Augustinian rather _than a Thomistic position. The contrast between Augustine and
Descartes is different from that between Thomas and Descartes - which is not to
imply that the Thomistic stance is an expression of the modem mind. By the late
thirteenth century, (neo-)Platonic Augustinianism had been supplanted by the
Aristotelian epistemology as the main bulwark of scholasticism. As in Descartes's
position, and as distinct from that of Augustine (e.g., Confessions, VI. 5 and City of
God, X.2), Aristotelian/Thomism treats human reason as essentially independent
of divine grace. But, unlike Descartes, it emphasizes the need for intimate
familiarity with traditional authorities, the chief of these being Aristotle himself.
Thus the difference between the latter two traditions does not lie primarily in the
emancipation of human reason from God, but from the tradition of rational inquiry.
Hence an oversimplified statement of these various relationships is that the
modem mind has affinities with the Augustinian mind in its view of the (for
Augustine, relative) unimportance of traditional thought but sharply differs from it
on the necessity of a relationship between human reason and divine grace, and has
affinities with the Thomistic mind on the (for Thomas, partial) irrelevance of a
relationship between human reason and divine grace, but sharply differs from it on
the importance of traditional authority.
Arnau Id and the Modem Mind 47
8 For a similar contrast drawn in recent literature between Augustine and Descartes,
see Charles Taylor. The Sources of the Self: The Making of the Modern Identity
(Cambridge, MA: Harvard University Press 1 989), 1 1 5-49.
9 This predominance is stressed in Ndiaye's La Philosophie d 'Antoine Arnauld. For
example, p. 1 1 : Amauld's writings are essentially polemical, but 'toujours en
rapport avec ses preoccupations morales et theologiques.'
10 For Amauld's use, see AT 7, 1 97, 2 1 4, 2 1 8; for Descartes's, see AT 7, 229, 246.
For an account of the relation between attention and will in Descartes's
Meditations see Peter A. Schouls, Descartes and the Enlightenment (Edinburgh:
Edinburgh University Press; Kingston and Montreal: Mc Gill-Queen's University
Press 1 989), 90--5 . That Amauld retained this position in later life is clear from
The Art of Thinking, e.g., Part IV, ch. 1 , the opening paragraphs.
1 1 For Amauld, see AT 7, 209; for Descartes, AT 7, 49.
1 2 For Amauld, see AT 7, 2 1 0, 2 1 2; for Descartes, AT 7, 49.
1 3 See Arnau Id's second paragraph, which echoes both the opening paragraph of the
Meditations and the opening paragraph of Part II of the Discourse on the Method.
1 4 AT 7. 2 1 6- 1 7.
1 5 AT 7, 2 1 5 .
1 6 AT 7, 2 1 4.
1 7 AT 7, 2 1 5, related to 2 1 6.
1 8 AT 7, 206--7.
1 9 AT 7, 2 1 7- 1 8.
20 There are other matters, two of which I indicated in note 5, above.
2 1 Whether the items mentioned under (4) are disagreements not coinciding with real
disagreement is an aspect of two other essays in this volume: Alan Nelson's 'The
Falsity in Sensory Ideas: Descartes and Amauld,' and Elmar Kremer's 'Amauld's
Interpretation of Descartes as a Christian Philosopher.'
22 Even when, two decades later, in The Art of Thinking, there are cogito-like
statements, these function quite differently from Descartes's cogito. In Part I,
ch. 9, it occurs only as an example of a clear idea: 'The ideas that each has of
himself as a thinking being is very clear . . . We cannot claim that there is no such
thing as thinking substance, since to do so we must in fact think.' And, in Part IV,
ch. 2, it functions as part of an illustration of the process of analysis, where the
illustration is in terms of the question ' Is the soul of man immortal?': ' In the first
place we note that thinking is the essential characteristic of the soul and that, since
doubt itself is a thought, the soul can doubt everything else without being able to
doubt whether it thinks.'
23 Amauld's defence of Cornelius Jansen is the best illustration of this point: he
found the defensibility of Jansen's position primarily in its comporting with that of
Augustine.
48 Peter A. Schouls
24 All quotations in th is paragraph are from Arnauld ' s reproduction of Augustine at

AT 7, 2 1 6- 1 7 .
25 In The Art of Thinking, Arnauld ' s position i s much closer to Descartes' s o n this
point. See again the first chapter of Part IV, which begins with two sentences
characterizing the first aspect of knowledge : ' When we know any maxim to be true
in itself - that is, when the evidence offered by the maxim suffices to convince us
of its truth - then we know by intellection. First principles are known in such a
fashion. ' He had, by that time, read Clerselier's copy of the manuscript of
Descartes' s Rules for the Direction of the Mind, in which the notion of knowledge
per se, of items grasped intuitively rather than discursively, occupies a prominent
position. Arnauld - although he does not adopt Descartes' s word ' intuition ' but
instead uses ' intellection ' - here explicitly accepts this notion of non-discursive
knowledge which, for him, as for Descartes, is knowledge of items foundational to
the sciences. In the sentence immediately fol lowing those quoted above, Amauld
introduces as a second item that which, in the Objections, he had (in Augustine ' s
formulation) stated as the first and only item sufficient b y itself t o characterize
knowledge and belief: ' If the maxim alone does not convince us of its truth, we
require some further basis for accepting the maxim as true; and this basis is to be
found either in authority or in reason.' Since the ' further basis' can under certain
circumstances be ' authority , ' Arnauld ' s pre-modem stance is sti ll intact on this
point at this later stage of his career. He is still closer to Augustine than he is to
Descartes, for it is only the former of these for whom ' understanding ' can be ' the
reward of faith ' so that one does ' not seek to understand in order to believe, but
bel ieve that you may understand ' (Homilies on the Gospel ofJohn, 29. 6. John
G ibb and James Innes, trans., in A Select Library of the Nicene and Post-Nicene
Fathers, Vol . 7, Philip Schaff, ed. (New York. 1 888), 7-452.
26 See the last few sentences of the preceding note.
27 On 'generosity ' see The Passions of the Soul, Art. 1 53 : ' true generosity . . . has
only two components. The first consists in his knowing that nothing truly belongs
to him but this freedom to dispose his volitions, and that he ought to be praised or
blamed for no other reason than using this freedom well or badly. The second
consists in his feeling within himself a firm and constant resolution to use it well
- that is, never to lack the will to undertake and carry out whatever he j udges to be
best. To do that is to pursue virtue in a perfect manner. ' (AT 1 I . 445--6; CSMK 1 ,
3 84)
28 In the Preface to the Principles, Descartes reiterates his position on the nature of
the moral rules of the Discourse and, by implication, the need for him to be
concerned with moral good and evil in addition to epistemic truth and falsity . He
there writes that, in the Discourse, 'I summarized the principal rules of logic and
of an imperfect moral code which we may follow provisionally while we do not yet
know a better one. ' (AT 98, 1 5 ; CSMK 1 , 1 86)
Amauld and the Modem M ind 49
29 Amauld' s quotation is from Augustine's On Free Will, II, ch. 3. There are other
passages in Augustine' s works which contain a cogito-like statement. There is On
the Trinity, X, 1 O; e.g., in Philosophy in the Middle Ages: The Christian, Islamic,
and Jewish Traditions, 2d. ed, Arthur Hyman and James Walsh, eds. (Indianapolis:
Hackett, 1 973), 7 1 . And there is the City of God, XI, 26; e.g., in The Essential
Augustine, Vernon J. Bourke, ed.; (Indianapolis: Hackett, 1 964), 33. Moreover,
Amauld was not the only one to draw Descartes's attention to this similarity; it
was pointed out by both Mersenne and Colvius, not with respect to the
Meditations but with respect to the Discourse, where the cogito appears in the
form / am thinking, therefore I exist. Mersenne's reference concerns (say Adam
and Tannery) 'sans dout le fameux passage Civil. Dei, lib. XI, cap. 26'; Descartes
gives it short shrift: 'ii ne me semble pas s'en servir a mesme usage que ie fais'
(25 May 1 637; AT 1 , 376). Colvius's reference, as well, is the City of God.
Disingenuously (given his correspondence with Mersenne) Descartes intimates
ignorance of the passage and again distances himself from its import: ' I am obliged
to you for drawing my attention to the passage of St Augustine relevant to my / am
thinking, therefore I exist. I went today to the library of this town to read it, and I
do indeed find that he does use it to prove the certainty of our existence. He goes
on to show that there is a certain likeness of the Trinity in us, in that we exist, we
know that we exist, and we love the existence and knowledge we have. I, on the
other hand, use the argument to show that this I which is thinking is an immaterial
substance with no bodily element. These are two very different things' ( 1 4
November 1 640; AT 3, 247; CSMK 3, 1 59).
30 Perhaps it is both of these. This is, after all, the third time that a critic confronts
him on this matter; and it is certainly the case that Descartes distances himself
from this identification. On both points, see the preceding note.
3 1 These statements are from pp. 36, 1 33-4, 1 42-3, and 1 45 of Gareth B. Matthews's
admirable Thought 's Ego in Augustine and Descartes, (Ithaca and London: Cornell
University Press I 992). See as well what Matthews has to say (on pp. 1 50 and
1 90) about the appearance of a cogito-like statement in Augustine's Soliloquies.
32 I use the plural ('some things') because, both in the Meditations and in the
restated argument of the First Book of the Principles of Philosophy, there are two
ways in which the cogito is established as first item of certain knowledge: first, in
terms of volition, and, second, in terms of intellection. See my 'Human Nature,
Reason, and Will in the Argument of Descartes's Meditations,' in Reason, Will,
and Sensation: Studies in Cartesian Metaphysics, John Cottingham, ed. (Oxford:
Clarendon Press 1 994 ), 1 59-76.
33 See note 25, above.
34 Pascal's Lettres provinciales ( 1 656--7), written in defence of Arnauld's Jansenist
views, are probably the result of colaborative work at Port-Royal among Pascal,
Amauld, and Pierre Nicole (Arnauld's co-author of the Art of Thinking).
50 Peter A. Schouls
35 Blaise Pascal, Pensees. H . S . Thayer, ed. (New York : Washington Square Press
1 965), Section 2, 77, p. 26.
36 The role of non-corporeal imagination in Descartes is a fascinating though largely
ignored topic. I am in the process of completing a monograph - provisionally
cal led Descartes on human nature and progress in science - in which this role is
the central focus of attention.
3 7 See Matthews' s Thought 's Ego, 1 7 1 -4 .
3 8 The opening paragraph o f Kant's What Is Enlightenment? reads: ' Enlightenment is
man ' s release from his self-incurred tutelage. Tutelage is man ' s inability to make
use of his understanding without direction from another. Self-incurred is th is
tutelage when its cause l ies not in lack of reason but in lack of resolution and
courage to use it without direction from another. Sapere aude! "Have courage to
use your own reason ! " - that is the motto of enlightenment. ' The translation
quoted is that of Lewis White Beck, appended to his translation of the
Foundations of the Metaphysics of Morals (New York : Macm illan 1 98 5 ), 83 .
3 9 Apparently as unaware as Amauld of Descartes 's contribution on this score, Peter
Gay has cal led this balancing act the new intellectual style of the eighteenth
century Enlightenment. See Peter Gay, The Enlightenment: An Interpretation,
Vol . 2 : The Science ofFreedom (London: Weidenfeld and Nicolson 1 970), x i i i, 1 60.
40 Nicholas Jolley, ' The Reception of Descartes ' Ph ilosophy, ' in The Cambridge
Companion to Descartes, John Cottingham, ed. , 3 93-423 , 40 1 . Jolley ' s essay
contains a broader statement of several of the points of my conclusion.
4 1 This sentence expresses the central theme of my Descartes and the Enlightenment.
4
Arnauld and Scepticism :

Questions de fait and
Questions de droit
T H OM A S M. LEN N ON
Amauld was viewed in his time as an extraordinary genius of great learning

and insight. This was a view shared by Pierre Bayle, who expressed it in so
many words. 1 Amauld's initial celebrity has not been fully retained by the his
torical sieve, however. Although Bayle's assessment of him might be justified
just by Amauld's long polemic with Malebranche, for example, there is only
one argument nowadays connected with his name: the argument concerning the
circularity of Descartes's reasoning about clear and distinct ideas and the exist
ence of God in the Meditations. Otherwise, there is no concept, no innovation or
rupture that history has come to associate with him. There is no equivalent for
h im of Descartes's cogito, for example, or Pascal's wager. If there is an excep
tion to this untagged intelligence, to this failure of history to find a niche for the
name of Amauld, it would perhaps be the distinction between questions de fait
and questions de droit. In the period, at least, this distinction would have been
the likeliest conditioned response to the name of Arnauld.2
I first explored the fait-droit distinction fifteen years ago in an article on
Jansenist scepticism. 3 My thesis there is that a sceptical outlook was logically,
temperamentally, and strategically characteristic of Jansenism: logically, in that
it followed from their views on grace; temperamentally, in that it sat well with
their rejection of speculative theology; and strategically, in that it obscured their
unorthodox views. As a Cartesian, even as a Cartesian fellow-traveller, Arnauld
ought to have been the arch-opponent of scepticism; but especially with respect
to Amauld's strategy, the article draws attention to a distinction of sorts be
tween the question de fait of what Arnauld asserted and the question de droit of
what he was committed to. The thesis is that Arnauld's use of the fait-droit
distinction itself committed him to scepticism.
52 Thomas M. Lennon
Here I expand that thesis, both historically and structural ly, especial ly with
respect to Amauld's use of the distinction in response to criticism from Bay le
and from Pascal. My thesis is that the distinction implies that the church's infal
libility is, at best, hypothetical in the sense that it is infallible with respect to a
proposition P if, but only if, in asserting its infallibility it means P, but that we
can never be sure that it means P. The result is an open-ended challenge to
church authority. None of the church's pronouncements need be binding be
cause of open questions about what any of them might mean. For anyone relying
on church infal libility to overcome doubt, scepticism in matters of religion is
the result. This result is very important in the larger, philosophical context. To
take a notable example from the seventeenth century, Gassendi's objection to
Descartes was that, even if clarity and distinctness were reliable criteria of the
truth of propositions, we would stil l need a means of determining which propo
sitions are clear and distinct. To put it more general ly, a standard does not guar
antee its own application. The issue raised by Amauld, then, is scepticism in a
sense wider than the religious context. But first, some background.
In 1653, the bul l Cum occasione of Innocent X condemned the famous Five
Propositions that everyone took to be definitive of the Jansenist position on
grace - everyone, that is, but a group from among the Jansenists headed by
Amauld. Faced with the inconsistent triad of ( l ) papal infallibility, (2) papal
condemnation of the propositions, and (3) the Augustinian, hence orthodox,
character of the propositions, this group chose to deny that the Pope had con
demned the propositions in the sense in which they accepted them. Following
the suggestion of Nicole, apparently, Amauld distinguished between questions
de fait, about which the church is fallible, in this case the question whether
certain propositions are to be found in a certain book, or are held by certain
people, and questions de droit, about which the church is infallible, in this case
whether certain propositions are heretical. Amauld claimed to cede on the
question de droit, but maintained a respectful silence on what he took to be the
question de fail. The Pope had no doubt condemned something, and correctly,
but he had not condemned Jansenism - at least not as Amauld understood it,
which, of course, is what (he thought) it really was. What the Pope thought was
Jansenism is not Jansenism, according to Amauld. 4
On 12 December 1661, the Jesuits published and defended a thesis, Assertiones
Catholicae, in which they argue that the Pope has the ' same infallibility' that
Christ had, that the Pope is ' an infallible judge of controversies of faith, ... as
wel l in questions de fait as de droit. Therefore, since the Constitutions of
Amauld and Scepticism 53
Innocent X and Alexander VII, we may believe with a divine faith that the book
entitled the Augustinus of Jansenius is heretical, and the Five Propositions
drawn from it, are Jansenius's and in the sense of Jansenius condemned. ' 5 Ferrier,
for instance, argued that, if indeed the Pope had not understood Jansenius in
condemning him, he might not have understood Augustine, for example, in
approving him.
In reply, Arnauld argued, ad verecundiam, citing even people like Bellarmine,
that the Pope is not infallible with regard to matters de fail, and, ad absurdum,
that the Jesuits might as well have argued that the Pope is unable to sin as that
he is unable to err with respect to matters de fail - history shows the one to be
about as likely as the other. The Jesuit position amounts to idolatry; it deifies the
Pope, endowing h im with omniscience. According to Amauld, the role of the
Pope is to interpret what has already been revealed by God. Nicole replied to
Ferrier that the questions he raises must be handled on a case-by-case basis,
relying on reason; history shows the church sometimes to have been right, some
times in error concerning these questions defait. 'One cannot therefore know in
general whether it be lawful or not to affirm that an author has been ill under
stood by the Church, since it depends on the particular reasons that induce one
to say it. ' 6
Not irrelevantly, Bayle indirectly accepted the Jesuits' premise and supported
the thesis I defend in my article, at least with respect to religious knowledge. He
argued that there is no church infallibility. For him there is no infallibility de
fait (as he thought most Catholics other than the Jesuits admitted), and there
fore no infallibility at all. However certain we may be about some matter defait,
we might always be mistaken.
We shall return to Bayle's position below. Meanwhile, the upshot of Amauld's
use of the distinction would seem to be, not just fallibility, but religious
scepticism: we must suspend belief with regard to matters of faith because we
literally do not know what to believe. Before we can determine whether a
believed proposition is true or false, we must determine what it means, but this
is always problematic. No appeal to an external criterion of religious knowledge
such as the Pope or church councils could be certain, even if the criterion were
certain, because no use of the criterion could guarantee its own relevance to an
individual belief. (That every criterion is crippled in this way is the case
that sceptics had traditionally argued against the dogmatists - for example, as
noted above, Gassen di against Descartes.) Given Amauld' s distinction, the
applicability of the criterion must always be infected with matters de fait: inter
pretation, vagueness, ambiguity, etc. But since Amauld did accept this external
criterion, the effect of his distinction and the extreme uses to which it was put
was really to question the possibility of any kind of religious knowledge, and
54 Thomas M. Lennon
perhaps of any knowledge at all. One observer remarked that 'the [Jansenists]
are come now generally to disavow, not only the Popes, but all human
Jnfallibititie. This is one of the last refuges they have made use of against their
adversaries. 7 7 The extreme use of this argument, however, was made by others.
Popkin points to three works of 1688, 1700, and 1757 where the following
argument was employed. 'The Pope and no one else is infallible. But who can
tell who is the Pope? The member of the Church has only his fallible lights to
judge by. So only the Pope can be sure who is the Pope; the rest of the members
have no way of being sure. '8 Yet even the knowledge by the Pope that he is the
Pope is a matter de fail. This is no artificial question. The more general version
of the argument concerns the characteristics of the true church, how they are to
be determined and applied, etc. In this dispute, of course, Arnauld and Bayle
were as prominent as anyone addressing the issue.
II
Soon after its appearance in 1685, Arnauld's Reflexions philosophiques et

theologiques was reviewed by Bayle in the Nouvelles de la Republique des
Lettres.9 As seems to have been typical of all his reviews, Bayle's account of
Arnauld's book is complete, fair, and accurate, but also generally sympathetic.
Alas, on one point, and on one point only, among many others, Bay le' s fairness
led him to defend Malebranche. He went so far as to accuse Arnauld of quib
bling in an effort to taint Malebranche's religious orthodoxy.
Bayle found eminently defensible Malebranche's view that pleasure makes
those who enjoy it happy while they enjoy it. The various kinds of pleasure, and
their moral status, are distinguished on the basis of their efficient cause, but not
their formal cause. There may be no difference in the pleasure itself caused by
the sovereign good and by some passion. In addition, the intrinsic distinction
Arnauld tries to draw between spiritual and bodily pleasures is not going to be
very effective in converting voluptuaries, who are ensconced in their pleasures. 1 0
Nadler has shown that the difference between Arnauld, on the one hand, and
Bayle and Malebranche, on the other, is with respect to the intentionality of
sensations, particularly pleasure and pain. 1 1 For Bayle and Malebranche,
pleasure and pain are non-intentional - at least considered in themselves. For
Arnauld, all conscious states are intentional, so he is in a position to claim
that there is an intrinsic difference between physical and spiritual pleasures
based on the intentional relation that each bears to its object. (Without this
distinction, he thinks, concupiscence would be justified, and one of the main
proofs of original sin would be upset.)
Arnauld was not long in responding. He produced an anonymous twenty
three-page pamphlet dated 10 October of the same year. 1 2 It is no quibble, he
Arnauld and Scepticism 55
says, to point out that the term 'happy' in Malebranche's position can be taken
in two senses - one popular, according to which those are happy who think
themselves happy, the other philosophical, according to which the only happi
ness is enjoyment of the sovereign good - which is the only sense in which he
took the tenn. 1 3 He goes on to insist that, for the proper evaluation of his argu
ments, two questions must not be confused: 'One, whether he has understood
[bien pris] his opponent's sense. The other, whether in the sense he has under
stood his opponent, he has refuted him.' 14 The first is a question de fait, but of
the Five Propositions to which Arnauld conveniently reduced Malebranche's
view on sensory pleasure, three are 'independent of this question de fait; that is,
concerning them one can not claim that he has not attacked the true position of
Malebranche.' Amauld continues, confusingly, as follows: 'thus, in order for
what [Bayle] says about his [Arnauld's] attack to be true, viz. that what the
attack says about sensory pleasure is neither evident nor reasonable, his [Bayle's]
criticism must extend also to these three propositions.' 1 5
Bayle's reply to Arnauld's invocation of the fait-droit distinction makes a
perhaps obvious but important point. On the question de fait whether Arnauld
properly understood Malebranche, Bayle thinks that Malebranche did not mean
by happiness what Amauld means, and thus that what Amauld says proves
nothing against him. Amauld fails to prove anything against Malebranche be
cause he fails to understand him. Bayle does not develop the point here, but the
drift is towards saying that the question defait must be presupposed if there is to
be refutation, or even disputes, at all, or perhaps (or there to be communication
of any sort. (The point becomes even more obvious when put in terms of confir
mation and agreement. ) However obvious or not, the point comports with Bayle's
position that restricting the church's authority to questions de droit eliminates
its authority altogether.
On the other hand, Amauld did not claim scepticism on the question de fail
concerning the more important set of Five Propositions, although he was re
spectfully silent on the matter. On the contrary, he must have believed, or have
been committed to believing, that the Five Propositions condemned by the Pope
are not to be found in Jansenius - otherwise he would have accepted the con
demnations without any fait-droit distinction. This is on the assumption, of
course, that Arnauld was sincere in his acceptance of church infallibility - an
assumption open to question. In any case, the point of relevance with respect to
scepticism is that the denial of church infallibility concerning a question does
not mean that it cannot be answered at all. The upshot may be that infallibility
takes on a hypothetical character: if the church says 'P' and means P by 'P,'
then P. That the questions whether the church says 'P,' and whether it means P
by 'P,' are questions de fail does not upset the infallibility of its pronouncement
that P. Still, this concession may mean only that, when the church speaks, what
56 Thomas M. Lennon
it says is true, not that we thereby are ever in a position to know what is true.
That is, the church's infallibility is by itself not sufficient to overcome doubt.
Amauld' s own responses to the church' s pronouncements on questions of grace
show that an answer to the question de fail making them sufficient would never
be forthcoming.
III
Arnauld' s response to the church' s attempted condemnation of Jansenius' s

doctrine is, apart from his respectful silence, essentially the following: 'I agree
that what you say is true, indeed must be true; but I don ' t agree with what you
say you mean. ' However much they may question Amauld' s motivation, com
mentators have been remarkably uncritical in taking his statement at face value
as embodying a transparently valid distinction. Yet, if it is not to be dismmissed
as a mere anticipation of Lewis Carroll's Humpty Dumpty who made words
mean whatever he wanted them to, the statement ought at least to be examined
on its own terms for its philosophical significance.
To begin, then, Amauld' s response to Bayle is confusing, for the fait-droit
distinction ought to apply to all five propositions that Arnauld attributes to
Malebranche, and not to just three of them. (Refer again to note 15 for these five
propositions, although what the propositions were is not crucial to the argument
here.) That is, of each of the propositions we should be able to ask whether it is
to be found in Malebranche, i.e., whether Arnauld and Malebranche meant the
same thing by the same words - in short, whether Amauld has understood
Malebranche (question de fait); and second, we should be able to ask of each of
the propositions whether in the sense intended by Amauld it is true (question de
droit). The only way to separate three of the propositions (the second, fourth,
and fifth propositions) as independent of the question de fait would be to make
their attribution to Malebranche irrelevant. But this would be paradoxical, since
Bayle did not undertake to defend whatever Malebranche said; that is, he was
not in a position like the Catholic' s position vis-a-vis the church, for the Catho
lic had to defend propositions simply because they were uttered by the church.
Rather, he was defending a proposition that was incidentally held by Malebranche.
So perhaps Amauld' s point, the most charitable interpretation of his remark, is
that in three of the propositions he was making points independently of what
Malebranche may have held. A charitable interpretation of what he thought he
was doing with respect to the other two then becomes problematic, however. For
it seems that he was refuting whatever Malebranche said, an impression not
without support from elsewhere in his exchange with the Oratorian. 1 6
To call the question whether a certain proposition is to be found in a given
book a question defait is, moreover, highly misleading. Consider Charles Dillon
' Casey' Stengle, who used to entertain sportswriters with stories involving unu
sual and often preposterous facts from baseball history. To quell the dismay
greeting his stories, his stock response was: ' Ya could look it up! ' The question
defait concerning Jansenism is not a Casey Stengle question, since the Augustinus
of Jansenius was obviously available to everyone for quotation. The real ques
tion that Arnauld tried to separate from the faith is what any quoted proposition
might mean. By contrast, the question de droit concerned its truth, to which the
church had infallible access in so far as it concerned faith or morals. Although
this was not the settled significance of the distinction, which tended to be used
in a bewildering variety of ways too numerous to be investigated here, Arnauld's
argument focused on the distinction in just these terms, i.e., fail - meaning;
droit - truth.
IV
Recall the church' s condemnation of the famous Five Propositions, which

assert, very roughly, that the grace merited by Christ is necessary and sufficient
for salvation. Arnauld claimed repeatedly, in so many words or in effect, that he
' sincerely condemned the five condemned propositions, in whatever book they
may be found, without exception.' 1 7 If what Arnauld said was true, did he con
demn Jansenius' s book or any part of it? He did so if the Five Propositions are in
Jansenius' s book; but he did not condemn Jansenius' s book or any part of it
under that description, because he restricted his condemnation to a question de
droit, and he took it to be a question de fait whether the Five Propositions were
in Jansenius' s book. Did Arnauld think that as a matter of fact the Five Proposi
tions were in Jansenius' s book? This is the question on which Arnauld pro
fessed a respectful silence, but, for the reason given above, it is hard to believe
that he did not, although the question was a matter of dispute at Port-Royal.
Certainly, the church must have thought that Arnauld, by abstaining on the
question de fait, agreed that the propositions were in Jansenius, but did not
take them to be heretical the way the church did. For Arnauld failed to explain
what he condemned as a Catholic, and what by contrast was in Jansenius. Nor
could he, in practical terms, have done so. For assuming that Jansenius is clear,
at least to smart theologians, and that Arnauld is a smart theologian, anything
relevance to the disputed questions of grace that he found in Jansenius is the
same as what the smart Catholic theologians would have found and would
have been condemned by the church. More mischievously, or perhaps more
realistically, even if Jansenius is not clear, any proposition on grace in Jansenius
as interpreted by Arnau Id would have been condemned. Arnau Id's refusal to
58 Thomas M. Lennon
recognize this while insisting on remaining in the church raises a question

repeatedly raised in the period, viz. the question of the sincerity of Amauld' s
intentions with respect to Catholic orthodoxy.
If it is hard to believe that Amauld did not think that the Five Propositions
were in Jansenius, it is also hard to believe that the church condemned a view
that it erroneously thought was in Jansenius. If the church erred by indirectly
condemning Augustine when it condemned the Augustinus of Jansenius, at least
the church knew it was directly condemning Jansenius. Such was the view of at
least some members of the Jansenist camp, who, in abandoning Amauld' s dis
tinction, not only raised questions about Amauld' s sincerity, but also precipi
tated deep philosophical questions. For in response, Amauld brought in a host
of issues from the philosophy of language and of mind.
The most important member of Port-Royal to throw in the towel on the fait
droit distinction was Pascal, who finally decided that it was of no use in justify
ing the signing the famous fonnulary whereby one accepted condemnation of
the Five Propositions. He made his case in the Ecrit sur la signature, published
for the first time by Brunschvig in 19 14. Even if Pascal did as much as anyone
to propagate the fait-droit distinction by introducing it right at the outset of his
first Provincial Letter, his own, unyielding position on the issue of grace is clear
at the outset of this Ecrit: 'the truth of the matter is that there is no difference
between condemning the doctrine of Jansenius on the Five Propositions and
condemning efficacious grace, St. Augustine, [and] St. Paul. ' 1 8 He is also clear
that even given the fait-droit distinction, the Pope and bishops have condemned
that common position. For 'they claim that it is a point of droit and of faith to
say that the Five Propositions are heretical in the sense of Jansenius; and Alex
ander VII has declared in his constitution that to be in the the faith, one must
say that the words "in the sense of Jansenius" express only the heretical sense of
the propositions, and that it is a fact that conveys a droit [unfait qui emporte un
droit] and comprises an important part of the profession of faith, as would be
the case in saying that the sense of Calvin on the Eucharist is heretical, which is
certainly a point of faith.' 1 9 For Pascal, the intention of the Pope and bishops is
clearly to promulgate as a matter of faith the condemnation of what was in
Jansenius, and not just propositions described as being in Jansenius, so that the
ambiguity introduced at this point by reservations in signing the fonnulary is
mischievous (mechant). Those who sign in this way take a path that is ' abomi
nable before God and contemptible before men. ' 20
Arnauld responded to Pascal with a work called The true meaning of the
words 'the sense of Jansenius ' in the Constitution of Pope A lexander Vil. This
work was to respond to the fundamental issue, for even I nnocent X had earlier
made clear that by condemning the Five Propositions he was condemning, or
intending to condemn, the doctrine of Jansenius. 2 1 Arnauld begins his response

in typically engaging fashion, accusing Pascal of 'pure sophisms' and claiming
that only inattention will prevent 'reasonable readers' from being persuaded of
his case to this effect. The case is strung out over eleven 'maxims,' some of
which are premises and others conclusions that become further premises. The
putative logical structure is inverted since only at the end are the two maxims
revealed that are said to serve as a foundation for all the rest, and the argument
is generally obscure. Fortunately, help in understanding it is to be had from the
Port-Royal Logic, which was published the year following the problematic
signings. 22
According to Arnauld, 'the sense of Jansenius, in the natural and proper
sense of those words [sic],' is too general and must be determined, because
everyone agrees that not everything meant by Jansenius is heretical, and only, at
most, the Five Propositions' (Maxims 1-3). Determination is one of the two
kinds of predication, or what the Logic calls addition, whereby complex terms
are produced. Addition is expressed by a subordinate clause introduced by a
relative pronoun. Determination is the restrictive sense and sometimes makes a
general word individual; for example, 'the man who is now Pope' specified
Alexander VII in 1622. This determination can be either expressed, as in this
example, or only understood, as when 'the king' is assumed to refer to the cur
rent king of France, viz. Louis XIV. An expression of this sort can pick out an
individual and yet be equivocal because of disagreement about which individual
it is - for example, 'the greatest geometer in Paris,' or 'the doctrine of an author
on a given subject.' (The other kind of addition, expressed by the unrestrictive
sense of the relative pronoun, is called explication, which 'only unfolds what is
contained in the comprehension of the idea of the first term, or at least what
belongs to it as one of its accidents provided that it belongs to it generally and in
its full extension'; for example, 'man, who is rational' or 'man, who is bi
ped. ' ) 23 The distinction between determination and explication is important, as
Jill Buroker has shown, for distinguishing in the Logic between judgments and
merely complex ideas. 24 Here its significance is that, for Arnauld, the expres
sion ' the sense of Jansenius' does not pick out an individual sense, at least not
the sense of Jansenius. Even if it did pick out exactly one sense, the expression
might still be equivocal and open to debate because there might yet be debate as
to what that sense is.
The same point can be made another way. A conclusion drawn by Arnauld's
response to Pascal is that the Pope's pronouncement supposedly condemning
Jansenius contains an 'incidental proposition,' which is that the condemned
doctrine is in Jansenius, and which is a question de fail. The subject of the
condemnation is the doctrine, and this is a question de droit. The Logic also
60 Thomas M. Lennon
discusses incidental propositions, and divides them according to the species of

addition. 25 In an obviously significant example, one of each kind is to be found
in the following proposition: the doctrine that places the sovereign good in
bodily pleasure, which was taught by Lucretius, is unworthy of a philosopher.
'Placing the sovereign good in bodily pleasure' detennines the doctrine, but
being taught by Lucretius does not. The test is substitution for the relative pro
noun. We cannot say, 'the doctrine that was taught by Lucretius places the
sovereign good in bodily pleasure,' for Lucretius taught other doctrines as well.
On the other hand, we can say, 'the doctrine which places the sovereign good
in bodily pleasure was taught by Lucretius,' for Lucretius taught that whole
doctrine. (The Logic also notes that to decide which sense is being employed,
we often need to take into account the speaker's sense or intention as judged
from circumstances, e.g., in using 'the king' to refer to Louis X IV. This require
ment, of course, introduces another element de fait.) The conclusion, again of
obvious significance, is that the assertion 'the doctrine of Lucretius concerning
the good is false' contains an incidental proposition, which determines what
that doctrine is. 26
Perhaps the simplest way to put the problem is in terms of a dilemma.
An incidental proposition is required in the statement of condemnation;
but either (more plausibly) the incidental proposition is determinative, in
which case the truth of the condemnation depends on a matter of fact, or (less
plausibly) the incidental proposition is explicative, in which case the condem
nation applies to whatever was said. Thus, Jansenius stands condemned either
only factually, i.e., fallibly, or by definition, i.e. , vacuously. Moreover, the
problem is one that, given the fait-droit distinction, emerges with respect to
the overthrow of Jansenius' s proposition on the basis of papal infallibility,
but more generally with respect to the overthrow of any proposition by appeal
to a cognitive norm, e.g. Lucretius's propositions on the basis of, say, clarity
and distinctness.
In the reply to Pascal, Arnauld says that the statement de fait implicitly
expressed in certain statements de droit is easily overlooked, as in the assertion
that the doctrine of Arius is heretical. But this oversight is attributable to
general agreement as a matter of fact on what that doctrine was (the kind of
agreement that Pascal thought there was on Jansenius's doctrine). Arnauld does
not quite say so explicitly, but it appears that no statement de droit can ever
be expressed alone and, as such, without any statement de fait. In principle,
the determinateness of a stated doctrine can always be questioned by asking,
what does it mean? And this question can always be repeated, however many
incidental propositions are supplied in answer to it. The effect of Amauld' s

distinction is, if not to silence authority, at least to make it inaudible as such.
The dilemma that Amauld precipitates is clear: droit is, although distinguish
able from fail, inseparable from fail; thus, either papal pronouncements are
infallible with respect to at least some matters de fait or not at all. This predica
ment, of course, is what led the Jesuits to argue in 166 1 that the Pope is infalli
ble on questions de fait. For Bayle, of course, this is not a predicament at all. He
rejects papal infallibility of any sort, and, moreover, famously accepts the wider
scepticism to which a general version of Amauld's distinction might lead. 27
Notes
I am grateful, once again, to R.A. Watson for many useful comments on an earlier
version of this essay.
Reponse aux questions d 'un provinciale, in Oeuvres diverses (The Hague, 1 737),
I, 446.
2 Question de fait might be unproblematically translated as 'question of fact'; but
the translation of question de droit depends on philosophical issues it is the
purpose here to explore. Both expressions will therefore be left untranslated.
3 ' Jansensism and the Crise Pyrrhonienne,' Journal of the History of Ideas 38
( 1 977), 297-306.
4 The papal bull itself might have suggested the distinction to Port-Royal, for it
comdemns the first four propositions as heretical, the last merely as false. For the
argument here, precisely what the Five Propositions were does not matter; suffice
it to say that they amount to the claim that the grace merited by Christ is necessary
and sufficient for salvation.
5 Assertiones Catholicae de Jncarnatione contra saeculorum omnium ab incarnato
Verba praecipuas haereses. Quoted by Amauld, La Nouvelle heresie des Jesuites,
. . . ( 1 662), OA 2 1 , 5 1 5 .
6 Les lmaginaires, ou Lettres sur / 'heresie imaginaire (s. 1.n.d.) 2nd letter (March
1 664) written under the pseudonym ' Le Sieur de Damvilliers. ' The letter was
important enough to be translated into English along with two others by Nicole,
Amauld's Nouvelle heresie, and other documents, by John Evelyn: Mysterion . . .
that is, another part of the mystery of Jesuitism, . . . (London, 1 664), 1 08.
7 Theophilus Gale, The True Idea ofJansenism, Both Historic and Dogmatic
(London, 1 669), 1 6 1 .
8 The History of Scepticism from Erasmus to Descartes (New York: Harper
Torchbooks 1 968; 1 st ed. 1 964), 1 3.
9 August 1 685, Art. iii; OA 1 , 346--9.
62 Thom as M. Lennon
I O OA 1, 348. Here, rarely, Bayle confuses the truth of a view with its usefulness.
1 1 Steven Nadler, Arnauld and the Cartesian Philosophy of Ideas (Princeton, NJ:
Princeton University Press 1989), 176---8.
12 Avis a I 'auteur des Nouvelles de la Republique des Lettres (Delf [sic], 1685); in
OA 40, 1-9.
13 Ibid., 3.
14 Ibid., 4.
15 Ibid., 5. Recall that Jansenism was thought to be condemned when the famous
Five Propositions were condemned. (See note 4, above.) That there should have
been exactly five propositions in Malebranche cannot have been anything but
sarcasm on Amauld's part. In any case, these five propositions are as follows:
( l ) Those who enjoy pleasures of the senses are happy in so far as, and to the
extent that, they enjoy them.
(2) They none the less do not make us permanently happy.
(3) Although they make us happy, they must be avoided for several reasons.
(4) They must not lead us to love bodies, because bodies are not their real. but
only occasional cause, God being their real cause.
(5) Pleasure is imprinted in the soul in order that it love the cause making it happy,
i.e., God. Reflexions sur le systeme de la nature et de la grace (Cologne 1685), in
OA 39, 362. Amauld says that he finds great difficulties in all these propositions;
the second, fourth, and fifth are the questions that are independent of the question
de fait.
16 Quite apart from Amauld's general belligerence and readiness to attack
Malebranche, which is a moral or at least psychological issue, there is surely a
logical, or epistemological issue here. If the tacit assumption is that a proposition
is false in so far as it is held by Malebranche, if a Malebranchean proposition is in
this sense false by definition, argument against (or for) it is vacuous. See below.
17 Seconde lettre, in OA 19, 455.
18 Pascal, Oeuvres, L. Brunschvig, P. Boutroux, and F. Gazier, eds. (Paris: Hachette
19 14), 17 1.
19 Ibid., 172.
20 Ibid., 173, 175.
21 For an entree to this issue, see Louis Cognet, Le Jansenisme (Paris: Presses
Universitaires de France 1968), 65.
22 Privilege, l April 1662; publication, 6 July 1662. Attention is drawn to the
connection by the editors of Pascal, Oeuvres; see vol. 10, 22 1-28.
23 Logic, Part L ch. 7.
24 Jill Vance Buroker, ' Judgment and Predication in the Port-Royal Logic, ' in The
Great Arnauld and Some of His Philosophical Correspondents, E.J. Kremer, ed.
(Toronto: University of Toronto Press 1994).
25 Logic. Part II, ch. 4. Thus calling them incidental is perhaps misleading because
the kind is anything but incidental.
26 Roughly, the unrestrictive incidental proposition is a strict proposition involving
j udgment; it asserts the containment of the subject's extension by the predicate's
extension. The restrictive incidental proposition expresses a complex idea and
involves only conception; it asserts that the predicate's comprehension, or
intention. is compatible with that of the subject. I am grateful to Jill Vance
B uroker for discussion on this topic.
27 This is not to say, however, that Bayle has no way of going beyond this stultifying
scepticism .
5
The Status of the Eternal Truths
in the Philosophy of
Antoine Amauld
ALOY S E- R AY MOND ND IA Y E
The philosophy o f Antoine Amauld has been described a s 'Cartesianism

without the creation ofthe eternal truths.' 1 In Robinet's words, 'Arnauld presents
a tranquil Cartesianism that is not troubled about the eternal truths. ' 2 In the
abundant, essentially polemical work of the Augustinian doctor, written in a
troubled period and unfavourable circumstances, we find no text sufficiently
explicit, clear, and precise to determine his position on the Cartesian doctrine of
the creation of the eternal truths. But it is hard to believe that he was indifferent
and without reaction to this important metaphysical question, while his fellow
theologians declared themselves for or against the Cartesian position. The fact
that the philosophers of Port-Royal, even the Cartesians among them, were not
of one mind on all questions, does not explain Arnauld' s silence on this point.
The rule of the Port-Royalists was to take an interest in philosophy only to the
extent that it has implications for theology. Arnauld himself took part in a long
polemic against Malebranche only because he was persuaded that the Oratorian' s
theological errors regarding grace were owing to his philosophy of ideas and his
position that we see the ideas in God. The theological implications of the Cartesian
doctrine of the creation of the eternal truths were too obvious not to draw a
reaction from Arnauld.
We will approach Arnauld' s position by considering whether he was aware of

the objections that his contemporaries, especially Malebranche and Leibniz,
raised against Descartes. But first we should review the Cartesian doctrine. In
The Status of the Eternal Truths 65
reply to a question posed by Mersenne, Descartes claims that the truths of

mathematics are created: 'The truths of mathematics that you call eternal, like
all the rest of God's creatures, have been established by Him and depend
entirely on Him. '3 Thus the eternal truths are assigned the status of creatures.
They are not independent. They are created. Later, in his correspondence with
Mersenne, Descartes returns to the question. In a letter of 6 May 1630, he
replies to a theological difficulty raised by Mersenne: ' What you say about
the production of the Word is not, it seems to me, contrary to what I say; but
I do not want to meddle in theology. I am indeed afraid that you will judge my
philosophy too impoverished to include an opinion on such lofty matters. ' 4 To
appreciate the difficulty, one must know that the creation of the eternal truths
contradicts the doctrine of the 'production' of the Word. For the theologians,
the eternal truths have their identity in the Word. Given this theological posi
tion, anyone who said that the eternal truths had the same status as creatures
would risk reducing the Word to the status of a creature. Descartes's doctrine
suggests that God could produce the Word in the same way as any other crea
tures. To avoid contradicting the traditional theological teaching in this way,
Descartes would need to distinguish, to separate the created eternal truths from
the eternal, uncreated Word. Otherwise, to assume that the eternal truths are
created would be to assume that the Word who contains them is himself created,
and that is unacceptable. The Word is consubstantial and co-eternal with God.
Hence, to say that God creates the eternal truths would be to say that God gives
being to his own essence - hence, that God is possible - and that is absurd. To
the question of his illustrious correspondent, in quo genere causae Deus disposuit
aeternal veritates, Descartes replies in his letter of 27 May 1630: 'As the same
sort of cause whereby he created all things, that is to say as an efficient and total
cause. ' 5 We will return to this reply later.
Was Arnauld aware of the content of these letters when Mersenne sent him
the manuscript of the Meditations, not yet published, and engaged him to give
his opinion of the work in writing? We know that he took pleasure in reading
the Meditations, and yet produced the critical comments which constitute the
Fourth Objections, to be published at the same time as the Meditations them
selves. In the letter that accompanied the Objections, Arnauld gave this testi
mony to Mersenne about his relations with Descartes: 'You know in what es
teem I have long held his person and the importance I attach to his mind and his
views. ' It is obvious that the relations between the two famous philosophers did
not begin with the Meditations and the Fourth Objections. In 1637, Descartes's
first essays had been published and Arnauld had read them. But what captures
Arnauld' s attention in 1 640 is the Augustinian tone of the Meditations: 'The
66 Aloyse-Raymond Ndiaye
first thing I find worth noting is that Descartes establishes as the foundation and
first principle of his entire philosophy the same thing that was taken as the basis
and support of philosophy by St. Augustine, a person of great intellect and note
worthy doctrine not only in theology but also in human philosophy. '6 Arnauld
immediately saw the likeness of Descartes's cogito to certain texts of St Augus
tine. In the Fourth Objections, he cites the text of On Free Will. But later he will
cite the texts of the De Trinitate more often and at greater length. Indeed, he is
persuaded that the two authors have something in common. Descartes's purpose
agrees with that of St Augustine. Both want to prove the existence of God and
the immortality of the soul. Indeed, it seems that Arnauld was attracted to
Descartes's philosophy at least in part because of the Augustinian spirituality
that was very clearly affirmed in the Meditations.
But to return to our problem, in 1630 Mersenne drew Descartes out on the
question of whether the eternal truths are created or independent, and in 1640
Arnauld proclaimed publicly in his Objections that Descartes's philosophy
deserved the attention of a theologian. Was he at that time ignorant of the
discussion between Mersenne and Descartes and about what was at stake in
the debate between them? It should be mentioned that Mersenne' s letters were
published by Clerselier in 165 7 and 1659. Some years later, Leibniz, in his
correspondence with Arnauld, tried in vain to elicit his opinion on the question.
As for Malebranche, he did not hesitate to give a public and forceful critique of
the Cartesian doctrine on the origin of the eternal truths, a doctrine that he
thought ruined science, morality, and religion. But the two adversaries of the
Cartesian doctrine, Leibniz and Malebranche, did not get Arnauld to take a
public position.
II
It was especially Leibniz, more than Malebranche, who tried to draw Arnauld
out. Let us consider the matter more closely. Leibniz distinguishes two sorts of
necessary truths: First there are those that can be called eternal, such as the
truths of geometry, arithmetic, and logic. Their necessity is absolute and their
negations imply a contradiction and hence are logically impossible. On the other
hand, there are contingent truths, whose negations do not imply a contradic
tion, and hence are possible. These are the truths of fact, the truths of experi
ence, the laws of nature, and the truths of history. Their necessity, in contrast
with the necessity of the eternal truths, is hypothetical, because they depend on
the free decrees of God. They are founded on God's will, whereas the eternal
truths are based on God's understanding. Two important Leibnizian principles
are connected with this distinction, the principle of reason and the principle of
contradiction. The principle of contradiction determines only the possibilities.
It does not provide a sufficient reason for the existence of anything. The reason
for existence is not provided by logical necessity. It goes back to what regulates
the divine choice. The principle of contradiction is the law of essences or of
possibilities, that is, of the eternal truths contained in the divine understanding:
'The understanding of God is the region of the eternal truths. ' 7
Malebranche adopts the same classification of necessary truths. In the
Rechere he de la verite. he distinguishes two sorts of truths, necessary truths and
contingent truths: 'There are two sorts of truths; one sort is necessary and the
other contingent. I call necessary those truths that are immutable by their
nature, or that have been fixed by the will of God, which is not subject to change.
All other truths are contingent. '8 Thus, for Malebranche, there are two sorts of
necessary truths, just as for Leibniz, truths necessary by their nature, which are
called eternal, and truths necessary because they have been established by God.
The first are uncreated; the second are created and depend on God's immutable
will. The eternal truths are in the Word, and it is there that we see them. They
include truths of mathematics and truths of morality, the former constituted by
relations of quantity, the latter by relations of perfection. But when we consider
the status of the necessary truths that depend on divine decrees, we see that they
have a certain relation to reason. For Malebranche the divine decrees are neces
sary because they are the work of the Eternal Reason. God's volitions and ac
tions are guided by Wisdom. God's choice of general laws is a function of the
end God seeks in his creation. When he chooses the general laws, he knows that
he will never want to revoke them. Consequently, from God's point of view, the
general laws are just as eternal and necessary as the truths of mathematics. The
laws are not the result of chance, but rather rest on the very Wisdom of God,
which ensures their intelligibility, their immutability, and their a priority. So
the relations that the physicist discovers among things are the same as exist
among their ideas and are contemplated by God before the creation of the world.
His knowledge of them is a priori, that is to say, mathematical. From the point
of view of God, the laws of physics are like the truths of mathematics. In both
cases, necessity and immutability are guaranteed by the Wisdom on which the
truths are founded.
Leibniz' s position is the same. Contingent truths depend on the principle
of sufficient reason. They express existence and reality. Although contingent,
they do not lack a reason. Thus there is a reason for the choice by which God
preferred this world and its laws to an infinity of other possible worlds. But
that there is a reason means that the predicate is contained in the subject. To
give a reason for the existence of our universe and its laws is to show that the
contingent truths can be known a priori. Does this not reduce the principle of
sufficient reason to the principle of identity? Since every reason can be reduced
to identity, that is, to the inherence of the predicate in the subject, we can no
longer oppose the principle of sufficient reason to the principle of contradiction,

of identity, the foundation of contingent existence to that of essences. So, just as
for Malebranche, the truths of physics are like the truths of mathematics. But it
remains the case that contingent truths are known by us only a posteriori, that
is, by experience. Under what conditions could they be known a priori? They
could be known in that way by one who knew the reasons for which God pre
ferred our universe to all other possible universes. Our universe is that which is
'the most prefect, that is, which is at once simplest in hypotheses and richest in
phenomena. ' 9 But the determination of this maximum, defined by a divine math
ematics, constitutes the notion or essence of the most perfect universe, the one
actualized by the will of God, and everything that happens in it is contained in
its notion as a predicate in a subject. Anyone to whom this notion is distinctly
present can have a priori knowledge of all these events; hence, for such a person
the corresponding propositions would be demonstrable, reducible to identities.
But Leibniz recognizes that this notion can be distinctly grasped only by God,
for its determination requires an infinite calculus not available to human under
stand ing.
It is clear that the approaches of Malebranche and Leibniz converge: the
truths created by God are based on reason and in this way they are assured a sort
of mathematical necessity. But this transformation of the truths of physics
into truths of mathematics, knowable a priori, presupposes that God' s under
standing is distinct from his will, and that his will is subordinate to his under
standing. Both authors establish a hierarchy among the d iving attributes,
although Malebranche introduces an almost real distinction among the divine
perfections, while Leibniz holds only for a distinction of reason. On all these
points Malebranche and Leibniz are opposed to Descartes. Their critique of the
Cartesian doctrine of the eternal truths was a logical consequence of the princi
ples of their own philosophies. But Leibniz, unlike Malebranche, wanted to
have Arnauld on his side. He does not despair of converting Arnauld to his own
view. On the occasion of their correspondence a propos Article 13 of the sum
mary of the Discours de metaphysique, which deals with the consequences of
Leibniz's conception of substance for freedom, Leibniz tries to draw Arnauld
out: 'I will say a word about the reason for the difference between the notions of
species and those of individual substances, in relation to the divine will rather
than in relation to simple understanding. The notions of species do not depend
on divine decrees (despite what the Cartesians say, which does not seem to be of
concern even to you on this point). ' 10 Leibniz reaffirms his own conception of
the distinction between kinds of necessary truth, but he also points out that his
correspondent, in the preceding letter, was not concerned about the Cartesian
view that makes all necessary truths depend on a single act of God's will. In
a first draft of his letter, Leibniz is more explicit: 'I notice that Arnauld has
forgotten, or at least is not concerned about, the opinion of the Cartesians, who
hold that God, by his will, establishes the eternal truths, like those regarding the
properties of a sphere; but since I do not share their opinion any more than
Arnauld ... ' 1 1 If Leibniz is claiming that his illustrious correspondent is on his
side against the Cartesians, Arnauld, for his part, refuses to confirm that he is at
one with Leibniz against the Cartesians. Why this reserve? In his reply of 28
September 1686, Arnauld avoids the question. He begins by expressing his sat
isfaction with Leibniz's explanations: 'I am satisfied with the way in which you
explain the point that shocked me about the notion of an individual nature.'
Next he approaches the problem under discussion and asserts that the difficulty
remams:
There remains for me only the difficulty about the possibility of things and about
that way of conceiving God as having chosen the universe he created rather than
any of an infinity of other possible universes that he saw at the same time and that
he did not will to create. But since that does not bear properly on the notion of an
individual nature and since I would have to wander too far in order to explain what
I think about it, or rather to explain what I must reject in the thoughts of others as
being unworthy of God, it is just as well, Sir, that I say nothing about it. 1 2
The disagreement with Leibniz remains. Arnauld does not share his conception
of the possible. He refuses to say why, for fear of having to 'wander' too far and
to expose publicly what he 'must reject in the thoughts of others.' He says that
the question does not leave him indifferent. He has an opinion on the matter. As
we have seen, Leibniz does not succeed in separating Arnauld from Descartes.
But neither does he obtain from Arnauld a statement of agreement with the
Cartesian position.
The results are no different in the case of Malebranche. Even in the heat of
the polemic, Arnauld does not abandon his attitude of reserve. The violence of
Malebranche's criticisms of the Cartesian doctrine do not deflect Arnauld from
his objective, which is to show that the Oratorian is mistaken when he claims
that his philosophy is in agreement with the Augustinian doctrine. From
Malebranche's point of view, the free creation of the eternal truths compromises
science, religion, and morals. All truths are infected by the contingency and the
arbitrariness of God's free choice. The doctrine has the deeper implication that
the truth of our science does not participate in the truth, because our reason is
not identical with the Reason of God. 'Thus everything is overturned,' writes
Malebranche,
no longer any science, any morality, any undeniable proof of religion. This conse
quence is clear to anyone who follows out the false principle that God produces all
order and truth by an extremely free will. And this is what made Descartes
conclude that God could have brought it about that 2 times 4 not make 8 and that
the three angles of a triangle not be equal to two right angles, namely his claim that
there is no order, no law, no reason of goodness or truth that does not depend on
God, and that from all eternity God, as the sovereign legislator, ordained and estab
lished the eternal truths. 1 3
According to Malebranche, the notion o f truth loses all sense. For what our
understanding perceives as true depends essentially on its own structure, consti
tuted by its innate ideas, which could have been different if God had so willed.
In other words, our science could have been different if God had created us with
different ideas or essences. Thus our science might not be true science, because
it might not coincide with God's science. Hence, Malebranche's violent reac
tion against the doctrine of the creation of eternal truths, which he does not
separate from the doctrine of innate ideas. Creationism and innatism are inti
mately connected. They lead to scepticism. Malebranche turns for inspiration to
Augustine, and finds the doctrine of the Vision in God. Ideas and truths are no
longer effects of God's will. They have their source and their foundation in
God's Wisdom, the eternal Word.
In his controversy with Malebranche, Arnauld avoids a direct confrontation
with Descartes. He does not take a position on the thesis that Malebranche at
tacks. Rather he goes directly after the Oratorian' s thesis. His aim is not to
support the thesis of Descartes, about which he says not a word, but rather to
show that Malebranche is opposed to Augustine. His tries to show his adversary
that his concept of the Vision in God is not at the same as Augustine's. He
avoids the confrontation between Malebranche and Descartes, into which the
Oratorian tries to draw him, and substitutes a different debate, one in which he
opposes Malebranche to St Augustine. According to St Augustine, the truths
that we see in God are 'certain truths of morals the knowledge of which God
impressed in the first man, and which original sin did not entirely efface from
the souls of his children. These are the truths that St. Augustine often says we
see in God.' 14 The truths we see in God are moral truths. Hence, St Augustine
excludes mathematical truths. He distinguishes between the eternal truths, those
which are properly divine and which reside in the essence of God, that is to say,
the truths of morality, and the truths of mathematics. His disagreement with
Malebranche is fundamental. Yet Arnauld will regret the fact that Augustine
did not explain his position on the way in which we see truths in God.
Thus we see that neither Leibniz nor Malebranche succeeded in drawing
Amauld into a public debate about the Cartesian doctrine of the creation of the
eternal truths. This silence on the part of one who had elsewhere forcefully
defended Cartesian orthodoxy may cause surprise, in view of the fact that the
partisans and adversaries of Descartes took clear positions on the value of the
doctrine. 1 5 Nor was Arnauld unaware of these positions. He was perfectly well
informed about the objections made to Descartes during his lifetime, and about
the reservations of some of his own friends from Port-Royal about the free crea
tion of the eternal truths. How, then, to explain the fact that the Augustinian
doctor did not seem concerned about this debate? We need to go back to 1640, to
the Meditations, and especially to the Fourth Objections.
III
In his Objections, Arnauld refers to the remarks directed against Descartes by

Caterus, author of the First Objections, about Descartes's application of the
principle of causality to God. In the Third Meditation, Descartes demonstrated
the existence of God from the idea of God, considered with respect to its objec
tive reality. This proof introduces a relation of causality between the idea of God
in me and God existing outside of my thought as the origin of the idea. Follow
ing out this line of argument, Descartes attempts a second approach to God as
the cause of me who has that idea. We must, he says, consider 'whether I myself,
who has that idea of God, could exist if there were no God. And I ask, whence
would I have my existence? Perhaps from myself, or from my parents or from
some other causes ... But if I were independent of everything else and were
myself the author of my being ... I would not lack any perfection.' The question
barely formulated, Descartes responds by recognizing that I am obviously not
the cause of myself. Hence, I depend on another cause about which we must
inquire whether it 'derives its origin and its existence from itself, or from some
other thing. For if it derives it from itself, it follows that it must be God.' 1 6
In his Objections, Arnauld notes that the author of the First Objections had
made the subtle remark that 'to be from itself' should not be taken 'positively'
but 'negatively,' so that it means not to be from another. As a good Thomist,
Caterus does not think that one can apply causality to God. According to St
Thomas, God is self-subsistent being, ens per se subsistens, being without a
cause. Descartes, in his reply to Arnauld, which I here summarize, maintains
that to be from itself should not be taken 'negatively' but 'positively,' even with
regard to the existence of God, in such a way that God does with respect to
himself what an efficient cause does with respect to its effect. This notion of
Descartes's seems to Arnauld to be 'a bit foolhardy and not to be correct.' 1 7
What he fears is that by applying the same principle of causality to God as to
other things, we would end up lowering God to the level of a creature. To say
that God is from himself positively is to admit that he gives himself a being he
did not fonnerly have, in the same way as he gives being to other things. He
concludes by noting that ' we can conceive that God exists from himself posi
tively only because of the imperfection of our mind, which conceives God in the
manner of created things.'
In support of his argument Amauld uses an example from mathematics: 'We
seek the efficient cause of a thing only on account of its existence and not its
essence. For example when we look for the efficient cause of a triangle, we want
to know what has brought it about that there is a triangle in the world, but it
would be absurd to seek the efficient cause of the fact that a triangle has three
angles equal to two right ones; and to one who seeks that, we would not respond
well by giving an efficient cause. Rather we ought to reply only that such is the
nature of a triangle.' Arnauld's position here is quite clear. The only cause
required for the eternal truths, the truths of mathematics, is the efficient cause.
But it is required in order to give a reason for their existence in the world. In
other words, God is the only efficient cause of the presence in our mind of the
truths of mathematics. He has placed them in us. They constitute the natural
light. While he is the cause of their existence in our mind, he is not the efficient
cause of what they are in themselves. 'To one who asks why the three angles of
a triangle are equal to two right angles, we ought not reply by giving an efficient
cause, but by saying, because such is the immutable and eternal nature of a
triangle.'
This assertion is clearly contrary to what Descartes would have said. For
Descartes, God is the efficient and total cause of the truths of mathematics,
since he has created the eternal truths and has decided that the sum of the angles
of a triangle equals two right angles. Arnauld, on the contrary, in the above text
from the Fourth Objections, does not make the nature of the immutable and
eternal truths of mathematics depend on the arbitrary will of God. He does not
apply efficient causality to them considered with respect to their nature or es
sence. Arnauld's statement stands in contrast to Descartes's statement to
Mersenne that God 'created all things as efficient and total cause, for it is cer
tain that he is the author of the essence as well as the existence of creatures; but
that essence is nothing other than the eternal truths.' 1 8 Clearly, Arnauld and
Descartes do not agree on this point. Amauld does not make the nature of the
eternal truths depend on the efficient causality of God. He says the same thing
about God as about the mathematical truths. 'If anyone ask why God exists, or
why he does not cease to be, we must not seek, in God or outside God, an
efficient cause or a quasi-efficient one (for I do not dispute here about the word,
but about the thing). Rather we must say, with all reason, that God exists be
cause such is the nature of the supremely perfect being.' 19 Thus the eternal and
immutable truths of mathematics function like the divine essence. They are to
be treated like God. With respect to their nature, they do not require an efficient
cause. And with respect to their existence? They do not require an efficient
cause in this respect either, except for their presence in our mind. And even in
this case, we cannot say that they are created. We must rather say that they have
been placed in us by God. This conclusion, inspired by Augustine, is not ex
pounded by Arnauld in the Fourth Objections. There he is content to show that
they do not require an efficient cause, hence that they are not created. They are
not lowered to the level of creatures.
Nothing in the text of the Fourth Objections tells us that Arnauld knew about
the discussion between Mersenne and Descartes regarding the creation of the
eternal truths. But even if we cannot take his remarks about the application of
efficient causality to God as a direct intervention in the debate, we can at least
recognize the striking similarity of his arguments to those used by Mersenne
against the production of the Word. Arnauld's fear that God will be lowered to
the level of creatures if we apply the same principle of causality to him and to
other things, is like Mersenne's fear with regard to the creation of eternal truths,
assuming that these truths, as traditionally taught in theology, are one with the
substance of the Word. Neither is it a mere coincidence that the example Arnauld
chooses in order to argue against the Cartesian conception of causa sui is drawn
from mathematics. The truths of mathematics do not function like created things.
These truths do not exist; that is to say, they are not created outside of God. They
are present in us, innately. But for Arnauld their being innate is not the same as
their being created. He takes his inspiration from Augustinian illuminationism.
Even if Arnauld has not directly confronted Descartes's doctrine on the creation
of the eternal truths, the fact remains that he has a settled position on the ques
tion. He does not share Descartes's doctrine. Neither does he find it dangerous,
once Descartes has given his guarantees to Mersenne, and a satisfactory re
sponse to Arnauld himself on the occasion of the Fourth Objections.
It may be asked how far Arnauld accepts the doctrine of his correspondent
when he says that he is satisfied with Descartes's explanation of the void, given
during their correspondence in 1648. Must we say that Arnauld also accepts the
creation of the eternal truths, the keystone of the Cartesian position? The an
swer is not certain.
In conclusion, we have seen that Arnauld often gives the impression that he is
ready to accept the confident explanations provided by his correspondents, as if
he did not have his own personal philosophical position to defend. We have
seen this in the case of both Descartes and Leibniz. But if we look more closely,
it is Arnauld who sets the question and fixes the rules for these discussions, not
his correspondents. They are put in the position of owing him explanations and
clarifications. In his controversy with Malebranche, he concedes nothing, and

he is the one who chooses the terrain for the discussion. On the question about
the eternal truths, Arnauld does not follow Malebranche. Neither does he follow
Leibniz or Descartes. Ifhe does not side with Descartes, or, later on, with Leibniz,
he does not for all that renounce his own convictions. He sticks to his own
positions, while remaining open to discussion. If the failure of understanding
between Arnauld and Malebranche was total, and their controversy aggressive,
something that Arnauld found painful, we must recognize that it was the
Oratorian who first lost his sense of moderation. Arnauld needs, indeed, to be
reassured. He wants to be sure that the theses of his correspondents will not
have consequences 'dangerous for weak minds.' 20 I n his theological undertak
ings and his works on spirituality, Arnauld shows that was concerned about the
salvation of souls. His preoccupation was pastoral. He thinks of the Christians
whose conscience has been shaken by the crisis that is crossing the church and
by the arguments that have been directed from all quarters against the faith. If
he has devoted so much time to philosophy, and in particular to Cartesian phi
losophy, despite the resistance of his friends at Port-Royal, it is because he is
persuaded that philosophy is useful to religion. What he always liked in Descartes
was his submission to the church and his constant care not to meddle in theol
ogy, that is, in the theology that arises from Revelation. Because he recognized
his incompetence in the area of theology, Descartes was assured of an ally of
great authority in Antoine Arnauld.
Notes
1 Henri Gouhier, Cartesianisme et augustinisme au XV/l e siec/e (Paris: Vrin 1 978),

1 56.
2 Andre Robinet, in the preface to A.-R. Ndiaye, la Philosophie d 'Antoine Arnauld
(Paris: Vrin 1 99 1 ), 8.
3 Descartes to Mersenne, 1 5 April 1 630, in Alquie, 1 , 259. Note that Descartes does
not say that the eternal truths are created. Mersenne fears that such a position
might follow from the formulae used by Descartes. Hence the difficulty he raises
regarding the production of the Word. Regarding Principles I, no. 22, which
Madame Rodis-Lewis takes to show that Descartes sets forth ' explicitly' the
doctrine of the creation of eternal truths, Arnauld retains nothing of this passage
except its Augustinian tone (see Ndiaye, la Philosophie d 'Antoine Arnauld, 342).
4 Descartes to Mersenne, 6 May 1 630, Alquie, 1 , 265 .
5 Descartes to Mersenne, 27 May 1 630, Alquie, I , 267.
6 Arnauld, Fourth Objections, Alquie, 2, 633.
7 Leibniz, Monadology, Article 43� to Amauld, 4/14 July 1686.

8 Malebranche, Recherche, L I no. 2, OC 1, 63.
9 Leibniz, Discours de metaphysique et correspondance avec Arnauld, edited by
Georges Le Roy (Paris: Vrin 1970), 1 15. Hereinafter 'Le Roy.'
10 Leibniz to Amauld, 4/14 July 1686, Le Roy, 105 .
1 1 Leibniz, Remark about Amauld's letter, Le Roy, 133.
12 Amauld to Leibniz, 28 September 1686, Le Roy, 133.
13 Malebranche, Recherche, eel. 8, OC 3, 84.
14 Amauld, On True and False Ideas, Ch. XIX, OA 38, 282.
15 Jean-Luc Marion presents all the texts on the creation of the eternal truths in
Sur la Theologie blanche de Descartes, (Paris: PUF 198 1), 270.
16 Meditations, I II, Alquie, 2, 449.
17 Amauld, Fourth Objections, Alquie, 2, 449.
18 Descartes to Mersenne, 27 May 1630, Alquie, 1, 267.
19 Arnauld, Fourth Objections, Alquie, 2, 65 1.
20 Ibid., Alquie, 2, 653.
6
Arnauld ' s Interpretation of Descartes
as a Christian Philosopher
E L MAR J . K R E M E R
It is well known that Arnauld was an admirer and defender of Descartes's

philosophy. Indeed, as Steven Nadler has said, 'Arnauld, of all the Port-Royal
ists of his generation, appears to be alone in his lifelong enthusiasm for
Descartes. ' 1 Nadler is here correcting a widely shared view that the Port-Royal
ists were, as a group, Cartesians. Indeed, Arnauld was criticized by others in the
Port-Royal group on account of his willingness to defend Descartes. His nephew
De Saci, together with whom Arnauld produced a translation of the Bible, was
especially vigorous in his criticism.
But although Arnauld admired Descartes' s philosophy and was prepared to
defend it against the charge that it was heretical, he did not accept some of its
most important principles, and cannot properly be said to be a disciple of
Descartes. He considered himself a professional theologian rather than a
philosopher, and his primary allegiance was to the Catholic theological tradi
tion as he understood it. Hence, he was prepared to depart from Descartes when
ever he thought it necessary to do so in order to maintain theological orthodoxy.
For example, from the Fourth Objections on, he rejected the method of doubt, at
least in part for theological reasons. 2Again, although he shared much of
Descartes's scorn for Aristotelian philosophy, his own late philosophy and the
ology are heavily influenced by St Thomas, and he did not hesitate to reject
important Cartesian doctrines, including the doctrine on free will and on judg
ment as an act of will, in favour of Thomistic ones. 3
My concern here is with Arnauld' s defence of Descartes as a 'Christian
philosopher. ' 4 When Amauld applies this phrase to Descartes, he means, not
only that Descartes was a philosopher and a Christian, but that his philosophy
Amauld's Interpretation of Descartes 77
was intended to be useful, and was indeed useful, to Christian theologians in

their task of defending the faith. Amauld thought that Descartes deserved to be
called a Christian philosopher in part because of the way he supported the doc
trine of human immortality by his argument for mind-body dualism. But he
especially commended Descartes as a Christian philosopher because Descartes
recognized it as a philosophical principle that 'whatever God has been pleased
to reveal to us about himself or about the extraordinary effects of his omnipo
tence ought to take first place in our belief even though we cannot conceive
[ concevoir] it, for it is not strange that our mind, being finite, cannot compre
hend [comprendre] what an infinite power is capable of.'5 Amauld was pleased
that what he considered staples of the Christian tradition were upheld by the
most brilliant thinker of his time. However, even while rejoicing in Descartes's
support for these doctrines, he found it necessary to modify the Cartesian
philosophy. I shall try to show, in particular, that Amauld's interpretation of
Descartes as a Christian philosopher led him to depart from Descartes's posi
tion on the nature of matter, on the relation of mind and body, and on divine
omnipotence.
A Modification of Descartes's Position on the Nature of Matter
Descartes helped defend the doctrine of immortality, says Amauld, by showing

that the soul and the body are two totally distinct substances, in such a way that
it is impossible that extension be a modification of the substance that thinks or
that thought be a modification of extended substance. 6 Repeating a line of thought
found in Descartes's Synopsis of the Meditations, Amauld says from the Cartesian
conclusion, together with a premise which any materialist will surely grant -
that nothing that exists goes out of existence in nihilo - implies that it is unrea
sonable to claim that the soul dies with the body.
But the Cartesian distinction between mind and body presupposes the
doctrine that the essence of matter is extension, and when the distinction is
viewed as part of Christian philosophy, the problem arises that the Cartesian
doctrine on the essence of matter contradicts authoritative church teaching
on the Eucharist. The first of Amauld's modifications of Descartes's philosophy
mentioned above is a response to this problem. It emerges in his attempt to solve
a difficulty he raised in his two letters to Descartes in 1 648. But the letters
of 1 648 are the second occasion on which Arnauld expressed misgivings
about the compatibility of the Cartesian view of matter with the doctrine of
transubstantiation, and, before considering the precise objection he raised
there, it will be helpful to summarize briefly the outcome of the first criticism,
78 Elmar J. Kremer
which was raised in the Fourth Objections of 1641.

In the concluding section of the Fourth Objections (AT 7, 214-18), ArnauId
briefly indicates four 'points which may cause difficulty to theologians.' The
fourth and last of these is, he says, 'likely to give the greatest offence ... ' It is
'that according to the author's doctrines it seems that the church's teaching
concerning the sacred mysteries of the Eucharist cannot remain completely in
tact.' He formulates the sacred doctrine as follows: 'We believe on faith that the
substance of the bread is taken away from the bread of the Eucharist and only
the accidents remain. These are extension, shape, colour, smell, taste and other
qualities perceived by the senses. ' He takes this to mean that such qualities,
understood as qualities present in the bread, remain after the bread is conse
crated, without inhering in any substance. The difficulty is that, according to
Descartes, all these qualities are reduced to 'shape, extension and mobility,' and
'yet the author denies that these powers are intelligible apart from some sub
stance for them to inhere in, and hence he holds that they cannot exist without
such a substance.' In reply, Descartes says the 'accidents' of the bread which
remain after the bread ceases to exist can be nothing other than 'the surface that
is common to the individual particles of the bread and the bodies which sur
round them' (AT 7, 251). Amauld, as we shall see, eventually accepted Descartes's
reply, and indeed considered it a theological advance over the explanations
previously available.
But this was not the end of Arnauld's worries about the compatibility of
Descartes' s philosophy with the doctrine of transubstantiation. He returns to the
problem in the first of the letters of 1648, and raises a new difficulty: " You assert
that a quantified thing [rem quantam] is not distinct in any way from local
extension [locali extensione]. Therefore I would like to know whether you have
thought of some way to reconcile that doctrine with the catholic faith, which
requires us to believe that the body of Christ is present on the altar without local
extension, just as you succeeded in showing how the absence of a distinction
between accident and substance could agree with the same mystery. Otherwise
you easily see to what great danger you expose the most sacred of all things.'7
This difficulty differs from that raised in the Fourth Objections in two ways. In
the earlier text, the focus is on the continued presence of the • accidents' of the
bread and wine when the bread and wine no longer exist, while in the present
text the focus is on the presence of the body of Christ. Second, the present text
for the first time refers to the Cartesian doctrine that the essence of matter is
local extension, which brings Descartes into conflict with the traditional teach
ing of the church that Christ is not present in the sacrament locally. Descartes
had recognized this church doctrine in his Replies to the Fourth Objections, 8
but this is the first time Amauld questioned its consistency with the Cartesian
theory of matter. The doctrine does not mean that the body of Christ is not
Amauld' s Interpretation of Descartes 79
present where, or even, in the place where, the bread had previously been. Rather,
it means that the body of Christ does not take on the dimensions of the place in
which it is present; it does not fill or occupy the place where it is.
It is not surprising that this doctrine seemed to Amauld incompatible with
Descartes' s position, for it implies that the body of Christ does not take on the
dimensions of the bread when it comes to exist where the bread was. But Descartes
holds that the essence of a body is an extension whereby it occupies a place.
Thus, in a letter of 1641, he says that a body, unlike a mind, has a 'true exten
sion, that is, an extension "whereby it occupies a place and excludes any other
thing from it" ' (AT 3 , 435). It is hard to avoid the consequence that, according
to Descartes, the body of Christ, as really present in the Eucharist, has the
dimensions previously possessed by the bread. So it seems that Descartes's
position is inconsistent with the teaching of the church.
Descartes declined to answer Amauld's question, citing the Council of Trent,
which ' did not wish to explain in what precise way the body of Christ is in the
eucharist, and wrote that it is there in a way of existing which we can scarcely
express in words' (AT 7, 251 ). 9 Now, in fact, Amauld himself thought that in
quiry into the precise way in which the body and blood of Jesus are present in
the sacrament was ill advised, and cited the Council of Trent in support of this
policy. 1 0 But he had not asked Descartes to explain precisely how the body of
Christ is in the Eucharist. He had asked a different question, namely, how the
church's doctrine that Christ's body is not present there locally is consistent
with the Cartesian doctrine that a quantified thing is not distinct from local
extension. He had reason, then, not to be satisfied with Descartes's response,
and, as we shall see, in his second letter he presses Descartes for an answer.
Immediately after raising the above difficulty about the Eucharist in the first
letter, Amauld raises an objection to Descartes's doctrine that there cannot be a
vacuum. It is clearly within God's power, he asserts, to cause a vacuum. For
example, God could 'annihilate the wine in a wine jar without producing an
other body in its place . . . ' ( OA 3 8, 73). He goes on to rebut the objection he
expects from Descartes, namely, that if there were a vacuum, there would be
properties such as length, breadth, depth, and divisibility in the vacuum, and
hence the vacuum would be something and indeed would be a body. 1 1 Against
this, Amauld says that, after the wine was annihilated, no properties would be
present in the vacuum. The distance between the sides of the jar, and other
spatial features of the cavity (concavitati) in the jar, would remain, but these
would be present in the jar, not in the vacuum.
Amauld does not explicitly link this new objection with his difficulty about
local extension and the Eucharist, but the two difficulties were almost certainly
connected in his mind. There is some reason to think that in 1648 Amauld held
some version of the Scotist theory of transubstantiation, according to which the
80 Elmar J. Kremer
substance of the bread and wine are annihilated at the time of consecration, for
this theory was influential in the seventeenth century. As Annogathe says, 'in
the seventeenth century the accepted scholastic doctrine was of the Scotist type
and thus includes the annihilation of the substance of the bread and wine.' 1 2 If
Arnauld did hold the Scotist theory, then his example, that God can annihilate
the wine in a container without producing another body in its place, would be
exactly what he thought occurred in the Eucharist. For according to church
doctrine, the wine and the bread are not replaced by a newly created body. The
bread and wine are replaced, in a sense, by the body and blood of Christ. But
according to the doctrine that Christ's body and blood are not present locally on
the altar, they do not fill or occupy the place where the bread and wine had been.
In his first reply, Descartes does little more than refer Arnauld to the Princi
ples of Philosophy regarding the impossibility of a vacuum. But Arnauld was
not satisfied. In his second letter, he returns to the point: ' About the vacuum, I
confess that I still cannot swallow the proposition that corporeal things are con
nected in such a way that God could not have created a world unless it was
infinite, or annihilate a body without by that very fact being bound to create
another of equal quantity, indeed without the space occupied by the annihilated
body being understood to be a real and true body in the absence of any new
creation' (OA 38, 83 ). He then adds that he would be very happy if Descartes
would provide an answer to the question he had raised in the first letter about
the Eucharist. But his repeated question was met by silence on Descartes's part.
Arnauld returned to the topic once again in 1 680 in a document quoted above,
in which he defends Descartes against the charge of heresy . ' 3 I shall interpret
this text in the light of Descartes's refusal to respond to Arnauld's request in
1 648. In 1 680, Arnauld puts much emphasis on the proposition, taken from the
Council of Trent, that the body and blood of Christ are present in the Eucharist
in a way that we can scarcely understand, and takes this to indicate that it is
unwise for theologians to attempt to explain the manner of this presence. But
this should not encourage us to think that Arnauld was content with the reply he
received from Descartes in 1 648, about the way in which Christ is present in the
sacrament. On the contrary, as I have pointed out, the fact that he raised the
question again in his second letter suggests that he was dissatisfied with
Descartes's response.
The charge of heresy against which Arnauld defended Descartes in 1 680 was
levelled by LeMoine, the dean of Vitre. LeMoine, as reported by Arnauld, took
the position that, according to the teaching of the church, the body of Christ is
present in the Eucharist in an indivisible point, without extension, and then
pointed out that this teaching is inconsistent with the Cartesian position that the
essence of body is extension. Arnauld begins his rebuttal by setting forth the
Cartesian position on the nature of matter, not as given by Descartes, but rather
by Malebranche, in Book III, Chapter 8, Section 2, of The Search after Truth. 1 4

Malebranche defines the essence of a thing as that property of the thing which
we recognize to be primary in the thing, which is inseparable from it, and on
which depend all its other properties. So, in order to find the essence of matter,
we should consider all the properties found in matter and ask which of them
satisfies all three criteria. He begins with inseparability from matter. Some
properties of matter, such as hardness, softness, and motion, he says are known
to be separable from matter because matter is sometimes found without them.
There remain four properties, he says, which 'we conceive as inseparable from
matter.' These are divisibility, shape, impenetrability, and extension. To dis
cover which of these inseparable properties constitutes the essence of matter, he
says, we should ask which are primary in matter and are such that all the other
properties of matter depend on them. He asserts that only extension satisfies
these further criteria.
After quoting Malebranche at length, Amauld returns to LeMoine's position
that, according to the teaching of the church, a body can exist without any ex
tension in a mathematical point. He says that LeMoine arrives at his position by
confusing extension with impenetrability. He agrees that, according to the teach
ing of the church, a body can exist without being impenetrable. But, in a sur
prising text. he says that this does not mean a body can exist without being
extended, for ' as we have just seen in the passage I have reported [ from
Malebranche ], only extension is the essence of matter, and impenetrability is
merely one of its properties. But only the essence is inseparable from it; regard
ing the properties, nothing prevents our saying that they can be separated from
it by the power of God, even if they naturally include the idea of shape and local
motion [/ors meme que naturellement elles renferment I 'idee de figure & de
mouvement local] . . . And consequently what the Fathers may have said about
penetrability of bodies by the power of God does not imply that they should have
said the same about their inextension, so to speak, that is to say, that they should
have said that God could reduce them to an indivisible point' (OA 38, I 05).
Malebranche had said that all four of the properties - divisibility, shape, impen
etrability, and extension - are inseparable from matter, but that only extension
constitutes its essence. Amauld accepts the latter point, that only extension is
the essence of matter, but then introduces a distinction not found in Malebranche
(or Descartes), between properties of matter that constitute its essence, and hence
cannot be separated from it even by the power of God, and properties that,
though they are inseparable from matter in the order of nature and can be said to
be part of the nature of matter, 1 5 can be separated from matter miraculously
by God's power. Consequently, on Amauld's view, even if the body of Christ is
present in the Eucharist without certain properties that are inseparable from
extension in the order of nature, this does not imply that the body of Christ is
82 Elmar J. Kremer
present on the altar in a mathematical point without extension. Furthermore,

Amauld seems to say in the above text, the body of Christ, as really present
on the altar, does have the property of extension, since extension cannot be
separated from body, even by the power of God.
The reason why extension cannot be separated from body seems to be that a
thing and its essence are identical. Arnauld puts the point explicitly in his late
Lettres au Pere Malebranche: 'Anyone who has accepted it as a principle that
extension is the essence of matter, ought to say that extension and matter are the
same thing.' 1 6 Thus the notion that the body of Christ is really present on the
altar in an indivisible point, without being extended, contains a contradiction,
and anything in whose concept there is a contradiction, according to Amauld,
is 'absolutely impossible.' 1 7 But there is no contradiction in saying that matter
is really present on the altar without impenetrability, or without any other
property distinct from its essence, and for this reason matter can be made to be
present without any such a property by the power of God.
Amauld mentions two miracles, cited by Augustine, in which impenetrabil
ity is removed from bodies: the birth of Jesus, which occurred 'without damage
to the virginity of his mother,' and Jesus' entry into a room where his disciples
were gathered despite the fact that the doors were closed. He also claims that the
body of Christ, as present in the Eucharist, lacks the property of impenetrability.
Thus he says that, when 'explaining the mystery philosophically, ' we cannot
avoid the consequences 'that God can bring it about that the same body exists in
different places and that the parts of a body penetrate one another [/es parties du
corps se penetrent] ' (OA 38, 109). So, once again, the doctrine of the Eucharist
seems to conflict with Descartes, who says, ' impenetrability belongs to the
essence of extension ...' 1 8 The apparent conflict is removed by the distinction
between extension itself, which, as the essence of matter. is absolutely insepara
ble from it, and the properties of matter, like impenetrability, which can be
miraculously separated from it.
As the passage just quoted makes clear, impenetrability is not the only
natural property of bodies that Amauld thinks is separated from the body of
Christ as it is present in the Eucharist. In other passages, he mentions the prop
erty of being enclosed in a place (renferme dans un lieu) and the property of
having a closed surface (une surface bornee). 1 9 So it looks as if Amauld is
concerned not only with LeMoine's attack, but with the difficulty he himself
had raised some thirty-two years earlier, about the compatibility of the Cartesian
theory with the doctrine that the body of Christ is present in the Eucharist
without local extension. For if the body of Christ is present on the altar with
out having a closed surface, then Christ's body is not present there locally.
The Cartesian theory of matter is brought into consistency with this revealed
truth if the presence of matter without its properties, as opposed to its presence
Arnauld's Interpretation of Descartes 83
without its essence, does not involve a contradiction.

The notion of a material thing present in a particular place without having
any particular dimensions there is not easy to grapple with. Even within
Descartes' s theory, there is one example of a material thing which does not have
any particular dimensions, namely, the material world, which, according to
Descartes, is indefinitely large. 20 But this could not be true of any part of the
material world in Descartes's account. Arnauld's position is closer to that of
Aquinas, who says, 'The place in which the body of Christ exists [in the Eucha
rist] is not empty [vacuus]. But neither is it filled by the substance of the body of
Christ, which is not there locally ...'2 1
A Modification of Descartes's Position on the Relation of Mind and Body
It is easy to see how this first modification of Descartes's philosophy arose when
Arnauld took seriously Descartes's claim to be a Christian philosopher. After
all, a philosophical position will not be of use to a theologian in defending the
truths of the faith if it is inconsistent with one or more of those truths. The
second modification, to which I now turn, is not so obviously related to the idea
of a Christian philosophy. In this case, Arnauld modifies Descartes so as to
bring him into line with the philosophical views of his two great predecessors in
the Christian tradition, Augustine and Aquinas. The problem arises as follows:
Despite his claim that all conscious activity, including sense perception, occurs
in the body, Descartes thought he could say that sense perception is caused by
material things. 22 But it was a deeply embedded principle in the philosophy of
both Augustine and Aquinas that a material thing cannot produce an immate
rial effect. Since the point at issue is largely, ifnot entirely, philosophical, Amauld
does not try to settle it by citing authorities. Rather, he produces the following
argument: 'Since the motion of a body cannot have any other real effect but to
move another body, it is clear that it cannot have any real effect on a spiritual
soul, which is by its nature incapable of being pushed or moved about' (OA 38,
146). 2 3
This departure from Descartes had two important consequences for Amauld.
First, like Malebranche, he had to deal with the question of how we could have
cognition of material things despite the fact that our cognitions are not caused
by material things. Second, when he tries to refute Malebranche's claim that we
cannot have demonstrative knowledge of the existence of material things,
he cannot rely on the sort of argument Descartes used in the Sixth Meditation,
which depends on the possibility that the active faculty corresponding to
our passive faculty of sense perceptions exists in material things. In the last
chapter of On True and False Ideas, Arnauld provides eight arguments for
the existence of an external world, but none of them relies on the notion that
84 Elmar J. Kremer
sense perception is passive. So it is clear that this was an important modifica

tion of Descartes's view. But does it tell us anything about Arnauld's idea of a
Christian philosophy?
I think it does. As I mentioned above, Arnauld settles the question not by
citing authorities, but by argument. None the less, he does cite authorities in
favour of his position, in particular, Augustine. 24 This may at first glance seem
to contradict the Augustinian dictum which Arnauld repeats on a number of
occasions, that questions of faith are to be decided by authority, but those of
philosophy by reason. However, he does not interpret this principle rigidly. Thus
he says: ' Be very careful about the nature of the question you are debating,
whether it is philosophical or theological. For if it is theological, it should be
decided principally by authority; whereas, if it is philosophical, it should be
decided principally by reason. I say, principally, because there is nothing to
prevent one from citing authorities also in philosophical questions, but this should
be done to throw light on them, not to decide them. '2 5 The philosophical
authorities he cites by far the most frequently, in addition to Descartes, are
Augustine and Aquinas. In fact, throughout Arnauld's discussion of Cartesian
philosophy, he shows himself eager to establish as much continuity as possible
between Descartes and the preceding Christian philosophical and theological
tradition. It would not be going too far to say that Arnauld questioned with
special care the points on which Descartes departed from that tradition. Once he
realized that Descartes had contradicted the view held by both Augustine and
Aquinas, that a material thing cannot produce an immaterial effect, he consid
ered the novel doctrine carefully, and found it wanting.
The first modification of Descartes's position showed that Arnauld could not
accept as part of a Christian philosophy a doctrine which was inconsistent with
authoritative church teaching. The second modification shows that, in his view,
a Christian philosophy is more useful to theologians to the extent that it remains
in continuity with the most important Christian philosophers of the past.
Arnauld's reading of Descartes as a Christian philosopher is a reading of
Descartes as someone in continuity with the medieval theological tradition.
Faith, Reason, and God's Omnipotence
I turn now to the second philosophical doctrine that Arnauld thought marked
Descartes as a Christian philosopher. I shall refer to it as the principle of divine
incomprehensibility (POI): (a) We ought to believe everything that is revealed
by God; and (b) the fact that we cannot understand how something revealed
by God can come about should not deter us from believing it because a finite
intelligence ought not expect to understand the nature or causal power of an
infinite being.
Amauld's formulation of POI (quoted in the third paragraph of this essay) is

quite close to that given by Descartes in Principles of Philosophy, I, nos 25 and
76. Where Amauld says that what God reveals ' ought to take first place in our
belief [ doit tenir le premier lieu de notre creance ], ' Descartes says that we should
make it 'the first rule [pro summa regula] ' that ' whatever God has revealed to
us ought to be accepted as the most certain of all things [ut omnium certissima
esse credenda] . ' Amauld's and Descartes' s 'ought' means that it would be irra
tional to believe that something is revealed by God and yet to reject it. How we
can tell that something is in fact revealed by God is, of course, another matter.
In this connection, Descartes says, 'Even with respect to the truths of the faith,
we should perceive some reason which convinces us that they have been
revealed by God, before deciding to believe them. '26 But once given that some
thing has been revealed by God, it would be irrational not to believe. Amauld
and Descartes also express (b) in quite similar terms: Where Amauld speaks of
what God reveals ' about himself or about the extraordinary effects of his om
nipotence, ' Descartes refers to revelation about 'his immense nature' and 'the
things created by him.' Again the point has to do with rationality in the forma
tion of beliefs: once given that something is revealed by God, it would be irra
tional to take the fact that we cannot see how it is possible as a reason for not
believing it.
Although Amauld was happy to have Descartes's support for POI, he does
not treat the doctrine itself as a distinctively Cartesian one. He points out that it
is a commonplace in the fathers of the church.27 What is distinctive in Descartes's
treatment is his claim that ' what an infinite power is capable of' includes mak
ing it possible that self-contradictory propositions be true and that what are in
fact necessarily true propositions be false. Descartes's strongest assertion of this
position is in a letter to Mesland:
The power of God cannot have any limits, and . . . our mind is finite and so created
as to be able to conceive as possible the things which God has wished to be in fact
possible, but not to conceive as possible things which God could have made possi
ble, but which he has nevertheless wished to make impossible. The first considera
tion shows u s that God cannot h ave been determ ined to make it true that
contradictories cannot be true together, and therefore that he could have done the
opposite. The second consideration assures us that even if this be true, we should
not try to comprehend it, since our nature is incapable of doing so. 28
Descartes addresses a related claim to Amauld in his second letter of 1648:
But it does not seem to me that it should ever be said of anything that it cannot
be done by God. For since every aspect of the true and the good depend on his
86 Elmar J. Kremer
omnipotence. I would not dare to say that God cannot bring it about that there is a
mountain without a valley or that one and two are not three. But I say only that he
has given me a mind such that I cannot conceive a mountain without a valley, or an
aggregate of one and two which is not three, etc., and that such things imply a
contradiction in my concept. 29
But Arnauld does not accept this distinctively Cartesian position on divine
omnipotence, and hence POI does not have quite ·the same force for him as for
Descartes. This is clear, I think, from his treatment of the Eucharist. Consider
his statement, quoted above: 'only extension is the essence of matter, and im
penetrability is merely one of its properties. But only the essence is inseparable
from it; regarding the properties, nothing prevents our saying that they can be
separated from it by the power of God ... '30 This passage certainly seems to say
that something does prevent our saying that the essence of matter can be sepa
rated from bodies by the power of God. Arnauld does seem to think that to bring
it about that a body is really present somewhere without being extended would
be to make a contradictory proposition true. The obvious explanation of the
above text, then, is that Arnauld thought that extension cannot be separated
from matter, even by the power of God.
Arnauld rejects Descartes's advice more openly a few years later while
commenting on Malebranche' s statement that God is 'powerless [impuissant] '
to do certain things. Arnauld accepts Malebranche's formula that God is power
less to do certain things, and gives as examples of things God cannot do, 'to
raise someone from the dead in answer to the prayer of a false prophet who
would lead an entire people into a false religion by means of this miracle, or to
commit a criminal action such as to make a person who holds a high position
die in order to put himself in his place.' 'All the theologians,' he says, would
accept these examples. Later on he adds the example that God cannot lie (OA
39, 213).
Of course, Descartes also holds that God is not a deceiver. But I know of no
passage in which he says flatly that God is powerless to lie. The strongest
Cartesian pronouncements on the matter that I know of are: ' I recognize that it
is impossible that God should ever deceive me [agnosco fiere non posse ut il/e
me unquamfallat]' (Fourth Meditation, AT 7, 53); 'It is impossible to imagine
that he [God] is a deceiver [deceptor fingi non posset] ' (Second Replies, AT 7,
144); and 'The will to deceive ... cannot be attributed to God [nunquam certe
fallendi voluntas . . . in Deum cadere potest] ' (Principles I, 29). If Descartes
means in these places to say that lying is not within God's power, then he is
acting contrary to his own advice to Arnauld in 1648. However, these passages
to not force us to that conclusion. They say only that we recognize it to be
impossible that God should lie, and hence cannot attribute lying to God. 3 1 But
Descartes also seems to hold that God could make possible what we recognize to
be impossible - indeed, that God could make it possible that contradictories be
true together.
Amauld disagreed with Descartes on this point from the very beginning of
his philosophical career. Thus in the Fourth Objections he attacks Descartes's
claim that God derives his existence, in a positive sense, from himself: 'I think
it is a manifest contradiction that anything should derive its existence positively
and as it were causally from itself' (AT 7, 208). Hence, he concludes, 'God
cannot derive his existence from himself in the positive sense, but can do so
only in the negative sense of not deriving it from anything else' (AT 7, 210). He
adds that ' it will scarcely be possible to find a single theologian who will not
object to the proposition that God derives his existence from himself in the
positive sense, and as it were causally' (AT 7, 214 ). Amauld cannot have been
unaware of the fact that Aquinas also says that God cannot cause himself to
exist, and includes this in his rather lengthy list, in the Summa Contra Gentiles,
of things God cannot do (SCG 1, 25). Indeed Amauld's notion of omnipotence
is close to Aquinas's: For any action A, God can do A so long as neither the
proposition that God does A nor the proposition that God wills to do A is self
contradictory. 3 2
Amauld's willingness, against the advice of Descartes, to speak about things
God cannot do suggests a broader disagreement between the two regarding
possibility and the status of necessary truths. These are large topics, well beyond
the scope of this essay. 33 The point I want to make here is that Amauld disa
greed with Descartes on divine omnipotence, at least in part for theological
reasons. Descartes's position brought him into conflict with the theological
tradition, with ' all the theologians.' This was enough to make Amauld suspect
that the Cartesian position, as it stands, is mistaken and, once the suspicion was
verified, to reject it.
Notes
Steven Nadler, 'Amauld, Descartes, and Transubstantiation: Reconciling

Cartesian Metaphysics and Real Presence, ' Journal of the History of Ideas 49
( 1 988), 229-46. See also Nadler's 'Cartesianism and Port-Royal, ' The Monist 7 1
( 1 988), 5 73-84.
2 Fourth Objections, Points Which May Cause Difficulty to Theologians, AT 7,
2 1 5- 1 6. In response to Amauld's misgivings, Descartes added a clause in the
Synopsis of the Meditations to make clear that the work dealt only with
88 Elmar J. Kremer
speculative truths knowable by the natural light, and not with matters of faith or
morals. See AT 7, 16; Cottingham 2, 1 1. But I know of no passage in his later
works in which Arnauld speaks favourably of the method of doubt.
3 I have discussed Arnauld's late (post- 1680) views on free will in ' Grace and Free
Will in Arnauld, ' in The Great Arnauld and Some of His Philosophical
Correspondents, Elmar J. Kremer, ed. (Toronto: University of Toronto Press 1994),
2 19-39. For Arnauld's rejection of the Cartesian d�ctrine that judgment is an act
of will in favour of the Thomistic one that judgment is an act of intellect, see his
Humanae libertatis notio, in Causa Arnaldina, Quesnel, ed. (apud Hoyoux:
Leodice Eburohium 1699), 99- 1 1 1. A French translation by Quesnel can be found
in OA 10, 6 14-24.
4 Arnauld refers to Descartes as a 'Christian philosopher' in Examen, OA 38, 90.
This work was written about 1680, and published for the first time in OA in 1780.
Arnauld's reference echoes Descartes's self-description, in the letter dedicating the
Meditations to Faculty of Theology of the Sorbonne, where he says that he had
carried out the injunction of the Fifth Lateran Council ( 15 13-17) that ' Christian
philosophers' should try to prove the immateriality of the soul: AT 7, 3.
5 I quote Arnauld's formulation of the principle in Examen, OA 38, 90. Amauld' s
words echo Descartes's Principles I , nos. 24 and 25.
6 Examen, OA 38, 137. Cf. 38, 145-6.
7 OA 38, 73. See Antoine Arnauld, On True and False Ideas, New Objections to
Descartes ' Meditations and Descartes ' Replies, Elmar J. Kremer, trans.
(Queenston: Edwin Mellen 1990), 187.
Descartes does not frequently use the expression ' local extension, ' but he
indicates in the Second Replies that the essence of a body is local extension: 'The
substance which is the subject of local extension and of the accidents which
presuppose extension ... is called body' (AT 7, 16 1). And again, in the Third
Replies, he says, 'Now there are certain acts that we call .. corporeal," such as size,
shape, motion and all others that cannot be thought of apart from local extension;
and we use the term "body" to refer to the substance in which they inhere' (AT 7,
176; CSMK 2, 124).
8 ' Christ' s body, however, is not supposed to be present in a place strictly speaking,
but to be present sacramentally . . . ' (AT 7, 252). The editors of CSMK say that
Descartes is quoting session 13 of the Council of Trent. But there is no reference
to ' local presence' or ' presence as in a place' in the document approved in that
session. However, the Catechism of the Council of Trent, produced under the
authority of the council, states quite explicitly that the body of Christ is not
present in the sacrament as in a place: ' Deinde vero docerant, Christum dominum
in hoc sacramento, ut in loco, non esse. ' Catechismus ex Decreto Concilii
Tridentini, ad Parochos, Typographia Pontificia, Eq. Petri Marietti, 1900, Part II,
Arnau Id's Interpretation of Descartes 89
Chap. IV, #44, 2 1 5 . This is almost a direct quotation from Aquinas, who had said,
' Unde, nullo modo corpus Christi est in hoc sacramento localiter, ' Summa
Theologiae III, 76, 5 .
9 He had taken quite a different attitude towards the same question some three years
earlier in a well-known letter to Mesland: 'As for the manner in which one can
conceive the body of Jesus Christ to be in the B lessed Sacrament . . . the Council of
Trent teaches that he is there "with that form of existence which we can scarcely
express in words" . . . All the same, since the council does not lay it down that "we
cannot express it in words", but only that "we can scarcely express it in words", I
will venture to tell you here in confidence a manner of explanation which seems to
me quite elegant and very useful . . . ' (To Mesland, 9 February 1 645, AT 4, 1 65).
10 Examen, OA 3 8, 1 2 1 . A similar attitude is present in Amauld and Nicole's Grande
perpetuite de la Joi sur l 'eucharistie, which was published in 1 669-73 and is
quoted extensively in Examen. It may be argued that Amauld had changed his
mind on this point between 1 648 and 1 669, but I see no reason to think that he had
done so.
11 Descartes in fact gives this argument in his second letter, with which the exchange
of letters comes to an end.
12 J.-R. Annogathe, Theologia Cartesiana (The Hague: Martinus Nijhoff 1 977), 1 2.
The Scotist theory was opposed to that of Aquinas, who held that, in
transubstantiation, the substance of the bread and wine was not annihilated, but,
rather, converted into the body and blood of Christ. The distinction is a subtle one,
as can be seen by considering the following statement of Occam, who is usually
said to follow the Scotist approach: 'I say that if by annihilation is meant that what
is annihilated is brought back to nothing [redigitur in nihi/o] and not converted
into something else, then in this sense the bread is not annihilated. But if it is
meant that what is annihilated is reduced to purely nothing [reducitur in ita purum
nihi{J such as it was before the creation of the world, then in this sense the bread
is truly annihilated (In IV Sent., Bk I, ordin, 1 , Q. 1 , A. I & 6).
13 Examen, OA 3 8, 1 00-24.
14 Ibid., OA 3 8, 1 0 1 -4. The text can be found in OC 1 , 459-66. This citation of
Malebranche by Amauld should serve as a reminder that the two Cartesian
theologians and philosophers, despite their long and rather bitter controversy, had
a great deal in common. On the friendship and subsequent controversy between
Amauld and Malebranche, see OC 1 7.
15 Examen, OA 3 8, 1 1 1 .
16 OA 39, 1 47. Cited by A.-R. Ndiaye in La Philosophie d 'Antoine Arnauld (Paris:
Vrin 1 99 1 ), 3 23 .
17 ' Dans l e premier exemple [que ce impossible queje ne sois pas, sije pense], c'est
une impossibil ite absolue; parce qu'il y a contradiction, que je pense & que je ne
90 Elmar J. Kremer
sois pas' (Ecrit du pouvoir physique, OA 10, 492).

18 Letter to More, 15 April 1649, AT 4, 342. See also note 9, above.
19 Examen, OA 38, 1 1 1-12.
20 Principles I, no. 26.
21 Summa Theologiae III, 736, 6, ad 2.
22 In the Sixth Meditation he argues that the 'active faculty' corresponding to our
·passive faculty of sensory perception' exists in corporeal things, which produce or
bring about our sensory ideas: AT 7, 79.
23 Cf. 38, 150: ' It is certain, as we have already said, that our body cannot act on our
soul as a physical cause.'
24 Examen, OA 38, 146, 160. It is clear, too, that Amauld realized that Aquinas held
that a material thing cannot prouce an immaterial effect (see Amauld's Dissertatio
bipartita, OA 40, 126--8).
25 Reg/es du bon sens, OA 40, 153.
26 Appendix to the Fifth Objections and Replies,' AT 9A, 208. In the following
sentence, Descartes suggests that people who accepted the truths of the faith
without having any reason to convince them that those truths have been revealed
by God would 'be behaving more like automatons or beasts than men.'
27 Examen, OA 38, 1 14-15.
28 To Mesland, 2 May 1644, AT 4, 1 18.
29 My translation, 195 .
30 See note 1 1, above.
31 Mersenne, in the Second Objections, says ' You [Descartes] say that God cannot lie
or deceive [Deum negas posse mentiri aut decipere]' (AT 7, 125). But, when
Descartes repeats the point before replying, he omits the 'cannot' and instead says,
' In saying that God does not lie and is not a deceiver, I think I am in agreement
with all metaphysicians and theologians past and future [cum nego Deum mentire,
vel esse deceptorum ... ]': AT 7, 142.
32 See the discussion of Aquinas on omnipotence by Cyrille Michon in Olivier
Boulnois, La Puissance et son ombre, de Pierre Lombard a Luther (Paris: Aubier
1944}, 206-- 16.
33 But see the essays by Carraud, Ndiaye, and Solere in this volume.
7
Arnauld: A C artesian Theologian?

Omnipotence, Freedom of Indifference,
and the Creation of the Eternal Truths
V I N C EN T C ARRAUD
The history of philosophy attempts to give an account of what the authors think
on the basis of what the texts say, and it is always dangerous to comment on
matters about which the authors remain silent. None the less that is what I
propose to do here, by raising the question: Why does the Cartesian primacy of
omnipotence among the divine attributes not appear in Arnauld's philosophical
works? In particular. why is the doctrine called ' the creation of the eternal truths'
not present there explicitly? We know that when Descartes took the position
that the eternal truths, though grasped by our mind as immutable and necessary,
are created by God, he was opposing both the mathematicians, like Kepler,
Mersenne, and Galileo, who referred the mathematical truths to the divine un
derstanding as their absolute foundation, and the philosophers, who attempted
to establish an ontology, that is, a common concept of being as thought, under
which to include both God and creation, the infinite and finite. 1 Indeed, we can
trace a series of attempts beginning in the twelfth century, always in philosophy,
to submit God to our concept of the possible, i.e. , to posit a single rationality for
both divine and human understanding. The first figure in this large movement
of thought is probably Abelard, who unhesitatingly submits God, not only to the
principle of non-contradiction, but even to what we may call, anticipating Leibniz,
the principle of reason: From the fact (noted by Augustine) that God cannot go
against his own wisdom and his own rationality, or again, that he cannot cease
following the order he has decided to follow, we jump to the assertion that God
must obey reason. Thus God's freedom is limited by his wisdom, at least that is
what Abelard's censors took his position to be. The questions that we find in the
seventeenth century had already been raised in the twelfth, in particular: Can
92 Vincent Carraud
God do anything but what he in fact does? Can God make anything other than
the best? And from the beginn ing, it is clear that the discussion depends on the
relation between theology and philosophy. It can be shown how this movement,
begun by Abelard, is developed in the commentaries on the Sentences of Peter
Lombard,2 and how the thesis of the univocity of logical truths becomes domi
nant. Without doubt it culminates in Suarez, with the affirmation of the univocity
of being, substance, and truth, both logical and mathematical. 3 To this affinna
tion, Descartes responds peremptorily: 'nulla essentia potest un ivoce Deo et
creaturae convenire.'4 Thus we see how Descartes, alone in his century , at least
in philosophy, resists this dominant movement, by philosophizing, if I may say
so, against the philosophers, i.e., by philosophizing in perfect conformity with
the theological condemnations tirelessly reiterated from the twelfth century on,
beginning with the condemnation of Abelard by the Council of Sens in 1141. 5
In sum, the history of the submission of God to logical necessity, beginning with
the principle of non-contradiction, forces us to place Descartes on the side of the
most official orthodoxy in theology, against the 'emancipation' of philosophy
that gave rise, at the beginning of the seventeenth century , to the concept of
ontology. After Descartes, it is obvious that the three great post-Cartesians,
Spinoza, Malebranche, and Leibniz, complete the history of that emancipation :
the possibles are imposed upon God; the truths that are in the divine under
standing are independent of the divine will. This can have occurred only with
the abandonment, by all of the great post-Cartesians, of the radical thesis
of 1630, the so-called doctrine of the creation of the eternal truths, itself a
consequence of the priority of incomprehensible omnipotence over the other
predicates of God.6 The historical question then naturally arises, whether there
is a post-Cartesian theologian ready to support Descartes' s new beginning, his
decisive break with the philosophical trend, in 1630.
Since we mean to speak about Cartesian theology, not merely about
Cartesian themes present in some work of theology or other, whether positive 7
or speculative, we cannot neglect the fundamental Cartesian assertion of the
incomprehensible power of God. Can we speak of Cartesian theology where the
fundamental and persistent Cartesian thesis that 'God is a cause whose power
surpasses the bounds of human understanding' is ignored, or even where it is
weakened?8 Can we speak of a Cartesian theology where what is for Descartes
the way 'to speak of God worthily' is abandoned?9 In other words, that thesis
ought to be found in anyone who is a theologian trying to develop a Cartesian
theology, and not merely a theologian who happens to be Cartesian in philoso
phy. This is the first motivation for studying Amauld, as well as Fenelon and
Bossuet, figures whom no one will deny were major theologians. 1 0
The complex history of the reading of Holy Scripture by philosophers could
provide a second motive for such an inqui ry. Against Malebranche, who refuses
Arnauld: A Cartesian Theologian? 93
to take account of certain scriptural texts on the pretext that Scripture is 'full of
anthropomorphisms, ' 1 1 especially 'the passages from Scripture which say that
God acts by particular volitions, ' Arnauld does not fail to call attention to the
texts which put his adversary in the wrong, especially concerning the submis
sion of God to the principle of contradiction and to the notion of the possible -
Luke, 1 8: 27: 'What is impossible for human beings is possible for God' ; 1 2 and
concerning the submission of will to wisdom in creation, Ephesians 1 : 1 1 : '[God]
... who works all things according to the counsel of his will. ' 1 3 We could show,
as Amauld does not, that the Cartesian doctrine of the creation of the eternal
truths provides a rigorous metaphysical formulation of several of these basic
scriptural passages, interpreted literally. We could begin with Genesis, cited by
Descartes himself in no. 8 of the Sixth Replies: 'because [God] decided that
certain things should be made, "they are good," as it is said in Genesis, because
their goodness depends on the fact that he willed to make them. ' 1 4
But there is a third motivation, which makes our initial question pressing,
and which will provide the context for its solution: Arnauld' s critique of
Malebranche after the publication of the Traite de la nature et de la grace in
1 680 ( and the Eelaireissements and additions of 1 683 and 1 684 ). In the face of
Malebranche' s emphasis on the claim that truths are imposed upon God him
self, that God is obedient to the single rationality of order, or again, that his
power is subjected to his wisdom (even so far as to make the fantastic statement,
in 1 684, that 'his wisdom makes him powerless' - a statement he qualifies in
1 7 1 2: 'his wisdom makes him, so to speak, powerless'), 1 5 the critical reaction of
a theologian who (like Amauld or Fenelon) counts himself a Cartesian, could,
or perhaps must, be based on the doctrine of 1 630, at least in its formulation in
the Sixth Replies. Descartes had there replied, in advance, to Malebranche, us
ing, in section 6, a set of three terms that Malebranche will also employ: 'Neque
hie loquor de prioritate temporis [ of an idea of the good or the true in the intel
lect over the determination of the will], sed ne quidem prius fuit ordine, vel
natura, vel ratione ratiocinata, ut vocant [that is, the Scholastics] ita scilicet it
ista boni idea impulerit Deum ad unum potius quam aliud eligendum. ' 1 6
Malebranche, like the Scholastics referred to by Descartes, speaks of the priority
of order, nature (that is, 'the inviolable law of creatures and even of the Crea
tor' ), and reason. 1 7 Similarly, in section 8, Descartes says, 'Attendenti ad Dei
immensitatem: manifestum est nihil omnino posse, quod ab ipso non pendeat:
non modo nihil subsistens, sed etiam nullum ordinem, nullam legem, nullamve
rationem veri et boni ... ' 1 8 In both cases we have the Malebranchian triad: order,
law, and reason. But what we expect does not occur. Although the critique of
Amauld and Fenelon bear upon what seems to them to be Malebranche' s denial
of the omnipotence of God, they do not rely at all upon the strongest Cartesian
thesis, and the one most opposite to Malebranche' s position, not even in its
94 Vincent Carraud
formulation in the Sixth Replies. Thus we face the following situation: the
Cartesian theologians, when refuting the Traite de la nature et de la grace,
ought to defend the doctrine of the creation of the eternal truths in order most
effectively to oppose Malebranche, both as Cartesians calling attention to
Malebranche' s abandonment of a constitutive thesis of Cartesian philosophy,
and as theologians upholding the primacy of omnipotence among the divine
attributes. We would expect to see the Cartesian doctrine reappear, at least in
this polemical setting, since it is the one Malebranche rejects most decisively
because it threatens most directly the whole of his system. But what we expect
does not occur. It is this silence that we would like to explain.
To this end, we will concentrate on the years 1685-7, the three decisive years
that follow the publication of the great texts responsible for the post-Cartesian
re-establishment of the univocity of being (or of substance) and the univocity of
truth. The authors are Spinoza, with the Ethics in 1677 (but the 'danger' of
Spinozism is not our concern here); Malebranche, with the publication of the
Eclaircissements to the Recherche de la verite in 1678; 19 and Leibniz, with the
completion of the Discours de metaphysique in 1685. But we will concentrate
on these three years above all because they contain Amauld' s Reflexions
philosophiques et theologiques sur le nouveau systeme de la nature et de la
grace (published in 1685), which crystallizes his opposition to the Traite de la
nature et de la grace, and his correspondence with Leibniz. 20
Philosophy and Theology
According to Arnauld, the relations between philosophy and theology are con
trolled by the Augustinian principle noted in the Fourth Objections: 'Quod
intelligimus igitur debemus rationi; quod credimus, authoritati, ' 2 1 a principle
applied in the division of the Objections themselves into two parts. The ques
tion de Deo, in which Descartes returns to the inexhausta Dei potentia, belongs,
of course, to the philosophical part of the Objections and Replies. Henri Gouhier
is referring to this principle when he speaks of the 'separation' of theology and
philosophy: 'Faith and reason, positive theology and philosophy - Amauld is in
profound agreement with Descartes about their separation as well as the caution
that ought to be inspired by speculative theology. The difference between the
two has to do with their situations. Descartes is a philosopher facing theology;
Amauld, a theologian facing philosophy. '22 If these comments were true in 1641,
do they remain true in 1685-7? Assuming the Augustinian principle, only the
requirements of polemic will justify the intrusion of one domain into the other.
Gouhier gives two examples: When Arnauld is forced to go beyond positive
theology in order to do philosophy, as in Volume III of La Grande Perpetuite,
he is clearly Cartesian; when, on the contrary, the identification of matter and

extension is attacked in the name of theology, as it was by LeMoine in the Traite
de / 'esssence du corps, Arnauld, in Examen d 'un ecrit qui a pour titre: Traite
de l 'essence du corps .. . accumulates references in positive theology in order to
show that the Cartesian philosophy is not opposed to them. If the Reflexions sur
le nouveau systeme de la nature et de la grace are Reflexions philosophiques et
theologiques, that is because Malebranche (and not Arnauld) acts sometimes as
a philosopher and sometimes as a theologian, without any clear principle de
marcating the two roles. Furthennore, says Arnauld (no doubt thinking of the
Meditations chretiennes), 'this extraordinary way of putting words into God's
mouth, in his philosophical discussions, is capable of taking people off guard. ' 23
The separation of philosophy and theology in the Reflexions philosophiques et
theologiques does not have the same theoretical status as the division of diffi
culties in the Fourth Objections. 24
Nevertheless, the last page of the Examen adds a complication to the princi
ple noted by Gouhier. The examination of the question of the Eucharist, like
that of the hypostatic union, allows Arnauld to conclude finnly: the incompre
hensible object of a dogma cannot be the nonn for the philosophical discussion
of natural things. So neither the Eucharist nor the hypostatic union is accept
able as a model for conceiving the natural union of mind and body. 2 5 Thus he
says, ' nothing would be more unreasonable than to hold that philosophers, who
have the right to follow the light of reason in the human sciences, are required
to take what is incomprehensible in the mystery of the Incarnation as a rule for
their opinion when they attempt to explain the natural union of the soul with the
body, as if the soul could do with regard to the body what the eternal Word could
do with regard to the humanity he took on, even though the power, as well as the
wisdom, of the eternal Word is infinite, while the power of the soul over the body
to which it is joined is very limited. ' 26 Arnauld concludes with a principle that
goes farther than would have been necessary, in so far as it posits equivocity
between the human and the divine, a principle which Descartes, no doubt, would
have accepted: ' We would not have those thoughts which mix up everything in
philosophy and theology, if we were more convinced of the clear and certain
maxim that Cardinal Belannine used against the quibbles of the Socinians: "No
inference can be made from the finite to the infinite," or, as others put it, "there
is no proportion between the finite and the infinite." ' 27 This principle is of in
terest to us for three reasons: ( I) it denounces as a confusion the attempt to use
an incomprehensible object of theology as a nonn in philosophy (I note in pass
ing that this criticism goes directly against Pascal); (2) it takes it as basic that
the power of God (here the power of the Word in the act of its own Incarnation),
like the wisdom of the Word, is infinite. Hence, there is no priority of the
96 Vincent Carraud
wisdom over the power of God to justify a priority of philosophy over positive
theology and the marvels that it explicates; (3) thus the two domains are kept
separate in the name of the infinite power of God, and it is our perception of the
incommensurability of the infinite and the finite that rules out an analogy be
tween the two: 'There is no proportion between the finite and the infinite. ' That
is, the affirmation of the incomprehensible omnipotence of God is prior to and
establishes the division of theology and philosophy from each other. Further, in
the Cartesian problematic, the first proposition of the Creed introduces the no
tion of a power that goes beyond the truths of the sciences - in particular, here,
the science of the union of mind and body. In sum, behind an apparently banal
theory of the separation of philosophy and theology, Arnauld, as a rigorous
Cartesian, points to the infinite power of God as the unconditioned condition of
the exercise both of theology (whose object is the mysteries of the faith) and of
philosophical rationality. 28
Thus the above principle has a twofold epistemic function: against LeMoine,
to separate the incomprehensible mysteries of revelation from the work of
philosophy so as to prevent the former from becoming the conceptual models of
the latter; against Malebranche, to prevent the application of philosophical propo
sitions to the divine, that is, to prevent the unconditioned from being condi
tioned by finite rationality. For the same reason, Arnauld, this time confronting
Malebranche, refuses to submit the divine will to causality. I cite the Reflexions
philosophiques et theologiques, Book II, Chapter 3: 'if we are asked why God
has created the world, we should only reply that it is because he wanted to; and
. . . if we are asked anew why he wanted to, we should not say, as the author
[Malebranche] does, that "he wanted to obtain an honour worthy of himself."
The idea of God does not permit us to accept Malebranche' s proposition. We
ought rather say that he wanted to because he wanted to, that is, that we ought
not seek a cause of that which cannot have one. ' 29 Arnauld here relies on a
passage from St Augustine, and again on Estius' s commentary on the same
passage, in which the repetition of the question 'Why?' is declared 'imperti
nent, because there can be no cause of God's volition. ' 30 Arnauld' s refusal, here
against Malebranche but also against Leibniz, to submit God to causality, that
is, to submit his will to rationality in the form of a principle of reason, brings us
back to the Fourth Objections and Replies, though without any mention of the
concept of causa sui, a silence no doubt connected with our present concerns. 3 1
Freedom of Indifference
I know of no text in all the writings of Arnauld that explicitly affirms the doc
trine of the creation of the eternal truths, in any form whatever. This point is all
the more remarkable in that one of the Cartesian passages which enunciate that
thesis is addressed to Arnauld. In his letter of29 July 1648, replying to Arnauld's
question about the impossibility of a void, which seemed to Arnauld to detract
from the omnipotence of God, 3 2 Descartes says: 'As for me, it seems to me that
one should never say of anything that it is impossible for God; for since every
thing that is true or good depends on his omnipotence, I do not dare even to say
that God cannot make a mountain without a valley, or that one and two do not
make three; rather, I say only that he has given me a mind of such a kind that I
cannot conceive a mountain without a valley, or that the sum of one and two do
not make three, etc. And I say only that such things imply a contradiction in my
thought. ' 3 3 Here I will only add a note that I have never found in the commen
taries on Descartes. In the case of the void, we are dealing with a reversal of the
ordinary way of thinking, which is not to appeal to the omnipotence of God.
Ordinarily, we think that one and two make three, and that there is no mountain
without a valley, and it takes nothing less than an exaggerated, hyperbolic, meta
physical doubt to shake us and to suspend our certainty. That is, it takes an
extraordinary appeal to the omnipotence of God, the turning-point which al
lows us to pass from a sceptical doubt to a metaphysical doubt in the First
Meditation. By contrast, the prejudice by which we think that a void is possible
consists in thinking too much about the omnipotence of God, through which the
void, to us unthinkable, would be possible. 'Furthermore, this difficulty arises
from the fact that we count on the divine power [ex eo quad recurramus ad
potentiam divinam] ; and because we know that it is infinite, we do not notice
that we are attributing to it an effect which includes a contradiction in its con
ception, that is, which cannot be conceived by us. '3 4 The possibility of a void is
a case in which we naturally think in a hyperbolic way, because our prejudice
assumes the infinity of God's omnipotence. Thus belief in the possibility of a
void, as an enduring prejudice, calls upon the same principle as the provisional
hypothesis of a God who could permit me to be deceived, namely, his infinite
power. 3 5
On the other hand, in a passage in the Defense ... contre la reponse au livre
des vraies et desfausses idees, Arnau Id uses an analogous reason to prove against
Malebranche that God is not extended and that he is not in an immense space.
The point of interest to us here is that Amauld cites articles 22 and 23 of the
first part of the Principles of Philosophy, which explicitly contain the Cartesian
doctrine of the creation of the eternal truths ('[Deum esse] omnis bonitatis
veritatisque fontem '): 36 'But, to add the philosophers to the theologians [Amauld
has just cited Denis the Areopogite ], I think everyone will agree that what
Descartes says on this point is more worthy of God than the new dogma of his
disciple [Malebranche' s dogma that God is extended]. It is found in the first
98 Vincent Carraud
part of the Principles, articles 22-23 . ' 37 To conclude this brief note: Not only is
Amauld, like all his contemporaries, perfectly well aware of the texts on the
creation of the eternal truths, but one of these texts was addressed to him. In
addition, he is not afraid to cite the passage from the Principles that expresses
the doctrine clearly.
I now tum to the Reflexions philosophiques et theologiques, where Arnauld
attacks the question that concerns us on the basis of God ' s freedom of indiffer
ence. ArnauId' s reading of the first Discourse of the Traite de la nature et de la
grace can be summ arized as follows: If God wills to produce any work wh atever
outside of himself, then he is obliged to produce the most perfect (and to pro
duce it most perfectly). By m aking the simplicity of means necessary, thus by
submitting power to order, that is, the Father to the Son, M alebranche denies
God ' s freedom of indifference. Here we do not need to evaluate the accuracy of
Amauld' s critique. 3 8 But we want to bring out the paradox of his argument: ( I )
The notion of God ' s freedom of indifference is basic in Descartes, a freedom of
indifference that does not have to do with choice, since for God ' s will there is no
question of choosing among possibles. Consider again this constant thesis of
Cartesianism in points 6 and 8 of the Sixth Replies, which we h ave already
cited: 'Repugnat enim Dei voluntatem non fuisse ab aetemo indifferentem ad
omnia quae facta sunt aut unquam fient . . . ' 39 Again, ' [Deus] fu isset plane
indifferens ad ea creanda quae creavit. ' 40 H owever, (2) when Amauld attacks
M alebranche, he relies, not on Descartes, the great theoretician of freedom of
indifference, but on St Thom as. This is all the more strange in that, forSt Thomas,
there cannot be any indifference on God ' s part towards the things produced (or
even towards truths) because there is no equality among creatures. Since things
always differ in degree of being because of their essence, that is, since form al
distinction always requires inequality (' distinctio autem forrnalis semper requirit
inaequalitatem ' - each form taking its place), all creation presupposes in
equality, or again, inequality pertains to creatures ( ' creatura, cui competit
inaequalitas ' ). 4 1 Therefore, although the F ather and the Son are equal within
the Trinity, there is no inequality among creatures, and consequently no possi
bility of indifference on God ' s part towards what is created. Amauld knows this
very well, and hence there is a certain difficulty when he tries to justify qualify
ing the freedom of God as indifferent. He is able to do so only by opposing
it to the necessity with which God loves himself. This appears in a text that is
especially confused : 'It is well to note how little freedom and indifference the
author [M alebranche] leaves to God, with regard to what he brings about out
side himself. I have added, and indifference, for we know that the Scholastics
believed that there was present in God a freedom without indifference and with
out contingency, as in the case of the love he necessarily bears for himself, and
the production of the Holy Spirit, the outcome of that love ... I do not mean to
speak of that sort of freedom, but rather of God's freedom with regard to what he
brings about outside himself, which must be accompanied by indifference [ my
italics], because God 'loves only his own nature invincibly and necessarily. '42
(3) Amauld then cites I, 19, 3, Utrum quidquid vult Deus ex necessitate velit?,
giving a translation/paraphrase: 'Since the divine will has a necessary relation
to his goodness, which is its proper object, it loves it with necessity. But since it
loves all other things only for the sake of his goodness, it does not love them
with necessity because they do not have a necessary relation to the divine good
ness, for that goodness can exist without them and receives no increase from
them.' And after posing the difficulty that what is capable of producing opposite
effects does not act unless it is determined by some other thing, Aquinas replies,
'That is true for a cause that is in itself contingent. But the will of God is God
himself, and consequently is a necessary being, so that it determines itself to
will with regard to the things to which it does not have a necessary relation. '43
Nevertheless it is clear that for St Thomas the determination of the will by itself
does not imply any indifference, for the reasons we have just given. Besides,
when St Thomas says that the divine will determines itself to that which it wills
(' voluntas divina, quae ex se necessitatem habet, determinat seipsam ad volitum,
ad quod habet habitudinem non necessariam '), he does not mean to oppose it to
the intellect, because the attributes of God must never be really distinguished,
above all not temporally (a principle taken up by Descartes). But Arnauld uses
the texts against Malebranche to emphasize the self-determination of the will
and thus imposes a Cartesian reading, in terms of indifference, on what is said
about the divine will: 'Note that he [St Thomas] does not say that it is the
wisdom of God that determines his will, by proposing that to which the will
ought to direct itself [Malebranche's position], but rather that the divine will
determines itself,freely and indifferently, towards all the things to which it does
not have a necessary relation, that is, towards everything that is not God. '44 To
be sure, the argument of St Thomas is entirely opposed to Malebranche, whose
position we stated above: If God wills to produce something outside of himself,
then he ought to produce the most perfect one (and ought to produce it most
perfectly). This is exactly the opposite of the reply to the second objection in
question 13, article 3: 'Deus ex necessitate velit bonitatem suam, non tamen ex
necessitate vult ea quae vult propter bonitatem suam. '4 5 But Arnauld puts a
Cartesian construction on the argument of St Thomas by saying 'freely and
indifferently.' For St Thomas, from the fact that God does not will necessarily
what he wills in view of his goodness, it does not at all follow that he wills with
indifference.
Two points are noteworthy here. First, the St Thomas that Arnauld opposes to
100 Vincent Carraud
Malebranche is a strangely Cartesian one. Even if Arnauld is trying to show

that Malebranche is not a good theologian - a rather easy task, after all46 - that
does not justify his Cartesian reading of St Thomas. Once more we look in St
Thomas for answers to questions he did not ask. 4 7 Second, Arnauld does not
make use of the self-determination of the will ('voluntas divina determinat
seipsam') to call attention to the doctrine of the creation of the eternal truths,
which he could have done without forcing the texts any more than he did when
he introduced indifference. Everything takes place as if it is Descartes who is
behind the text of Arnauld and who is the true origin of his criticisms of
Malebranche, as if the reference to St Thomas were not by itself sufficiently
anti-Malebranchian, and the Cartesian affirmation of freedom of indifference
had to be added in order to bring the attack on Malebranche to completion. 48 But
it seems that this presence of Cartesian theses must remain clandestine. 4 9
In conclusion, it looks as if Arnauld does not want to object against
Malebranche that the submission of the divine power to order, for example, to
the simplicity of means, is a direct denial of the omnipotence of God. He limits
himself to saying that it denies God's freedom of indifference, appealing to St
Thomas but using the vocabulary of the Meditations (to strengthen his polemic).
How should we understand this detour? The doctrine of the creation of the eter
nal truths does not seem to be connected, in Descartes, to one formulation rather
than another, to freedom of indifference rather than to omnipotence ( or vice
versa). Does Arnauld want to maintain God's indifference with regard to the
things created but to deny that the truths are among those things (in which case
he would not have accepted the reasoning about the void in Descartes's letter of
July 1648)? Or does he, by contrast, admit the creation of the eternal truths in
itself, but not dare to defend it, limiting himself to what is necessary for his
polemical purposes? That would be to suppose that Arnauld means to be a
Cartesian without considering himself able to defend philosophically the most
daring theses of Descartes, in the first instance that of the creation of the eternal
truths, but close behind it the thesis that God is causa sui. The correspondence
with Leibniz will enable us partly to clarify the question.
Arnauld's Silence
From the time he first received the summary of the Discourse on Metaphysics,
Arnauld focuses on the central difficulty of paragraph 13: How to understand
the proposition that 'the notion of each individual contains once and for
all everything that will happen to him.' He calls attention to the fact that
the freedom of God is once again in question: 'Therefore God is no more free
with regard to all that [is included in the notion of the individual Adam],
Amauld: A Cartesian Theologian? IOI
assuming that he willed to create Adam, than he would be not to make a being
capable of thinking, assuming that he had willed to create me. '50 The Leibnizian
problematic has to do, not with distinguishing one individual from another (within
a problematic of intersubjective multiplicity), but with understanding the rela
tion of an individual to itself, as a matter of intention, that is, with calling
attention to the ' intrinsic connection' between a possible individual and every
thing that will happen to it. 5 1 In this situation, Amauld takes the problem to be
that of analysing the subject as subject of representation and carries out a two
fold change of direction: (1)'The true notions' are to be sought, not in God, but
in myself. (2) When thinking about that intrinsic connection, we can compare it
to another type of connection, that between a sphere and its properties. I shall
summarize briefly. 5 2
( I ) God is not the one we should question in order to arrive at knowledge of
an individual nature. ' I can hardly believe that we philosophize well if we try to
find out what we ought to think by considering the way God knows things. '53
God' s way of knowing things is forever inaccessible to us: ' We ought to seek
true notions not in God, who dwells in a light that is inaccessible to us ... but in
the ideas that we find in ourselves.' 54 So it is necessary to separate the truth of
things in themselves to which we have no access and the truth of things for us,
or, to speak with Descartes, in relation to our conception. 'The divine under
standing is the rule of the truth of things quoad se; but it does not seem to me
that it is the rule quoad nos, as long as we are in this life. For what do we now
know about God' s knowledge? We know that he knows everything, and he knows
everything by a single, entirely simple, act, which is his essence. When I say
that we know, I mean that we are sure that it must be so. But do we comprehend
it? Ought we not recognize that however certain we are that we know that it is
the case, it is impossible to conceive how it can be?' 55 There follows a well
known criticism of the idea of 'purely possible substance.' We cannot represent
to ourselves the possibles that God has not chosen because that would imply
that we have at our disposal the very knowledge of God, and would imply that
the possibles are not the effect of God's power, but of his understanding
(which is the thesis of Leibniz, as of Malebranche: 'The place of possibles, that
is .. . his understanding'). So Amauld takes up the Cartesian refusal of all repre
sentations of the divine decrees and applies it to the Leibnizian concept of the
possible: 'I am strongly inclined to believe that those are chimeras that we form,
and that everything that we refer to as possible, purely possible, substances, is
nothing but the omnipotence of God which, as pure act, cannot contain any
possibility. '56 It would be hard to be more Cartesian than Amauld is here. The
possible is a concept which has value only with regard to my conception; it
makes no sense in relation to God.
I 02 Vincent Carraud
Since I cannot know 'in what way things are in the cognition of God,' I shall
seek the notion of myself in myself. Once again, one could not philosophize in a
more Cartesian way: ' But I find in myself the notion of an individual, because
I find there the notion of myself. Therefore I need not look elsewhere ... '57 The
development of this important point is beyond the limits of this essay, and I pass
at once to the second change of direction introduced by Arnauld.
(2) In thinking about this intrinsic connection of myself to myself, I can com
pare it with another type of connection: ' Therefore I need only consult this
individual notion in order to know what it contains, just as I need only consult
the specific notion of a sphere in order to know what is contained in it. ' 5 8 With
this astonishing comparison, Arnauld makes it appear that an individual has an
essence (like the essence of the sphere as distinct from its dimensions). The
application of 'the same rule to the individual notion of myself' allows Arnauld
to use the hypothesis of a journey I can take or not take, thus to use the assump
tion of freedom, to reject the inherence ofmy life in my concept. In sum, Arnauld
plays off freedom against individuality in the Leibnizian sense (all my life being
in my concept). We know Leibniz's answer: it is incorrect to consider the same
type of necessity with regard to a sphere and with regard to an individual: the
notion of an individual substance and of a species differ totally. 59 At the level of
the notion of a species, necessity has to do with eternal truths: ' The notion of a
species includes only certain eternal or necessary truths, but the notion of an
individual includes sub ratione possibilitatis matters of fact or what has to do
with the existence of things and time, and consequently it depends on certain
free decrees of God considered as possible.'60 That is why we must 'philoso
phize differently' about the two notions. And that is what Arnauld did not do: ' I
see that M. Arnauld did not remember or at least was not concerned about the
opinion of the Cartesians, who hold that God establishes the eternal truths, like
those that concern the properties of a sphere, by his will; but since I am not of
their opinion, any more than M Arnauld, I will only say why I think one ought
to philosophize differently about the notion of an individual substance than about
the sphere. '6 1 The letter that was actually sent to Arnauld puts the point differ
ently, not accepting Arnauld's initial silence: ' I will say a word about the reason
for the difference ... between the notions of species and those of individual sub
stances, in relation to the divine will rather than in relation to simple under
standing. The highly abstract notions of species contain only certain necessary
or eternal truths, which do not depend on the decrees of God (whatever may be
said by the Cartesians, about whose position on this point it seems that you also
are not concerned); but the notions of individual substances ... must include
[ envelopper] in addition ... the free decrees of God ... '62 There can be no doubt
that Leibniz repeats his point in order to challenge Arnauld to declare himself a
Cartesian on this decisive point.
Thus, Leibniz concludes that Arnauld has abandoned, or has forgotten, the
doctrine of the creation of the eternal truths, from Arnauld' s comparison
between the connection of the sphere and its properties and the connection of
myself and what will happen to me, a comparison Leibniz himself rejects. 6 3
Starting from the fact that Arnauld tried to separate the essential from the
inessential in the notion of the self, that is, the individual notion (to be myself)
from that which is free or decreed (the journey) by comparing the self to a
sphere, whose dimensions are inessential, Leibniz indicates to Arnauld that he
has forgotten that, for a Cartesian, even what is essential or definitional for
mathematical objects is decreed by God, and hence arbitrary and no doubt free.
In other words, Leibniz is convinced that Arnauld has confused physical kinds
with mathematical or metaphysical kinds, 64 and hence needlessly connects the
two terms of Arnauld' s comparison. I would like to conclude by showing
that this treatment of Arnauld is unjustified, and that Leibniz has ignored the
twofold change of direction that was introduced by Arnauld, as I noted above.
(1) In the letter of 26 September 1686, Arnauld writes: ' My only remaining
difficulty has to do with the possibility of things, and with this way of thinking
about God as having chosen the universe he created out of an infinity of other
possible universes that he saw at the same time and that he did not will to create.
But since that does not have to do with the notion of the individual nature, and
since it would take me too far afield to explain what I think about it, or rather
what I object to in the thoughts of others, because they do not seem to me worthy
of God, it is better that I say nothing about it. ' 6 5 Arnauld keeps his thoughts to
himself, both his criticism of the (unworthy) conceptions of God (in Malebranche
as well as in Leibniz) as needing to choose among possibles which are imposed
on him, and his (Cartesian) refusal to distinguish faculties in God, ne quidem
ratione: In God, wisdom is not prior to omnipotence, the understanding not
prior to the will. Can we go so far as to say that he keeps to himself his basic
agreement with the Cartesian doctrine of the creation of the eternal truths, the
only way to 'speak worthily about God' ?66
(2) Consider again the letter of 13 May 1686, more precisely the subtle pas
sage, which Leibniz read perhaps too quickly, that articulates our inability to
comprehend the how of God's knowledge (whence the distinction between quoad
se and quoad nos) and the criticism of our representations of creation: 'We
fancy that before willing to create the world, he [God] envisaged an infinity of
possible things of which he chose some and rejected others. ' 67 In this short
passage, Arnauld wrote: 'Can we so much as conceive that God has knowledge
of an infinity of things that he might not have known because they might not
have existed, he whose knowledge is his immutable and necessary essence? The
same point can be made about his will, which is also the same as his essence, in
which there is nothing that is not necessary. ' 6 8 God knows what he might not
have known because he knows what is. But what is and is known by God would
not have been if God had not willed it. And yet it is necessary for us. Hence
we propose to read this text as Arnau Id's statement, discreet but exact, of the
possibility of the Cartesian thesis of the creation of eternal truths. The inaugural
Cartesian thesis remains an open possibility, at least.69
What was Arnauld's motivation for 'not going further,' in Henri Gouhier's
words, and never maintaining explicitly the so-called doctrine of the creation of
eternal truths? Two sorts of motivation suggest themselves to me. The first was
recognized by Gouhier: the association of Cartesianism and Augustinianism is
possible only if the doctrine of the creation of eternal truths is ignored. 70 As we
have tried to show elsewhere,7 1 because of Descartes a certain number of Augus
tinian texts and theses acquired a new philosophical status in the seventeenth
century, and Amauld contributed to this development. It is thus understandable
that he wanted at all costs to avoid either directly contradicting St Augustine or
showing a fundamental disagreement between Descartes and Augustine.
The second motivation is pointed out by Arnauld himself. It is what we called
his first change of direction in his letter to Leibniz of 13. May 1686: It is not in
God, who dwells in inaccessible light, that we should seek the true notions of
things. There is too much obscurity and difficulty in that quarter. 72 Can this
fundamental reserve be called Cartesian? Certainly, even if Descartes is not its
only possible origin. Whatever the origin of the thesis that I must find in myself
the true notions of things and of myself, it represents an authentically Cartesian
way of philosophizing.
But behind the motives which are properly Arnauld's, I discern a more fun
damental reason. When Amauld falls back on the Cartesian concept of freedom
of indifference, is this not a weakening of omnipotence itself? In the face of the
priority of the divine understanding over the divine power, which characterizes
all the great post-Cartesian systems, that is, in the face of the definitive submis
sion of God, from that time on, to rationality, are the means still available, in
philosophy, to consider omnipotence in all its radicalness and thus to uphold the
creation of the eternal truths? I think not. Descartes, as I said at the outset, was
the only one in his century to do so. He was also the last. To put the point
differently, I see no theological reason that would stop Arnauld, or Bossuet or
Fenelon, from holding the doctrine of the creation of eternal truths. What stops
them is that they, like everyone else at the end of the seventeenth century, lacked
the means or the audacity to contemplate in philosophy the proposition that
essences cannot be attributed univocally to God and to creatures. Furthermore,
the new problematic of theodicy (God summoned to justify himself according to
the standard of common rationality) which will dominate the eighteenth cen
tury, requires a weakening of the unconditioned omnipotence of God. This weak
ening is anticipated, malgre lui, by Amauld's silence.
ArnauId: A Cartesian Theologian? I 05
Notes
It is my pleasant duty to thank Elmar J. Kremer on two counts: for translating this
paper into English, and for comments that led to improvements in the final version.
See Jean-Luc Marion, Sur la theologie blanche de Descartes, 2d ed. (Paris: PUF
1 99 1 ), especially Book I .
2 What is involved here is the application to theology of the methods used to study
the history of long periods of time. See La Puissance et son ombre. La Toute
puissance divine de Pierre Lombard a Luther: Sentences I, dist. 42 a 44,
introduction, translation and commentary under the direction of Olivier Boulnois
(Paris: Aubier 1 994 ). On the formula of Abelard that anticipates the principle of
reason ( ' God does nothing with reason'), known by Odon de Soisson, see ibid.,
Introduction, iv.
3 See Jean-Luc Marion, Sur la theologie blanche de Descartes, 1 1 0-39 and Jean
Fran<;ois Courtine, Suarez et le systeme de la metaphysique (Paris: PUF 1 990),
especially part I I .
4 Sixth Replies, A T 7, 433, 5-6.
5 See Boulnois, La Puissance et son ombre.
6 See Jean-Luc Marion, 'De la creation des verites eternelles au principle de raison.
Remarques sur l 'anti-cartesianisme de Spinoza, Malebranche, Leibniz, ' XV/l e
siecle 1 47, 2 ( 1 985), 1 43-64.
7 See for example the pages concerning Amauld in Francisque Bouillier, Histoire de
la philosophie cartesienne, 3d ed. (Paris: Delagrave 1 968), vol. 2, 1 56--77.
8 AT 1 , 1 50, 1 8- 1 9.
9 AT 1 , 1 46, 1 7.
10 Amauld, Fenelon, and Bossuet, as theologians, are the great lacunae in the classic
work of Henri Gouhier, Cartesianisme et augustinisme au XV/le siecle (Paris: Vrin
1 978), as well as in the articles of Genevieve Rodis-Lewis in ldees et verites
eternelles chez Descartes et ses successeurs (Paris: Vrin-Reprise 1 985). See
especially the article 'Polemiques sur la creation des possibles et sur l'impossible
dans l'ecole cartesienne, ' 1 39-57. Surely the study of this question in Amauld,
Fenelon, and Bossuet is more decisive than in those considered in these reviews of
the Cartesian doctrine, such as Wittich, Calley, Desgabets, Poiret, or even Regis.
11 See Reflexions, OA 39, 1 86, 234-5, etc.
12 A reference already present in Descartes's letter to Arnauld of 29 July 1 648: ' But
it does not seem to me that we should say about anything that God cannot do it' :
A T 5 , 223-4. Cf. the Gospel of Luke 1 : 37 and Matthew 1 9:26.
13 First cited by Arnauld in Reflexions, OA 39, 2 1 9; St Thomas cites it in the sed
contra of Ia, 1 9, 3. See also the Summa contra Gentiles II, 23.
14 AT 7, 435, 30; 436, 3. See Vincent Carraud, 'Descartes et la Bible, ' in Le Grand
106 Vincent Carraud
Siecle et la Bible, sous la direction de J.-R. Armogathe (Paris: Beauchesne 1 989),

277-9 1 .
15 Traite de la nature et de la grace, I , XXXVI II, addition; OC 5 , 47.
16 AT 7, 432, 5-9.
17 Recherche de la verite, Eclaircissement X; OC 3, 140; Traite de la nature et de la
grace, Eclaircissement XIX; OC 5, 1 70; Meditations chretiennes et
metaphysiques, VIII, x; OC 1 0, 66; Traite de morale, l , I, 7 and II, 4, I O; OC 1 1 ,
1 9 and 1 82, etc.
18 AT 7, 435, 22--6.
19 Similarly, the third edition of the Recherche in 1 678 contains the famous
correction of the ambiguous formula about necessary truths 'immutable by their
nature, and because they were fixed by the will of God': I, 2, no. 2; 1 :63. See
Gouhier, Cartesianisme et augustinisme, 1 60.
20 Bossuet's and Fenelon's opposition to Malebranche also occurred during these
years: the Refutation du systeme du Pere Malebranche sur la nature et la grace, as
well as the second part of the Traite de I 'existence de Dieu, probably date from the
winter of 1 687-8. See Henri Gouhier, Fenelon philosophe (Paris: Vrin 1 977), 39
and 1 30. The present essay can therefore be seen as the first tablet of a dyptich of
which the second will be the similar silence of Fenelon in his critique of
Malebranche.
21 AT ?, 2 1 6. 24-5.
22 Cartesianisme et augustinisme, 1 30. See the end of Amauld's letter to Du Vaucel.
29 April 1 680: 'The philosophy of M. Descartes is the most reasonable of all' ( OA
2, 245).
23 Reflexions, OA 39, 1 69.
24 AT ?, 1 97, 1 5-22.
25 Examen, OA 38, 1 56--7. According to Amauld, the Athanasian creed says nothing
about the way in which the soul is united to the body: " Far from agreeing with the
author that the words of the creed require us to believe that the soul and the body
are united in the human being in the same was as the human and divine natures
are united in Jesus Christ, the theologians declare the opposite, holding explicitly
that to understand Athanatius' comparison between the human being and Jesus
Christ in that way would be to fall into the error of Eutyches while trying to avoid
that of Nestorius . . . ' On this question, see the first two parts of the classic work of
Paul Galtier, L 'Unite du Christ (Paris: Beauchesne 1 939).
26 Examen, OA 38, 1 75 .
27 Ibid. ; emphasis added
28 If this Cartesian reading were not correct. we would have to say that the infinite
could not in any case be the object of philosophy, which would seem difficult for
Amauld to defend.
29 OA 39, 433. From the refusal to submit the will of God to causality, we pass to
Arnau Id: A Cartesian Theologian? I 07
Malebranche's application of the same refusal to reason itself: ' When we think of
order, of the eternal laws or truths, we do not naturally seek a cause, for they do
not have one' (Tenth Eclaircissement, OC 3, 133).
30 OA 39, 434. See Augustine. The City of God, XI, XXI.
3 1 See AT 7, 2 13, 8- 16.
32 ' Id vero omnipotentiae divinae derogare videtur' : AT 5, 190. 14- 15.
33 AT 5, 223. 3 1-224. 9: " Mihi autem non videtur de ulla unquam re esse dicendum,
ipsam a Deo fiere non posse� cum enim omnis ratio veri et boni ab eius
omnipotentia dependeat, nequidem dicere ausim, Deum facere non posse ut mons
sine valle. vel ut unum et duo non sunt tria; sed tantum dico ilium talem mentem
mihi indidisse. ut a me concipi non possit mons sine valle, vel aggregatum ex uno
et duobus quod non sunt tria, etc., atque talia implicare contradictione, in meo
conceptu. '
34 AT 5, 223, 27-3: "deinde ex eo quod recurramus ad potentiam divinam, quam
infinitam esse scientes, effectum ei tribuimus, quern involvere contradictionem in
conceptu, hoc est a nobis concipi non posse, non advertimus.' See also the letter to
Mersenne of 27 May 1638, part 2, AT 2, 138, 1-15.
35 See F. de Buzon and V. Carraud, Descartes et /es Principes II: Corps et
mouvement (Paris, PUF 1994), 66--9. On the relation between the doctrine of the
creation of the eternal truths and the question of the existence of a void, see A.-R.
Ndiaye, La Philosophie d 'Antoine Arnauld (Paris: Vrin 199 1), 323-32.
36 AT 8A, 13, 19-20.
37 OA 38, 543. Amauld cites the Picot translation of the first part of the Principles,
and ties together articles 22 and 23.
38 The same reproach is made by Fenelon, in Chapter 6 of his Refutation du systeme
de la nature et de la grace. Regarding the absence of divine freedom of
indifference in Malebranche, see for example the Meditations chretiennes, IX, as
well as the Entretiens sur la metaphysique, IX: ' It is with complete freedom that
God determines himself to create the world': OC 12, 202. Once the world is
created, however, the very idea of freedom of indifference ceases to have any sense
(we could add that it is doubly impossible, since there are no singular thoughts in
the Word).
39 AT 7, 43 1, 27-432, 1.
40 AT 7, 435, 27.
4 1 Summa Theologiae, I, 47, 2, ad 2.
42 Reflexions, OA 39. 598.
43 I, 19, 3, ad 5.
44 Reflexions, OA 39, 599.
45 See also the reply to the first objection in the same article: • Ad primum ergo
dicendum quod ex hoc quod Deus ab aetemo vult aliquid, non sequitur quod
necesse est eum illud velle, nisi ex suppositione.'
46 Arnauld does not fail to note, a propos the Traite de la nature et de la grace, that
'it is quite strange that a theological system, full of so many new thoughts on the
most important questions of religion, has been printed four times without the
approbation of any bishops or doctors [of theology] ' : Reflexions, OA 39, 847.
47 The same point can be made about the creation of eternal truths, I, 44, I , ad 3. The
Third Objection was the following: In mathematics nothing is demonstrated by
reference to an agent cause. The reply requires that we distinguish mathematica ut
abstracta secundum rationem from mathematica in so far as they have being
(whether separate or not). If the mathematica have being, they have an agent cause
in so far as they beings, since they are involved in matter. But if they are
considered in so far as they are abstract, they are not created, hence they have no
agent cause. But for St Thomas, God remains subject to the principle of non
contradiction.
48 The complete elucidation of this hypothesis would require a twofold evaluation
that is far beyond the limits of this essay: on the one hand, an estimation of the
relation of Malebranche to St Thomas with regard to creation (and with regard to
the concept of order that is implicit here); on the other hand. the relation of
Amauld to St Thomas regarding the question of freedom. On the latter point, I
should thank Elmar Kremer for having drawn my attention to Humanae /ibertatis
notio, published in 1699 by Quesnel in the Causa Arnaldina. (A French
translation by Quesnel can be found in OA 10, 6 14-24 .) This 'small work in
Latin' (written on the basis of a collection of Thomistic texts that constitutes the
Disquisitio ... , OA I 0, 625-40), which, according to Kremer, dates from 1688,
seems to indicate a rather precisely Thomistic development in Arnauld on the
topic of human freedom. (See the convincing contribution by Elmar Kremer,
'Grace and Free Will in Amauld,' in The Great Arnauld ... , (Toronto: University of
Toronto Press 1994 ), 2 19-39.) If we can believe a letter to Vuillaret of 2 1 June
1692 (OA 3, 4 9 8 , cited by Kremer in n. 3, p. 232), Amauld's basic work on the
Thomistic theory of freedom would date from 16 84-5 ( ' six or seven years ago'),
hence immediately before the composition of the Reflexions, with which I am now
dealing and in which Thomistic and Cartesian elements are combined into an
ambiguous argument.
49 See also the classic discussion of the third Eclaircissement of the Traite de la
nature et de la grace, which breaks the Cartesian rule that we should never say
that God is not able to do something: 'The wisdom of God thus renders him
powerless in this sense, that it does not permit him to will certain things, or to act
in certain ways' (OC 5, 1 8 1). The point is discussed by Amauld in the Reflexions,
where he mentions lying, causing miracles in order to authorize error. damning the
blessed, etc.
50 G I I, 15.
51 G II, 29.
Arnau Id: A Cartesian Theologian? 109
52 Here I will only point out certain elements of the first letters in the
correspondence. For a commentary on the entire debate (but one that does not deal
with the point on which I am concentrating) see R.C. Sleigh, Jr, Leibniz and
Arnauld: A Commentary on their Correspondence (New Haven, CT: Yale
University Press 1 990).
53 G IL 1 9.
54 G IL 32.
5 5 G 11. 3 1 . It is not necessary to insist once again on the Cartesian Thomism of
Arnauld.
56 G IL 32 1 .
57 G IL 32.
58 Ibid.
59 The compete notion determines an individual by the conjunction of all its essential
and existential attributes: The notion of a circle is not complete if it contains only
its essential predicates (see G. II, 39, 45, 52). That is why there is no complete
notion of geometric figures, or of species of substance (see also the Textes inedits,
published and annotated by Gaston Grua [Paris: PUF 1948), I, especially p. 3 1 1.)
Recall that, for Leibniz, in contrast to Amauld, the cogito does not give a complete
notion of my individuality (G I I, 32,33, 45. 52, etc.).
60 Remarks on a letter of ArnauId, in the letter of 21/3 1 May I 686 to the Landgraf
Ernst von Hessen-Rheinfels, G II, 39.
6 1 G I I, 38-9. These remarks of Leibniz have been noted and commented on, in a way
close my own, by Ndiaye in La Philosophie d 'Antoine Arnauld, 332-43, where he
discusses a ' certain difficulty' of Amauld's.
62 G IL 49.
63 This point all the more remarkable in that Leibniz himself began paragraph XIII of
the Discourse on Metaphysics by comparing the consideration of the notion of an
individual substance with that of the nature of a circle: 'by considering [the) notion
[of an individual substance] one can see there everything that can be truly said of
it, just as we can see in the nature of a circle all the properties that can be deduced
from it. ' The comparison has to do with vision (that is, with the intelligibility of
the determinations), not with deduction (because the one consititutes a necessity
ex hypothesi and the other an absolute necessity). But what is at stake in the
correspondence is the contrast between the notion of a species and the notion of an
individual.
64 G II, 13 I .
6 5 G I I, 64.
66 AT 1, 1 46, 1 7.
67 G II, 3 1.
68 Ibid.
69 One might cite against this hypothesis, a difficult passage in the Reg/es de ban
1 10 Vincent Carraud
sens, where Arnauld writes: 'Things, properly speaking, are the substances created
by God; whereas truths, and propositions demonstrated in the sciences, are not
things created by God ' (OA 40, 1 67). In response, two main points need to be
made:
( 1 ) The question Arnauld is dealing with here has to do first of all with the use of
'thing, ' a notion strictly defined as 'substance' ; and while Descartes grants to
truths the same status as 'all the other creatures' (AT 1 , 1 45, 9- 1 0), and does not
hesitate to say that God 'is the sole author on which all things depend ' (AT 1 , 1 50,
7-8), indeed, goes so far as to use the vocabulary of creation with regard to truths
(AT 1 , 1 52, 20 and 27, for example), he would clearly not say that truths are
created things if by that is meant that they are substances, that is, existents; his
intention is to assign the same ontological status to essences as to existents (AT 1 ,
1 52, 2-5, 28-9). This is the origin of the two types of eternity distinguished by
Arnauld in the Dissertatio bipartita, which he uses to 'Cartesianize' a Thomistic
thesis: that essences, in so far as they are possibles, do not have the same eternity
as God.
(2) In 1 693, the date of the Dissertatio bipartita, followed by the Reg/es de bon
sens, Arnauld, faced with his new adversaries, Gommaire Hyygens and Franc;ois
Lamy, settles on positions that are ostensibly Thomistic, once again as a polemical
strategy, at least at the outset, and the usual Augustinian corpus is, rightly or
wrongly, conceded to his adversaries. The fact remains that Arnauld is silent
precisely at the point where we might expect him to join battle, namely, at the
point where, in order to set himself at once against what he will present as two
parallel errors, he opposes Malebranche and Huygens. Malebranche' s error is the
vision of all creatures in God, because bodies are not visible through themselves,
not the vision of truths in God, for truths are merely relations. (On this last point
Malebranche 'is entirely right' : OA 40, 1 58.) Huygens, by contrast. holds that we
see only the necessary and immutable truths in God. In the face of these upholders
of the vision in God, vision of things or vision of truths, Arnauld avoids both
alternatives with the help of St Thomas, not Descartes. Nevertheless. it seems to
me that a careful analysis of the Dissertatio bipartita would enable us to find there
the same ambiguous Cartesian Thomism that we found earlier in the Reflexions.
70 Cartesianisme et augustinisme, 1 56. See also the illuminating remarks of
Genevieve Rodis-Lewis in ' Augustinisme et cartesianisme. ' L 'Anthropologie
cartesienne (Paris: PUF 1 990), 1 0 1 -25 .
7 1 ' Arnauld: From Ockhamism to Cartesianism, ' in Descartes and His
Contemporaries, R. Ariew and M. Grene, eds. (Chicago: University of Chicago
Press 1 995).
72 G II, 32.
8
Arnau Id' s Defence of
Miracles and Its Context
GRAEME HUNTER
The Background
It is sometimes supposed that. if a thing is to excite our wonder, it must be rare.

In particular the wonderfulness of miracles has been attributed by some philoso
phers to their being seldom seen. But during the Middle Ages this inverse law
seems not to have held, for, as is well known, miracles at that time were both
highly regarded and frequently attested. 1 Moreover, at the zenith of that age, St
Thomas Aquinas provided the philosophical background against which mira
cles could credibly be said to be both commonplace and provocative of wonder.
The crux of his account is his definition of miracles as events which it exceeds
the power (facultas) of nature to perfonn. 2 Paraphrasing him, we could say that
miracles are natural events with supernatural causes.
This definition raises the question of what is meant by nature. St Thomas
understands it to be the order of what he calls 'secondary causes,' an order
created and presided over by the primary cause, God. 3 All creatures are second
ary causes, endowed by God with characteristic powers according to their kind.
Within a stipulated range, the use of these powers is left up to the discretion of
creatures, at least of animate ones. A dog, for instance, can bark or not when a
rabbit runs by. The rabbit can increase its pace or dodge into a hole. A com
pletely natural event is then a transaction among creatures which can be wholly
explained by reference to their creaturely powers. Now, since God is free to
suspend, limit, or enhance his creatures' powers from time to time as it suits his
purposes, when he does so, by definition, a 'supernatural' event occurs. And
such events are what are properly called 'miraculous, ' because they depend on a
1 1 2 Graeme Hunter
power outside the natural order of secondary causes.

Yet to confine miracles outside the natural order is not to banish them from
experience. On the contrary, St Thomas's account of them could be called
'centripetal,' for it drives miracles right into the middle of things. M iracles
always occur just when intervention is least expected, right in the midst of ordi
nary events. That is why they are sources of wonder. Miracles so understood are
also by the same token local events. If, for example, Lazarus is raised from the
dead, it is not a victory over death everywhere and forever. It is a reversal of the
process of bodily decay in only a single human body, and only for a season.
Some theologians distinguish therefore between resurrection, which is global,
and resuscitation, which is local. For St Thomas, resuscitation, so understood,
would be a model of miracles generally. They are mainly local disruptions in the
natural order.
This is why miracles can be simultaneously frequent and wonderful. It is not
their rarity, but the intrinsically awesome fact of supernatural intervention in
ordinary life, which is the object of human wonder. 4 Thus the proliferation of
miracles in late medieval practice finds a basis in rigorous Thomistic theory. I
shall refer to that happy combination of theory and practice as the Catholic
status quo.
In the modem period the Catholic status quo was under attack on two fronts.
Protestants were proposing a more stringent criterion for miracles, namely,
that they not merely be wonderful but in addition pass the test of conformity
to the teachings of scripture. Calvin refers to this desideratum as 'the test of
Doctrine,'5 and Luther formulates it as follows: 'the rule is this: Regardless of
their size and number, no wonders or signs are to be accepted contrary to estab
lished teaching ... The only preventive is to have a good grasp of doctrine and to
keep it before your eyes continually. '6
The reasons given by the reformers themselves for their more demanding
criterion were not accepted by their Roman Catholic critics. The reformers said
that they were trying to purify the church of superstition and idolatry; the coun
ter-reformers alleged that the Protestants were only attempting to make less
conspicuous their own lack of miraculous confirmation, or, as Luther himself
put it in his inimitable parody of their attacks, the fact that 'the Lutherans had
never even cured a lame horse.' A neutral way of characterizing the Protestant
doctrinal test of miracles would be to say that it grew out of Luther's general
principle: so/a scriptura.
The second attack on the Catholic status quo originated with the apologists
for the modem scientific outlook and was due to the new science's reconception
of nature. What for the medievals had been a community of creatures (second
ary causes) came to be understood by the modems as a vast system of matter in
Amauld's Defence of Miracles 113
motion, obedient to simple physical laws and describable in mathematical terms.

Divine power seemed to be the on ly motor either necessary or sufficient to
account for the motion of matter, and accordingly the doctrine of secondary
causes plummeted into disrepute. 7
Some of the early modems wished to explain miracles away altogether. This
seems to be pretty clearly the case with men like Spinoza, Hobbes, and Thomas
Burnett. Others were willing to keep them, provided they were rare and deco
rously absent from local events of concern to the sciences. To the latter group of
philosophers must be counted such figures as Malebranche, Leibniz, and Samuel
Clarke. Of course these very different thinkers differed also in the detail of their
accounts of miracles. Robert McRae has recently clarified the extent to which
Clarke ' s (and hence, Newton's) understanding of miracles, which could be called
· statistical, ' was at odds with Leibniz's power-based account. 8 But they were
able to agree at least that there is a general order of things which, due to God's
foresight, holds no (or negligably few) miraculous surprises for the practising
scientist. Even those of the modems who are sympathetic to miracles do their
best to confine them to the periphery of experience. This centrifugal account
is prefigured, like so much of modem philosophy, in Part V of Descartes's
Discourse on Method. The only reference to miracles made in that place is
to the ' miracle of creation,' which, though conceded to be a fact, is compared
unfavourably from the scientific point of view with a kind of universal,
mechanical evolution which, Descartes says, was a possible blueprint of our
origins, though not the one actually chosen by God. 9
It is instructive to note in this connection that St Thomas does not recognize
creation as a miracle at all, since it establishes, and hence cannot contravene,
the natural order. 1 0 Yet creation is the only miracle that the advocates of the new
science can comfortably admit. These philosophers demand a tame, predictable
first mover who initially will oblige them by getting things off to a good start
and thereafter will govern in undeviating adherence to a scientifically detect
able pattern. Such a creator has the twin virtues of accounting metaphysically
for the order of nature and of guaranteeing its permanence. And he accom
plishes both of these tasks without any of the erratic intrusions on the system of
matter in motion which caused the God of Abraham, Isaac, and Jacob to fall
from the scientists' favour.
Arnauld
Antoine Amauld had some of the right credentials for being on the modem
team. As one suspected of sympathies with the Reformation he might have been
supposed to share some of the reformers' concerns about medieval excesses
114 Graeme Hunter
in the matter of miracles. Since he was a Cartesian, one would anticipate his
allegiance to the demands of the new science. But Arnauld surprises us on both
counts. He defends instead the church's historic position on miracles, holding
them to be interventions of God in the normal course of events, and allowing
them to be frequent, local, and confirming of doctrine, rather than dependent on
it. The questions to be examined in this essay are, first, why Arnauld defended
the Catholic status quo, and, second, how well his defence of it holds up in the
modem setting.
La Sainte £pine
To guess at thinkers' motivations is often a risky business, but it is less so in the

present case. The circumstances of Amauld's life easily explain his dissent from
the modem consensus on miracles. He was personally involved with a sequence
of miracles which fascinated Paris, rocked the Jesuit establishment, and deter
mined to a large extent the fortunes of Port-Royal, beginning with the so-called
miracle of the holy thorn. From the viewpoint of posterity, the most significant
achievement of this miracle was undoubtedly that it inspired Pascal's Pensees.
But it also led Arnauld to defend it against the polemics which it excited and so
awakened his philosophical interest in the question of miracles. In order to
evaluate his apologetic position, it is important to be familiar with the main
circumstances of this story.
On 24 March 1656 Mademoiselle Marguerite Perrier was just approaching
her tenth birthday. She was a pupil in residence at one of the 'petites ecoles'
attached to Port-Royal, probably so that she could be under the supervision of
her uncle, Blaise Pascal. Mlle Perrier was originally from Clermont, where her
family still resided. At the age of six she had developed in the corner of her left
eye, next to the nose, what the doctors of the day called a 'fistula lachrymosa,'
which means literally a conduit out of which tears (and other fluids) continually
leaked. Pascal, who had seen his niece in Clermont, at the beginning of her
illness, described it as having originally consisted in 'some drops of water which
fell out of the corner of the said left eye next to the nose, the discharge rapidly
growing more frequent and viscous and eventually changing into something
like mud [boue] . ' 1 1
She was seen by some of the best doctors of France, who unanimously
recommended cauterization, though they could not guarantee that the eye, or
even the child, would survive the ordeal. They knew, however, of no other treat
ment. It is not difficult to see why this severe and risky recommendation led the
family to try first some of the gentler remedies proposed by doctors of lesser
stature. These were pursued right up until August 1655, with the result that the
Amauld' s Defence of Miracles 115
eye got continually, though slowly, worse. By that time Mlle Perrier was a pupil
at Port-Royal, and her father recommended, by the intermediary of Pascal, that
all treatment be discontinued, to see whether unaided nature might not effect a
cure. Instead a tumour about the size of a filbert nut developed, and the girl
began to lose her sense of smell. In addition the fluid, which by now gave off
such an odour that Mlle Perrier had to be segregated from the other girls, was
also flowing into her nose and mouth. She had grown weak, pale, and emaci
ated and had difficulty sleeping because of the constant discharge. Once again
the unanimous recommendation of the doctors was cauterization. Pascal was
alarmed about his niece's condition and wrote a letter to her father in Clermont,
requesting that he come as soon as possible to Paris for the operation. The latter
set out for Paris on 29 March 1656, still believing that his daughter would be
operated on when he arrived.
Meanwhile, the monastery of Port-Royal had received on loan from a certain
M . de la Potterie, a high-ranking Parisian churchman, a relic from his collec
tion, said to be a thorn from the crown of Christ. Apparently it had been making
the rounds of some monasteries prior to its arrival at Port-Royal. It was deliv
ered there on the Friday of the third week of Lent, 1656, which in that year was
the 24 March. The poet Jean Racine, in his account of the miracle, notes that the
introit of the prescribed mass for that day was drawn from the words of Psalm
86: 1 7, " Fae mecum signum in bonum ... ,' in the King James version: ' Shew me
a token for good; that they which hate me may see it, and be ashamed: because
thou, Lord, hast holpen me, and comforted me.'
To the members of the community of Port-Royal, these words would have
been deeply significant, because their monastery was under steady and bitter
attack from the Jesuits on points of doctrine. The community was thus collec
tively hoping for a sign of some kind to vindicate them in what they regarded as
the unjust attacks of their enemies, one thrust of which, incidentally, was to
demand the closure of the very school which the Perrier girl attended.
Following vespers that evening all the schoolgirls filed past the relic and,
when it came the tum of Mlle Perrier, the school-mistress said to her: 'Com
mend yourself to God and touch the thorn to your eye.' She did as bidden.
Following the ceremony the girls returned to their rooms. Scarcely had Margue
rite Perrier reached her room when she said to one of her companions: 'My
sister, I am no longer in pain. The holy thorn has cured me.' Someone, who may
have been Amauld, responding to a later Jesuit attack on the miracle, describes
the rapid sequence of events as follows:
The source of the mud [boue] which was flowing continually from her eye, her nose
and her mouth, and which had stil l been flowing on her cheek a moment before the
116 Graeme Hunter
miracle, as she herself declared later in her statement, was completely dried up.
The bone, which had been decayed and rotten, was restored to its original condi
tion. All the stench which had been associated with her wound, and which was so
intolerable that the girl had to be separated from the others by order of the doctors
and surgeons, was changed into breath as sweet as an infant's. At the same time
she recovered her sense of smell, and none of the ills which were a result of the
main one returned. Even her colour, which had been pale and leaden, became as
lively and clear as it ever had been. 1 2
The miracle of the holy thorn became a cause celebre in Paris, to such a
degree that the Queen Mother and the King sent their personal physicians to
join the large party of other prominent doctors and religious authorities investi
gating it. The physicians examining Mlle Perrier after the event included sev
eral who had also attended her before it. They concluded unanimously that there
were no medical causes for the cure. Even the hostile Jesuits were forced to
recognize that a miracle had occurred, and they confined their criticisms to
saying that there can be miracles which vindicate neither sects nor doctrines. 1 3
The fact that the Jesuits took up this typically Protestant position in attacking
the miracle of the holy thorn meant, of course, that anyone setting out to defend
it would have to occupy different ground. And it is not surprising that advocacy
of the miracle fell to the lot of Arnau Id, it being a crucial event in the life of the
Port-Royal community, to whose defence Arnauld had consecrated his life.
Against this background, then, it is understandable that the entire argumenta
tive strategy of Arnauld's apologetical essay on the authority of miracles con
sists in proving that the Jesuits have taken up the cause of the Calvinists, while
he (Arnauld) is defending, not merely the Port-Royal Community, but the church
and God himself. 1 4
Circumstances thus determined not only that Arnauld would defend local
miracles, but also that his apologetic would take the traditional theological slant
that it did. Furthermore, the fact that the miracles connected with the holy thorn
seemed to come dramatically at the moment of Port-Royal's greatest need would
certainly have encouraged Arnauld to look upon them as doctrine-confirming,
rather than as in need of doctrinal testing. He writes: 'Since these miracles
occurred in the middle of the worship services and in such propitious circum
stances, there was no one who failed to recognize them as marks of God's pro
tection of this community ...' 1 5 In short, Arnauld's deep commitment to the
monastery of Port-Royal led him to understand the events connected with the
holy thorn as local miracles and to see them as direct interventions of God
intended for confirming doctrine, specifically the Jansenist doctrines which in
formed both the religious life of Port-Royal and the teaching at the petites ecoles.
Amauld's Defence of M iracles 1 17
A defence of particular miracles is unlikely to advance very far, however,

in the absence of a general account of them. The writing of the Port-Royal Logic
in the years following the healing of Marguerite Perrier gave Amauld an oppor
tunity to formulate just such a general account.
Miracles in the Port-Royal Logic
The Art of Thinking 1 6 was, of course, co-authored by Amauld and Pierre Nicole.
However, we have it on the authority of Racine that Amauld was singlehandedly
responsible for its fourth part, concerning method, and it is there that the chap
ter devoted to miracles is found. 1 7 Briefly, Part IV of the Logic studies each of
the three ways in which belief can qualify as knowledge. If a belief is self
evident. it is knowledge of the type called in French intelligence; if it is arrived
at by sound proof, then the conviction so produced is called in French science.
In the third place, if the belief in question is based on authority, then the knowl
edge it produces is called 'testimonial' (in French /o i). 1 8
The chapter on miracles i s found within Amauld's discussion of this third
branch of knowledge, the one that is based on authority. That setting militates
against a correct understanding of the chapter for two reasons, which ought to
be noted in passing. The first is that we tend, by an inveterate modem prejudice,
to regard all appeals to authority as suspicious, if not actually fallacious. With
sufficient attention, that prejudice can be overcome, but the second difficulty is
more serious : we are likely to be led astray by a deceptively ambiguous term
which is pivotal in Amauld's discussion of this type of knowledge.
The word in question is Joi, which, in some contexts, must be translated as
'faith' and, in others, as 'testimony.' 1 9 What makes it treacherous in the present
setting is that in a number of cases both translations make sense, though each
leads in a different direction. The sentence with which Amauld introduces the
third (authority-based) type of knowledge is a good example. The original French
says : 'Si c'est l'autorite qui fait que l'esprit embrasse ce qui est propose, c'est ce
qu'on appelle fo i .'20 This has been translated as : " If it is authority which leads
the mmd to embrace what is proposed to it. this is what is called faith. '2 1 So
translated, it yields a not implausible defin ition of faith: faith is believing on the
authority of another. But the word 'testimony,' or its adjectival derivative 'testi
monial,' fits here just as well. In fact, when I paraphrased the same French
sentence above, I said that, when belief was based on authority, the knowledge it
produced was called testimonial. When the sentence is translated in this way, it
becomes, not a defin ition of faith, but instead a classification of knowledge.
Although the sentence taken in isolation makes the translation of ' foi' by 'faith'
look, if anything, more plausible, its context, which is precisely a classi fication
1 18 Graeme Hunter
of types of knowledge, makes ' testimony' (or 'testimonial') a better translation.

Similarly, in Chapter 1 2 of Part IV of the logic, when Arnauld returns to this
matter, the same decision must be faced time and again by the translator.
The title of that chapter ought to read (in English): 'Of what we know through
testimony, whether human or divine.' However, a usually reliable English trans
lator, the one with whom I have already quarrelled regarding an earlier occur
rence of ' foi,' once again puts 'faith' where ' testimony' should be, despite the
fact that no literal sense can be made of the idea of '·divine faith.' ' Foi' is a small
word, but this is not a small point. The proprieties of translation have to be
established already, because the discussion of the connection between knowl
edge and testimony in Chapter 1 2 is the first step on the way to Arnauld' s
discussion of miracles in Chapter 1 4 .
In the intervening chapter (IV, 1 3), Arnau Id formulates rules for the accept
ance of human testimony. Where they are followed, Arnauld suggests, it is pos
sible to attain to moral certainty in matters of this kind. In Arnauld's concise
discussion, one can discern three rules for reasoning well about testimony. The
first rule is to recognize a realm of merely contingent matters, including
paradigmatically those which depend on free human decisions. To call them
contingent is to say that they are not necessary, and therefore to imply that it is
futile to expect apodictic certainty in any belief which we may form concerning
these events. Arnauld understands such contingency (or non-necessity) to in
clude independence of physical, as well as logical necessity, as he explicitly
says. 22 Rule two is to recognize that the fact that we abandon the search for
necessity does not mean that we have to settle for bare possibility. Testimony
does not become credible merely because the events it describes are possible.
Rule three is formulated against the background of its predecessors: To know
whether or not to believe testimony concerning a certain purportedly contingent
event, we must first examine both the internal and the external circumstances of
that event. By 'internal circumstances' Arnauld means what belongs to the event
itself, such things as its consistency and plausibility; by ' external circumstances'
he means especially the testimony which leads us to believe that such an event
took place, and the reliability of the witnesses who are responsible for that testi
mony. If our examination reveals both the internal and the external circum
stances to be such that falsity in similar cases would be very rare, then we can
have moral certainty that the witness's testimony is true. This is the highest
degree of certainty which we are able to attain, since, by hypothesis, we are
dealing with matters of contingent fact.
An example is desirable at this point, and one is provided by Arnauld in the
subsequent chapter, entitled ' Application of the foregoing rule to the belief in
miracles.' Miracles are paradigm cases of events which are (in Arnauld's sense)
Arnauld's Defence of Miracles 1 19
contingent with respect to the order of nature, and of which we become aware by
testimony. Arnauld, of course, did not intend to exclude the possibility of one's
being an eyewitness of a miracle, and hence not needing testimony for it, but,
for all the miracles on which the Christian faith depends, reliance on human
testimony is necessary and the same would apply to any contemporary miracle,
if it were to be widely used for the confirmation of doctrine. In considering
when testimony concerning miracles should be accepted, and when not, we must
apply the rules discussed in the previous chapter.
It must be remembered that the general aim of the Port-Royal Logic is to
' form our judgment' in such a way that we will exhibit 'good sense and accu
racy of thought, in discriminating between truth and falsehood. '23 In the case of
miracles, then, this places us under a twin obligation. The first is to avoid un
warranted scepticism in the consideration of miracles. They cannot be ruled out
a priori, because the events they describe are not logically impossible. To try to
eliminate them in this way is to fail to recognize the domain of the contingent,
to which the previous chapter of the Logic carefully laid claim. But, in the
second place, one must also avoid the opposite extreme of credulity, the vice of
believing things on the strength of their possibility alone. Where scepticism
perversely overlooks the rightful inhabitants of the realm of contingency, credu
lity populates it with fabulous ones.
The key to avoiding both extremes lies in observing the third rule from the
previous chapter, the one concerning the examination of the internal and exter
nal circumstances surrounding a miraculous event and the testimony in its fa
vour. In order to permit his readers to apply this rule, Arnauld mentions several
miracles and recounts at length the details of one performed by St Augustine
and described by him in The City of God (BK. 22, ch. 8). The purpose of the
itemized description is to enable the reader to decide whether there is any a
priori reason why the events could not have happened precisely as Augustine
described them. Briefly and neutrally, the facts are that a brother and a sister on
two public religious occasions, which followed each other in close succession,
were released from a condition of convulsive trembling, common to their entire
family. Augustine was the eyewitness of the second miracle, which occurred
while he was conducting a service of thanksgiving for the first. There is, Arnauld
assumes, no a priori reason why such things could not have happened. And if
they happened in the religious context described, Arnauld claims, 'no reason
able person could fail to recognize the finger of God in them. '24
With the internal circumstances of these miracles thus satisfactorily exam
ined, Arnauld turns to the external ones, that is, the reliability of Augustine's
testimony. Amauld first points out the psychological improbability of any sober
person's lying about so public an event. In the second place he reminds us of the
1 20 Graeme Hunter
general probity of Augustine in particular as a witness to anything, of his hatred

of lying as a crime and his fear of it as a sin. The conclusion of Amauld is that
• every reasonable man, even if he is not pious, must recognize as true the mira
cles that St. Augustine describes ... ' 25
No further study of chapters 13 and 1 4 is necessary to see that Amauld's
interest in the problem of miracles is an epistemological one. He stipulates the
existence of a general field of contingent events, which would include any that
were miraculous, i.e., any which were inexplicable by reference to secondary
causes alone. The philosophical task, as Amauld conceives it, is only to show
that testimony presenting the details of a putatively miraculous event must be
evaluated according to the criteria usually applied to testimony about contin
gent matters of any kind. Presumably this would entail such things as making
sure that our informants are not providing the description for frivolous or obvi
ously self-serving ends. Of course, even a testimony which was self-serving or
frivolously given might still be true, but it would clearly be unworthy of belief
by a third party. Next we must ascertain such things as whether the witness's
general integrity and competence is a matter of established fact and whether
there are corroborating witnesses. If the answer to all such inquiries is affirma
tive, then we have grounds for believing that the event attested to really oc
curred. I f, in addition to that, the event (which by hypothesis is scientifically
inexplicable) occurred in appropriate religious circumstances, such as those in
which it appears to be an answer to prayer, or to obvious needs of a religious
community, or to unresolved questions of doctrine, then we may be morally
certain that it is a miracle. Such, essentially, is Amauld's apology for miracles
in the Port-Royal Logic.
One thing may puzzle us about this account of miracles and several aspects of
it may trouble us deeply. The puzzling thing is why Amauld never directly
mentions the miracle of the holy thorn, though some of what he says strongly
suggests that he is thinking of it. The most likely explanation for the omission is
that any testimony he had given in favour of the holy thorn would have ap
peared to be self-interested. 26 It is a less straightforward matter to reply to the
philosophical reservations which many may have about his account of miracles.
To them we now tum.
It is not necessary to be creedally opposed to miracles in order to be hesitant
about Amauld's defence of them. In the first place it is disconcerting to see him
approach the matter as a logician or epistemologist, rather than as a metaphysi
cian or philosopher of science. For, if the critics of miracles in the seventeenth
and eighteenth century agreed upon anything, it was surely that the possibility
of miracles, as tradition conceived them, meant the impossibility of science as
they understood it. Thus it is hard to think any treatment of miracles complete
Arnauld's Defence of Miracles 121
that does not discuss them in scientific terms. A second (related) difficulty with
Amauld's account is that he seems to evaluate testimony concerning miracles as
if it were no more intrinsically problematic than testimony on any other contin
gent subject. We post-Humeans, on the other hand, have been conditioned to
think that we have a priori reasons for being particularly suspicious of such
testimony, reasons which derive from the a priori improbability of miracles as
such.
The criticism of Hume's which touches more than any other on both these
points is the one in which he says: 'no testimony is sufficient to establish a
miracle unless the testimony be of such a kind that its falsehood would be more
miraculous than the fact which it endeavors to establish. ' 27 What Hume implies
is that, since a miracle is 'a violation of the laws of nature,'28 it has a probability
of 0.29 And, of course, no testimony in favour of anything, let alone miracles,
will ever have a probability of 1 . As Hume says: 'it appears that no testimony for
any kind of miracle has ever amounted to a probability, much less to a proof;
and that, even supposing it amounted to a proof, it would be opposed by another
proof derived from the very nature of the fact which it would endeavor to estab
lish. ' 30 Thus even a testimony that could be faulted from no other angle would
be disqualified for the very reason that it was attesting to a miracle. In other
words, miracles, according to Hume, can be ruled out a priori, and therefore can
never be fit subjects of testimony in the first place. The cultural persuasiveness
of this argument is one factor, I suspect, in the hesitation many people would
feel about Arnauld's account of how knowledge of miracles is possible.
Though Hume' s probabilistic criticism of miracles and Arnauld's epistemic
discussion of testimony have quite different thrusts, it is not impossible to ex
press their differences in a common idiom. The basis of their disagreement is
that Arnauld affirms what Hume denies, namely, that there can be events, con
tingent with respect to physical laws, the cause of which is a direct act of God.
Hume, by insisting on the exceptionless regularity of nature, seems to be
stating only a scientific truism, and therefore to be in a more defensible position
than Amauld, who allows for scientific anomalies. However, on closer scrutiny,
Hume loses his advantage. Let us see why.
First, an assumption in Hume's account can be challenged. He assumes that
the witness to a miracle must be claiming that a miracle occurred. But this is not
so. The witness need only be testifying to the occurrence of a naturally inexpli
cable, i.e., contingent, event, such as happened with the application of the holy
thorn. That the sequence of events there described could occur, and could do so
in the order described, is not a logical impossibility. And it is not primarily the
witness, according to Arnauld, but the events themselves, which testify to their
being miraculous. Regarding the miracles of the holy thorn, I have already quoted
122 Graeme Hunter
Arnauld as saying: 'Since these miracles occurred in the middle of worship

services and in such propitious circumstances, there was no one who failed to
recognize them as marks of God's protection of this community ... ' Likewise,
concerning the miracles described by Augustine, Arnauld was quoted above as
saying that every reasonable man would have to recognize them as miraculous.
If we now shift the focus away from the events and onto the testimony about
them, it is Amauld's account that is universal and regular, and Hume's which
admits of exceptions. To stipulate the exceptionless regularity of nature, as Hume
does, thus comes at the cost of introducing differential rules for testimony. In
other words, Hume cannot countenance any general theory for the evaluation of
testimony, but must evaluate it ad hoc, according to its subject. Arnauld inverts
that pattern. All testimony is evaluated by him according to its internal and
external circumstances. This admits the possibility of there being true testimony
regarding irregularities in nature, such as would lead any reasonable person to
conclude that a miracle had occurred.
The miracle of the holy thorn must have strengthened Amauld's conviction
that such testimony was possible. And, indeed, there have been few events in all
of recorded history which are attested to simultaneously by the best doctors
(pace Sainte-Beuve), 3 1 philosophers and poets of an era, and witnessed in
addition by a convent full of women of renowned probity. Should a rational
person with an antecedent belief in the existence of God not conclude that the
testimony is accurate and that it likely betokens a miraculous event? Only the
unsupported (and even on Humean assumptions unsupportable) 32 assertion that
there are no contingent events of the sort that Amauld describes could rule
this out a priori. Thus Amauld's assumption of the regularity of the rules of
testimony in the Logic is no more intrinsically controversial than Hume's
assumption of the regularity of the laws of nature in the Inquiry. And Amauld
has a distinct advantage in his account of nature. He need only maintain that
contingent events, as he defines them, are possible, whereas Hume must either
defend the stronger claim that they are impossible, one for which he has no
evidence, or else lose his a priori argument against miracles. Amauld's general
doctrine of miracles, though conceived in a limited apologetic context, thus
proves defensible, even against the assaults of the ablest of the later modem
sceptics.
Malebranche and Frequent Miracles
The miracle of the holy thorn thrust upon Arnauld the task of defending
miracles as local interventions of God, useful for the confirmation of doctrine.
The writing of the Port-Royal Logic gave him the opportunity of securing that
Amauld's Defence of Miracles 1 23
position within a larger framework of epistemology. But neither occasion

required Amauld to take a position as to the remaining aspect of what I have
been calling the Catholic status quo - the frequency of miracles. An opportunity
to do so was provided, however, in 1 684, with the publication by Nicolas
Malebranche of an expanded fourth edition of his Treatise on Nature and Grace.
A major novelty of that edition was an added fourth Eclaircissement with the
title 'The Frequent Miracles of the Old Law in no way mean that God often acts
by Particular Volitions. ' Malebranche ingeniously argues that the frequent bib
lical miracles can all be accommodated without supposing that God willed each
one separately . God could, instead, have made a general resolution to act on
every occasion as demanded by the intended interventions of his angels in hu
man affairs. In this way God would remain the sole real cause in the universe,
creatures would be only occasional causes, the world would be governed only
according to general laws. and yet frequent m iracles would still be possible. The
ingenuity of this arrangement lay in its potential for satisfying several seem
ingly irreconcilable parties. Modem metaphysics, as we have seen, had no place
for real secondary causes. This account of m iracles satisfied its demand that
God be the sole efficacious cause. Modem science's requirement that the uni
verse be exceptionlessly regular was also satisfied by Malebranche's position,
since it permitted God to act only according to general laws. Finally, a mini
mum demand of theological orthodoxy was that the miracles recounted in the
Bible be deemed to have happened, and Malebranche was able to satisfy that
too, yet without jeopardizing modem metaphysics and science. The key to this
elegant solution obviously lays in the proof that there could be frequent m iracles
without frequent particular volitions on the part of God.
Amauld lost no time in attacking it. In fact, he couldn't wait for his attack to
be pub l ished in its natural place within the two book- length studies of
Malebranche' s Treatise which were in preparation in 1 685, and so he brought it
out separately in that year under a windy title, which began: ' Dissertation of M.
Amauld, Doctor of the Sorbonne, on the manner in which God performed the
frequent m iracles of the Old Law by the Ministry of Angels ... ' In that work he
was able to produce no knockdown conceptual argument against Malebranche's
position, though he exhibited his wonted acumen in finding logical difficulties
with it. His principal contention, however, rem iniscent of Pascal, was that
Malebranche' s conception of divine concourse was a philosopher's pipe-dream
and quite at odds with anything to be found in scripture or in the traditions of
the church.
A . -R. Ndiaye has shown, in his study of Amauld's thought, that the latter's
doctrine of general providence entails that God acts by particular volitions. 3 3 It
is not surprising, therefore, that his doctrine of special providence, or miracles,
124 Graeme Hunter
should make the same claim. And if God acts by particular volitions, then the
dreams of the new science are not true, for the general laws it seeks will not be
a comprehensive guide to all that happens in the universe.
Though the attack on Malebranche is philosophically the weakest of Amauld's
writings on miracles, it is of a piece with the others and adds a theological and
biblical dimension desirable in dealing with a question of this nature. It is also
important in being the only writing of Amauld explicitly to defend the tradi
tional understanding of the frequency of miracles. ·
The three texts on miracles examined here were written over a period of a
quarter of a century and in different contexts. Yet they display sufficient conti
nuity to form a unified defence of what I have called the Catholic status quo. It
is characteristic of Amauld to have mounted such a defence at a time when
every aspect of the thing defended was under attack, and yet to carry it off with
such aplomb that it challenges the most formidable sceptics. Nevertheless, to
reason well is sometimes to do no more than spit against the wind. A defence of
the Catholic status quo, however well executed, was not what was called for in
the age of reason. Neither has subsequent history looked kindly upon it. If there
is a time for Amauld's theory of miracles, it must therefore be still to come.
Notes
See Benedicta Ward, Miracles and the Medieval Mind (Philadelphia: University of
Pennsylvania Press 1 982). 1 f.
2 St Thomas Aquinas, Summa Theologica, I q. 1 05. a. 8, resp.
3 Ibid., I, q. 1 05, a. 6, resp.
4 Ibid., I, q. 1 05, a. 7, resp.
5 Jean Calvin, Institutes of the Christian Religion, vol. 1 , J.T. NcNeill, ed.,
F.L. Battles, tr. (Philadelphia: Westminster Press 1 960), ' Prefatory Address. · 1 5 .
6 Martin Luther. cited in What Luther Says. vol. 2. E.M. Plass. ed. ( Saint Louis:
Concordia 1972), #3006. p. 957.
7 See for example Malebranche' s critique of secondary causes in Ec/aircissement 1 5
of the Recherche de la verite.
8 Robert McRae, ' Miracles and Laws. ' in The Natural Philosophy of Leibniz,
K. Okruhlik and J.R. Brown. eds. (Dordrecht: Reidel 1 985), 1 7 1-8 1 . See p. 1 76.
9 Descartes, AT 6, 45. It is possible, of course, that Descartes is speaking ironically
here and that his real meaning was that the evolutionary course was the true one.
10 Thomas, Summa. I. q. I 05, a. 7, ad 1 .
11 Quotes in Tetsuya Shiokawa, Pascal et !es miracles (Paris: Nizet 1 977). 79.
12 The editors of the Oeuvres of Antoine Arnau Id give some very convincing reasons
Amauld's Defence of Miracles 125
for thinking that the author of the article from which this excerpt is taken entitled
' Reponse a un ecrit . . . ' is indeed Amauld (See their introduction in OA 23, viii ff. )
Modem scholars however are not convinced (see Shiokawa; Pascal, 1 06). The
cited passage at any rate is found on p. 1 1 of the ' Reponse . . . '
13 The sources of the account given here are Arnau Id ( or Pseudo-Arnau Id)
C Reponse . . . ,' iii- I O); the historical and critical preface to the 'Reponse . . . , '
v i ; Racine, Abrege de 1 'histoire de Port-Royal (Paris: Editions d'aujourd 'hui
1 98 1 ), 8 1 -90; Pascal (see Shiokawa), and Shiokawa, Pascal, ch. 3 .
14 Amauld. · oe r Autorite des miracles. ' OA 2 3 , 33-86, esp. 3 5 .
15 Ibid .. 67 : ' Mais puisque ces miracles n e sont arrives qu'au milieu d e leurs [i .e., of
the religious] adorations & de leurs hommages, & dans les conjonctures si
etranges, qu ' i i ny a personne qui ne les ait pris pour des marques de la protection
de Dieu sur cette maison. '
16 All citations of this work are according to the critical edition of it entitled La
Logique ou 1 'art de penser, Pierre Clair and Fran-;ois Girbal, eds. (Paris: Presses
Universitaires de France 1 965).
17 Jean Racine, Abrege, cited in Amauld, Logic, 365.
18 Logic, IV, 1 , 29 1 f.
19 The problem here is analogous to that faced by the French translator in translating
different occurrences of 'you. ' To the English speaker they seem uni vocal; to the
French speaker some of them seem to mean 'tu, ' others ' vous. ' The word 'foi' may
appear to French speakers to be used univocally throughout the passages in
question, but the work demands different English translations in different contexts
and remains perplexingly ambiguous in some.
20 Logic, 292.
21 This is the translation given by the usually reliable translator Thomas Spencer
Baynes in The Port Royal Logic, 5th ed. (Edinburgh : Sytherland and Knox 1 86 1 ),
300.
22 Amauld, Logic, IV, 1 3, 339: ' J ' entens tout ceci selon leurs causes prochaines, ... '
23 Ibid., Premier Discours, 1 5 .
24 Ibid., I V, 1 4. 346: ' i i n 'y a point de personne raisonnable qui n 'y dotve reconnoitre
le doigt de Dieu. ' I assume that Arnauld would counter the claim that atheists are
reasonable people who would not recognize God 's finger in the events by saying
either that such events would convert a fully rational atheist or that there are no
fully rational atheists.
25 Ibid., IV, 1 4, 345 : 'Mais je soutiens que tout homme de bon sens, quand ii
n 'aura pas de piete, doit reconnoitre pour veritables les miracles que S . Augustin
raconte . . . '
26 Sainte-Beuve makes this point in general about the Jansenist use of this miracle,
in spite of disconfirming instances like the present one (see Port Royal, v. 3, ch.
126 Graeme Hunter
XII [Paris: La Connaissance 1 926], vol. 3 , ch. 1 2, p. 52).

27 Hume, Enquiry § I 0, part I (Indianapolis: Library of Liberal Arts 1 976), 1 23 .
2 8 Ibid., 1 22.
29 I am aware that weaker interpretations of Hume' s meaning in this chapter are
possible. I do not bel ieve them more conformable to Hume' s text, however, and of
course they would be of less use in the present context, where Hume is to act as a
foil for Arnauld.
30 Hume, Enquiry § 1 0, part II, 1 3 7.
3 1 Sainte-Beuve regards the miracle as a ' fait nature) ' exploited by the Jansenists in
order to save themselves from extinction (Port Royal, vol. 3, ch. xii, p. 52).
However, the case he makes for this is so slight that it is better seen as an
expression of his own struggles with orthodox Catholicism.
32 Within the scope of Hume' s general sceptical treatment of causality a denial of the
possibility of miracles is ruled out, because it allows for exceptions to the causal
order. But, even within the chapter on miracles in the Inquiry, Hume seems only to
want to prove their a priori physical impossibility. It is as if Hume wanted to say
that miracle stories contain synthetic a priori falsehoods.
3 3 A.-R. Ndiaye, La Philosophie d 'Antoine Arnauld (Paris: Vrin 1 99 1 ), 259.
9
Arnauld versus Nicole:

A Medieval Dispute
J E A N- L U C SO L E R E
Antoine Arnauld was known as a fierce controversialist. At the end of his life,
he fought, not only with his lifelong opponents, but even with his friends. Hence,
we can read the report of his controversy with Pierre Nicole, Gommaire Huygens,
and Fran�ois Lamy. The first, Arnauld' s longtime companion, needs no intro
duction. The other two were also sympathizers with what has come to be known
as Jansenism; they belong, at least, to the Augustinian sphere of influence whose
champion was the Master of Port-Royal.
Thus we will observe the persecuted quarrelling among themselves. Their
inflexible theological theses will be found to rest on rather unexpected philo
sophical positions; we will find to our surprise that Arnauld has, at this stage,
abandoned St Augustine in favour of St Thomas Aquinas. It will emerge, then,
that whatever their enemies may have said, the 'Jansenists' did not form a doc
trinally homogeneous group or 'party. ' This may help us revise our historio
graphical categories, which we too often draw from the polemical vocabulary of
the times, the summary labels proposed by one or another litigant, thus compro
mising our historical neutrality.
Nicole and General Grace
The dispute originated in Pierre Nicole's theory of general grace. Around

1688-9, in his XIIlth Theological Instruction, it became glaringly obvious that
this eternal second to Arnauld had his own views on the matter. 1 Nicole is con
cerned to explain the following: Every descendent of Adam shares in the fall of
human nature, yet any one of them can do good and save himself, though he
1 28 Jean-Luc Solere
will never actually want to do so unless he receives a special efficacious grace.

No one who lacked a sufficient capacity to do good could be considered guilty of
wrong doing. But the simple physical power to choose recognized by Jansenius
and Arnauld is not enough; in the absence of the knowledge of good and evil,
the simple power seems to Nicole ' like a healthy eye which is deprived of light,
like healthy legs which are well tied up. ' 2
Thus Nicole proposes the hypothesis of a (non-Molinist) supernatural assist
ance sine qua non, given to all, a light from God shining on every person,
whether Christian or not (even if he does not actually make use of it). This aid is
a general grace, a reception of eternal truth in all minds: 'It consists in truths we
do not have to learn from men' 3 because they are communicated directly to us by
the Creator.
Is it likely that every person is enlightened in this way, even if he is not
conscious of it and does not recognize the source of the enlightenment? So it
would be for the vast number of non-Christians, and indeed for many Chris
tians. Arnauld raises this issue in lemma 5 of his Ecrit geometrique sur la grace
generale. He objects that ' a soul has not been enlightened with regard to an
object when it has not known it and has not a single thought of that object . . . ' In
other words, for Arnauld, this ' general grace' is but a name: It is supposed to
consist in knowledge bequeathed supernaturally and universally, whereas m any
minds have evidently never received it, since they have no notion or knowledge
of true good and true evil. To know something is to think about it and to be
aware of thinking about it.
Nicole admits that this is the main point, and that the doctrine of general
grace assumes that a soul is never in the state of invincible ignorance with
regard to moral truth or with regard to natural law. (He cites the teaching of the
theologians of Louvain.)4
In reply to Arnau ld's objection, he proposes a distinction between two types
of knowledge: direct knowledge of an object or truth ' by means of an exact and
distinct idea, which will make us know it distinctly, expressly and immedi
ately, ' 5 and indirect or shrouded knowledge, by dint of a general principle in
which the truth is contained really but confusedly and implicitly. In the latter
case, we can know the truth expressly by deriving it from the principle commu
nicated to us by God, and we are prevented from doing so only by our passions
and vices, which divert our attention from the principle, which is none the less
always present in our mind. 6
The question, then, is whether the impl ic it, confused, even unnot iced,
apprehension of a truth in its principle is indeed knowledge. Nicole is not in
doubt: ' We conceive many truths without tying them to words, without noticing
that we know them, and many people who have never heard the principle that
the whole is greater than its part have judged that a part is smaller than its
Amauld versus Nicole 129
whole ... This is not done by means of distinct propositions, but by intellectual
perceptions that yield the same conviction. ' 7
Nicole identifies this kind o f unformulated or implicit knowledge with
the divine illumination taught by St Augustine. 8 The assumed ' intellectual per
ception' is a perception of God himself, or rather of the eternal ideas in God
(equality, order, justice, piety ... ). It is not the result of effort or research, which
obviously could not be carried out by everyone; on the contrary, the ideas are
unveiled by God, qua veritas docens et illuminans. Thus, ' God as truth, as
justice, as light, shines forth in the spirit of infidels and of ungodly people. ' 9
Again, ' since this illumination ordinarily takes place without words, the mind
does not perceive it distinctly, and so we should pay no attention to the account
of those who say that they have never had thoughts and illuminations of that
kind, for they might have had them without knowing it. ' 1 0
In order to describe these notions, unnoticed yet given by God to every mind,
Nicole proposes the category of ' imperceptible thoughts, ' which is, thus, the
keystone in the system of general grace. True knowledge and true thoughts dif
fer from clear and distinct ideas in that they are not dressed in language and
spread out in consciousness; these ' imperceptible thoughts' are in fact knowl
edge by feeling: ' that is to say, delicate thoughts, prompt, confused, indistinct,
and afterwards forgotten. ' 1 1 Nicole refers explicitly to Pascal's theses on the
grasp of principles: many things are known by feeling alone, 'by which is meant
that we do not have the fully developed idea of them ready to hand. ' 1 2
G. Huygens, Divine Illumination and Its Thomist Interpretation
In his discussion of lemma 5 in Amauld, Nicole refers, we saw, to the teaching

of the theologians of Louvain. More particularly, he refers to a thesis advanced
by one of their number, Gommaire Huygens: de veritate aeterna, sapientia et
justitia aeterna. 13 Wishing to undermine such a foundation for the doctrine of
general grace, Amauld went directly after Huygens' s thesis. In this way he at
tacked one of the chief supporters of Jansenistic Augustin ism in Louvain, one of
his own allies, with whom he was well acquainted. 1 4 Worse still: Huygens was
only repeating Jansenius' s ideas, contained in the De statu naturae purae (I, c.
7-8). It is therefore against Jansenius that Amauld waged war, and behind him,
against St Augustine himself.
In that thesis, Huygens had held, first, that it is in uncreated truth, namely,
God, that we see all necessary, immutable truths, and, second, that when we
love a virtue for itself, we in fact love its archetype in God, so that it is God
himself that we love. Hence, Amauld attacked him in a Dissertatio bipartita,
intended to combat those two assertions one after the other. 1 5
The second assertion i s a consequence o f the first, and for Arnauld the
130 Jean- Luc Solere
more embarrassing theologically, for, like Nicole's general grace, it entails

recognition of the virtue of pagans. 1 6 From his earliest works, ArnauId had taken
a clear stand on that issue, one inspired by Augustine (or at least by certain of
his writings): pagans cannot practise any genuine virtue, for they do not fulfil
the condition for all virtue, which is to love God as the only end in itself; even in
their heroic, outwardly disinterested, actions, they remain subject to the sin of
pride, to self-love (love of their own glory, of their soul's beauty), for they lack
the power to overcome concupiscence through faith and grace. If, by contrast,
one supposes with Huygens that the pagans' love of virtue is an implicit love of
God because it is directed, without their knowing it, to the idea of virtue which
is in God and which is God, there would be no denying that their conduct was
morally good and, finally, that they could be saved without faith in Jesus Christ
and without grace.
The idea that grace might not be absolutely necessary is intolerable to Amauld.
He tries to undermine that claim by contesting the first part of Huygens's thesis,
namely, the view that every person discovers in God, or in God's light, the
necessary truths he perceives. Thus Amauld, paradoxically, finds it necessary,
in order to maintain the strictest Augustinism in moral theology, to break with
another, no less characteristic, Augustinian thesis on the metaphysical level:
the thesis of illumination.
To that end, and here is the surprise, he uses Thomas Aquinas against Augus
tine: 'Truth is principally in the intellect, and secondarily in things.' 1 7 Truths
are so called only with reference to the intellect upon which they depend: to the
divine intellect in some cases (one speaks of 'true gold' if the metal concerned is
in keeping with its archetype in God, in accordance with which God created it),
to the human intellect in others. Now, all true propositions are of the latter sort,
even when they state necessary and immutable truths: 1 8 'The truth of the propo
sition is not other than the truth of the intellect.' 19 Propositional truth (which is
truth proper, 'secundum propriam rationem '), Amauld comments, is not a thing.
It is nothing other than a relatio, 'a simple relation.' between subject and at
tribute in the judgment of the intellect, in so far as it is consonant with the thing
it judges. It should therefore not be sought elsewhere than in the intellect. 20
Hence, the truths perceived by different intellects are not reducible to one sin
gle, selfsame, subsisting truth, God himself: 'There are as many truths, properly
spe·aking, as there are judgments formed about things whether by diverse minds
or by one and the same mind. '2 1
This is a curious revival of a quarrel four centuries old, in which St Thomas
undertook to reduce Augustine' s illumination to the light of the agent intellect
proper to each person: 22 'For there is in each and every man a certain principle
of science, namely the light of the agent intellect, through which certain univer
sal principles of all the sciences are known at once, in principle, naturally.'2 3
Arnauld versus Nicole 131
Only qua cause of our cognitive ability, and not qua first known object, is God
the source of all our knowledge: Instead of giving himself in an illumination, he
gives us our inner light, our knowing/acuity: 'By God we are given the interior
light of the intellect, which is the principal cause of knowledge [scientiae]. '24
Having conceded that man is endowed with his own knowing power, we must
apply the principle of economy and eliminate every superfluous hypothesis, such
as the hypothesis that we need direct divine illumination in order to acquire
natural knowledge: ' It is useless to have recourse to eternal truth above our
mind if we find in our mind itself all that is necessary in order that we should
judge true those things that are demonstrated with necessity in the sciences. '2 5
Now, if it looks into itself, the mind finds in itself perceptions or ideas, the
faculty to join these together and to separate them Uudgment), and finally the
faculty to link two judgments by means of a third (reasoning). 'Assuming only
these things, we easily understand how the human mind can acquire the demon
strative sciences, such as geometry and arithmetic. '26 For they proceed through
definitions and demonstrations. Definitions merely awake in our mind the ideas
of the tenns, which it grasps by simple apprehension. 27 Demonstrations merely
lay down relations among those ideas. That is enough for the discovery of any
truth, even it is necessary and immutable: 'But here there is only created truth,
which is in my mind: the ideas are in my mind, their connection is a work of the
mind, as is the assent by which it adheres to that connection. '2 8 Every intellect
finds in itself, not its truth, but the truth; agreement among minds on the true
does not result from contemplation of a common object, but from parallel proc
esses. Accordingly, Arnauld rejects the classical Augustinian argument: 'If both
of us see that what you say is true and we both see that what I say is true, where,
I ask, do we see it? Certainly I do not see it in you or you in me, but both of us in
the unchangeable truth that is above our mind.' To this Arnauld replies: 'That I
do not see it in you or you in me, I agree . . . For each of us sees it in his own
mind: I in mine, you in yours. '29
One may with reason conclude that in God there exist eternal reasons for
things, their archetypes, considering that God is the cause of all things and that
things have not been fonned at random. But it does not follow that we have
knowledge of the archetypes themselves: they cannot be perceived by us in this
life, but only by the blessed who see God face to face.
Fran�ois Lamy and the Defence of Augustinism
We are told that, when Nicole read the Dissertatio bipartita, he 'had the sincer
ity to acknowledge that he could not see what could be said in reply.'30 He showed
Amauld's work to Father Frarn;ois Lamy (the Benedictine). The latter, both an
Augustinian and a Cartesian, was a close friend of the renowned members of
that cultural milieu: Nicole, of course; Amauld (whose friendship antedated his
departure into exile); Bossuet; and Malebranche. But that did not deter him
from entering into lively disagreements with them. 3 1
The controversy would thus spread to include a person closely connected to
the previous protagonists. In fact, Lamy hastened to Nicole's rescue. Not that he
adhered to his views: quite the contrary, for after Nicole provided him with the
manuscript of the Traite de la grace generale, he wrote Refutation du systeme
de la grace generale! 3 2 However, he felt it his duty to defend, against Arnauld,
both the Augustinian doctrine of illumination and the consequences drawn from
it by Huygens concerning the love of God. Therefore, he wrote an answer to the
Dissertatio bipartita. It remained unpublished and seems today to have been
lost. 3 3
We are in possession, though, of Amauld's reply to Lamy. In addition to an
answer on the heart of the matter, it contains a reflection on the procedures it
should be adopted in that type of discussion, so as to put an end to misunder
standings, sophisms, and endless disputes. This is the Reg/es du hon sens, pour
bien juger des ecrits polemiques dans /es matieres de science, appliquees a une
dispute entre deux theologiens touchant cette question metaphysique: Si nous
ne pouvons voir /es verites necessaires et immuables, que dans la verite
souveraine et incree. 34
Arnauld and the Rules of Common Sense
Joining theory to the practice of controversy, the Rules state fifteen general pre
cepts, each illustrated by discussion of Lamy's objections to the Dissertatio. Or,
to put it more accurately, Amauld picked out certain objections and prefaced his
discussion of them with ad hoc methodological considerations. Once again the
work is divided into two main parts, reflecting the two parts of Huygens' s the
sis: There is a purely metaphysical question, that of the knowledge of truth, and
a moral question that depends on it. We shall inspect a few of the articles.
ART I: Diverse kinds of disputes
FIRST RULE: 'Take good heed of the nature of the question being disputed: is it
philosophical or theological? For if it be theological, it must be decided chiefly
by recourse to authority; whereas if it be philosophical, it must be decided chiefly
by recourse to reason ... ·
The first reproach Lamy levels at Amauld is that he abandoned St Augustine 'to
follow St. Thomas, thus preferring the opinion of the Disciple to that of the
Arnauld versus Nicole 133
Master' (p. 154). But it is the very spirit of Augustinism that allows Arnauld to
move away from its letter; while opposing Augustine on a particular point, he
remains faithful to a deeper Augustinian principle: 'quod credimus, debemus
auctoritati, quod scimus rationi.' On the matter here in dispute, nothing is re
vealed by God or decided by the church; hence, the light of reason alone can and
must show what side to take. Augustinism itself is not an indivisible block, to
take or leave. One must distinguish within it what belongs to the exposition of
Catholic doctrine, and what belongs to philosophical speculation.
Besides, Arnauld notes, 'should one wish to stop at authority,' that of St
Thomas is far from being 'as contemptible as our Friend would make us be
lieve,' for 'he was a very great mind, and held St. Augustine in most special
esteem and respect' (p. 154). Scholasticism is often believed to be dead and
gone in the seventeenth century. We can see, on the contrary, that Thomas's
teaching lives on and plays a role in philosophical reflection. Indeed, we see it
defended and illustrated by someone who is not outmoded, but has, rather, fought
repeatedly in favour of Cartesianism.3 5
ART II: The state of the question
SECOND R ULE: 'To consider carefully whether the state of the question has
been well put, and to take care that it is not subsequently changed by passing
insensibly from the point at issue to another point not at issue.'
Arnauld acknowledges that the state of the question has been well formulated
by Lamy: ' Whether all necessary and immutable truths are seen in a sovereign
and uncreated truth' (p. 156). But we need to consider what is and what is not
contained in that formulation, and to ' sort out the equivocation which might lie
in the words: "to see immutable truths in uncreated truth",' for this phrase may
be taken in a more or a less proper sense. Now Lamy, according to Arnauld,
kept changing the state of the question and passing from one problem to an
other, for want of a full elucidation of the content of that phrase.
Arnauld starts out with an important remark: He sets the present discussion
apart from his famous controversy with Malebranche, which also had to do with
'vision in God.' Despite the apparent similarity, this is not in fact the same
problem, or the same vision. For Malebranche, what we see in God are not
truths, which, for him, are mere relations, but rather the creatures themselves.
By contrast, Lamy expressly denies that ' those works of God ... are seen in God;
that is to say in the ideas according to which those bodies were formed'; for him,
only (necessary and immutable) truths are seen in God. His Augustinism is far
more orthodox than Malebranche's (even though he was known as a disciple of
134 Jean-Luc Solere
Malebranche). His case seemed less extreme to Amauld, who saw nothing but
wild imaginings in the Oratorian's metaphysics.
Nevertheless, the root of the error is the same. It is the confusion provoked by
the attractive fonnula of the Augustinians: to see (truth or creatures) in God.
Beneath the apparent clarity of the expression there are fonnidable ambiguities.
Amauld will dissect these with great skill.
Applying a principle of his own logic (to have a clear and distinct idea of the
meaning of the words one uses), Amauld asks for the general meaning of such
a fonnula. 'When we say, especially with regard to spiritual insight, that we see
one thing in another, B in A, the proper, natural sense of those words is, that
knowledge of the one gives us knowledge of the other' (p. 158). For instance, in
the case of the 'demonstrative sciences,' theorems are known 'in their princi
ples,' i.e., are known by means of axioms, which are known by themselves. 'It
follows that A must be known, and better known than B, if we are to speak
properly of knowing B in A, tamquam in objecto cognito' (p. 158). 3 6 A is the
'efficient cause of the knowledge of B,' or its 'condition sine qua non.'
On the other hand, to say that we see invisible corpuscules with the naked eye
in a microscope (an allusion to a recent invention) would be to speak improp
erly: we see them through the microscope, for we do not see the microscope
itself or even its eyepiece, but only what it shows us. Likewise, we cannot say
that we see objects in the sun, but rather, thanks to the sun, light coming from it
enables us to see them; it is a cause of our seeing, but we do not necessarily see
it. We should speak of seeing objects in the sun only with the qualification
" causaliter, non vero objective, seu tamquam in objecto viso' (p. 159). 37
This being understood, let us now ask those who hold that we see certain
truths in God: ' Do you mean that we see them in God, tamquam in objecto
cognito, or only causaliter, and because God is cause of the fact that we see
them?' (p. 159).
In the second case, they would be in agreement with St Thomas (and Amauld),
who accepts the Augustinian thesis of a vision of certain truths in God if it is
taken to mean that 'God is the efficient cause of the knowledge we have of those
truths, since the natural light of our mind, thanks to which we know them, is a
participation of uncreated light' (p. 159).
But, as Amauld points out, that is not in fact their thesis. Malebranche, in the
quarrel 'on true and false ideas,' held that modifications in our minds are not of
an essentially representative nature, i.e., are not really and fonnally our light,
which implies that God is, not only the cause of our light, but fonnally our light.
For his part, Lamy specifies that if we do not see contingent truths in God, but
only immutable and necessary truths, it must be because we see his essence, not
his decrees; in other words, he holds that we see the divine essence tamquam
objectum cognitum.
ART Ill: St Thomas 's definitions
Lamy appears to have challenged the definitions of truth borrowed by Amauld

from St Thomas, on the grounds that they are 'all founded on purely arbitrary
definitions, and on notions forged recklessly, offering to the mind no more than
a pure scholastic jargon that has nothing solid in it' (p. 16 1 ). A special target is
the idea that ' truth proper is only to be found in [human] understanding. ' ' Why, '
Lamy asks, ' can we not say, with just as much or more reason, that truth proper
is to be found properly and chiefly in things only; that it consists in their rela
tion to the originals, the eternal ideas of them, and that in all the rest it is only
improperly to be found; and that our judgments are thus true solely in so far as
they express the relations of things to their ideas and their originals?' (p. 162).
But that, says Amauld, is transcendental truth, which lies in the correspond
ence between things and their model, namely, the divine idea for natural things,
and the craftsman's idea or the copied model for artificial things (as when we
speak of the true portrait of a man). St Thomas speaks of that truth also, but for
him it is not truth proper. Truth proper is found primarily in our judgments
(p. 162). The question then is whether the notion of truth primarily means truth
in the latter, predicative, sense, or in the transcendental sense. Is truth primarily
logical or ontological?
Lamy appears to have reasoned as follows: Since correspondence between
things and divine ideas is far more noble than correspondence between our
judgments and their objects, we have better reason to say that truth proper is
that which is in things (p. 163). But the rule here invoked is wrong. The greater
or lesser propriety of the use of a term hardly depends on the greater or lesser
dignity of the subject it is applied to: Scripture speaks of the wrath of God or of
his arm, but God's superiority to men does not entail that those expressions are
said with greater propriety of God than of men - quite the contrary.
One should therefore return to a simple thesis regarding the origin and nature
of language: Words are conventions for the expression of ideas, and the exact
sense of words, especially in real definitions, is fixed by common usage, unless
otherwise indicated. 3 8 'Now one of the first things that men need to convey is
whether they or someone else are mistaken in their judgments of the things they
talked about' (p. 163).
The words for truth and falsehood in all languages first signified quite sim
ply that we are or not mistaken, i.e., that our judgments correspond or not to
things. And this original meaning remains the chief and immediate meaning.
The real definition proposed by St Thomas rests on th is fact about ordinary
usage and is therefore well grounded in common sense.
Of course those judgments cannot be correct unless they also correspond to
divine ideas, to which correspond in tum the things whose ideas they are. But it
136 Jean-Luc Solere
is not necessary to know these ideas in order to know things. Or again, arriving
at truth consists, not in knowing the relation of things to their archetypes, but
rather in judging of their nature 'through experience and conjecture' (p. 164).
We are, says Amauld, in the situation of someone who wants to unravel the true
meaning of a coded message not intended for him. He does not have the key to
the code, but he has guessed it 'by means of certain rules, found out by those
whose profession it is to unravel letters of that kind' (p. 165). The world is like
a cryptic letter, but we do not understand it from . its archetypes in God, which
would be to know its code in advance. We can only trace the scrambled message
back to its code, or rather assume that we are deciphering according to the code
when we obtain coherence and sense, like Descartes who 'having laid down
very simple and very evident principles, ... very clearly deduced from them a
great number of the most beautiful phenomena of nature' (p. 165).3 9
ART V: Imperceptible thoughts
We now arrive at the crucial moment of the dispute, at which the plausibility of
direct illumination of minds by God is now at stake.
We know that Nicole is the one who put forward the postulate of impercepti
ble thoughts to prop up his system of general grace. The same postulate is ap
pealed to by Lamy to make it possible for us in fact to see every truth in God,
even if we are not conscious of doing so. For, in order to make up for the lack of
evidence for that thesis (since, as with Malebranche, most men have the vision
in God unwittingly), we are forced to suppose that when we know geometrical
truths, we are thinking of uncreated truth without being aware that we are thinking
of it.
That hypothesis is purely and simply incredible, says Arnauld. How could
anyone accept the claim that, having seen truths such as two and three make
five ' a thousand and a thousand times' I have had uncreated truth present to my
mind a thousand and a thousand times 'without my ever noticing that I was
thinking of it' (p. 171)?
Arnauld thus uses the question of imperceptible thoughts to test of the
Augustinian theory of illumination. The latter must be rejected if it can be held
only at the expense of conceding an unacceptable hypothesis: in this case, the
existence of imperceptible thoughts, i.e., of thought contents which, although
they take place in thought itself (understood as substance), are not perceived
by it.
That postulate is not acceptable, for it goes against other well-established
principles, whose suppression would ruin Cartesian gains in philosophy and
the sciences. Indeed, the very cornerstone of the new science would thus be
attacked, namely, the transparency of thought to itself, the possibility of evident

knowledge of oneself as a thinking substance. The Cartesian soul, once it de
fined as res cogitans, is freed from the obscurity involved in the assumption that
it animates the body. It is this possibility of knowing oneself completely which,
after the cogito and the end of hyperbolic doubt, authorizes the mind's self
certainty as pure thought (it is not in myself that I see the notions that make me
think I am a body), and authorizes the second proof of God through effects
(there is not within me an unconscious power to give myself being). It also
enables us to proceed through clear and distinct ideas. 40 The metaphysics of
Cartesianism has as much horror of imperceptible thoughts as its physics has of
occult qualities. 4 1 Thought is pure consciousness, and cannot admit to its fold
any foreign element, anything opaque to it, which it could not perceive within
itself: in short, nothing unconscious within consciousness itself; that would be a
contradiction in terms. This important point, among others, set the seal on
Leibniz's break with the Cartesians (in his theory of small perceptions) as well
Malebranche's (in his thesis that the soul is obscure to itself).
By contrast, Amauld relies on the Cartesian principle of the total transpar
ency of thought to itself. 42 According to this principle, I am certain, thanks to
the witness of my conscience, that I have seen many necessary and immutable
truths of geometry and arithmetic with the the eyes of my mind, even though I
did not have the slightest thought of a sovereign and uncreated truth. Now, ' to
say that I think of a thing is the same as to say that the thing is present in my
mind' (p. 170). 43 To entertain a thought is to have it present in the mind, so that
it, like every other content of consciousness, is ipsofacto an object of conscious
ness. When I think of something, I know at the same time that I think of it. Now
I can have knowledge of truths without thinking of uncreated Truth. Therefore,
I can have that knowledge without having uncreated Truth present in my mind,
for if I did consider it in any way I would be conscious of seeing it.
Amauld emphasizes that his critique is concerned only with entirely imper
ceptible thoughts: 'Certainly some thoughts are less perceptible than others,
and we are far more aware of clear and distinct thoughts than of thoughts which
are obscure and confused, more aware of thoughts on which we reflect deliber
ately than of those on which we reflect only virtually, a sort of reflection that
must be judged inseparable from thought' (p. 172). But an unperceived or fugi
tive thought is not for that reason imperceptible.44 Virtual reflection, i.e., the
capacity to make thought an object, is inseparable from thought because it con
stitutes the very nature of consciousness. What makes us say that no natural or
artificial body, however well organized, is endowed with intelligence, is the fact
that no such machine is aware of what it does: non est conscia suae operationis
(p. 173). 4 5
Amauld seems to weaken his position, however, when he gives the following
argument: Everyone allows that it is really possible for a person to be of good
faith, to know and to state the very core of one's thought. 'But how could a
man swear that he has not had the slightest thought of this or that thing, with
out risking perjury, since he could have had imperceptible ones' (p. 173)?
Psychologists would doubtless fault Amauld here. There appears to be a misun
derstanding between Amauld, at ease with epistemological problems and logi
cal dilemmas, and his two adversaries, Nicole, author of the famous Essais de
morale, and Lamy, author of a treatise De la connaissance de soi-meme, who
are accustomed to probing the innermost recesses of the heart and the subtleties
of feeling, of muted and clandestine thoughts. 46 Are they really talking about the
same thing? As a matter of fact, Nicole invokes 'the example of certain secret
feelings which we have in the heart and which the soul does not discern; for this
reason its own depths are unknown to it and what swims on the surface of
thought is quite often different from what dominates the heart, as St. Gregory
the Great says. '47 Does not Amauld suffer a setback on this terrain?
Yet Nicole adds: 'Those secret feelings are always accompanied by secret
knowledge because those feelings are loves, and there is no love without knowl
edge. '48 It would seem, then, that he is the one in a weak position, since he goes
back to the level of knowledge, of intellectualism, of thought, and therefore of
consciousness. If secret feelings belong to the domain of knowledge, they might
be unperceived, but they cannot be imperceptible. Besides, however profoundly
self-love might be hidden under apparent virtues, La Rochefoucauld would not
have written his Maximes had he not succeeded in tracking it down, i.e., in
making it pass into conscience. Amauld could then send the psychologist back
to the preceding paragraph.
Finally, Arnauld reminds us that 'we must acknowledge the presence of
mechanical acts in man, which take place without thought, and which cannot
therefore serve as proof that there are imperceptible thoughts.' If Lamy is a
Cartesian, he believes that beasts do not think and that they do an infinity of
actions 'solely by the disposition of their machine. � He must admit, Amauld
concludes, that the same holds for many human actions, which assume no thought,
not even imperceptible thought, such as chewing or walking, or even writing.
The mechanicistic physics of Descartes renders useless the hypothesis of a psy
chological unconscious. To suppose an imperceptible thought guiding thought
less action is to appeal to an occult quality, a hypothesis that is superfluous,
given the capacities and the autonomy of the corporeal machine.
Having proved that imperceptible thoughts have no reality whatsoever, Amauld
goes on to show that, even if they existed, they would be of no use to a system of
general grace. An imperceptible thought would be an ineffectual thought, since,
not being conscious, it could not give me knowledge of any sort. 'Is there a
difference between saying nothing at all to me about my danger of falling into

an ambush, and warning me in a language I do not understand' (p. 182)? Hence
' imperceptible thoughts' cannot be used to accuse men before God.
ART XIII. The Iroquois ' conception of God as justice
The second part of the work is aimed at the moral corollary of Huygens's thesis:
that one cannot sincerely love a virtue without loving that virtue's eternal form
and reason, which is in God, so that to love a virtue is to love God. As we have
seen, this assertion seemed to Amauld even less tenable than what was said
about truth, because it had the consequence that even pagans are capable of
loving God.
Of course, Amauld acknowledges, one may say that God is justice itself. But
one may not draw the consequence that anyone who has some notion of justice
has the notion of God. Amauld cites the case of the Iroquois. Although they
knew full well whether a contract was just or unjust, they had, according to the
missionaries, no idea of any divinity whatsoever, whether true or false� indeed,
they did not even possess a word for a divinity in their language (p. 239). 49 If it
is replied that they may have known God without being aware of it, one could
just as well say that they may have been geometers or astronomers without
being aware of it (p. 240).
ART XIV: On the love of justice
This article gives us the rare spectacle of Amauld attacking Jansenius. Augus
tine, the doctor of Port-Royal reminds us, often says that one must do good for
love ofjustice, and not for fear of punishment. What is this justice one must love
while doing good? Jansenius has successfully shown against Vasquez that it is
not merely the justice of the action itself, but God himself. ' But when he wants
to explain how that justice is God, he resorts to Platonic thoughts that Augus
tine did not use when giving his excellent explanation of the nature of the genu
ine virtues' (p. 241 ).
Indeed, in his Contra Julianum (III, 4 ), Augustine shows that we must distin
guish, within virtue, between the duty and the end, between what must be done
and the motive for doing it ('Officium est quod faciendum est; finis, propter
quod faciendum est'). Pagans may have done good works secundum ofjicium;
but those works, though good, were badly done, because they had no knowlege
of God and so failed to link them to their genuine end. The first principle of
Christian morality is that, since God alone is our sovereign good, he must also
be the ultimate end of all our actions (p. 242-3 ).
That demonstration is sufficient. Hence, Arnau Id confesses, Jansenius' s
140 Jean-Luc Solt�re
additions 'are painful to me.' Indeed, in the De statu naturae purae (I, 8),
Jansenius claims that we see the eternal rectitude, the immutable and eternal
rule of justice which radiates within us and, thanks to which we form our judg
ments, not only through faith, but also through an intuitive insight (p. 243). But
if that is true, the love of justice as Jansenius understands must be found
'in all men, including infidels and the impious, because that primitive and eter
nal form of justice is exposed to the spiritual insight of all men.' Now Jansenius
himself proves in the De statu naturae purae (I, 5) 'that such a love of justice
can only be had by a true and supernatural grace of God, which is not common
to all men, and which is certainly not to be found in infidels or in the impious'
(p. 245).
Conclusion
What lessons should be derived from the controversy we have seen unfold?
First, it is clear that internal tensions existed in the 'Jansenist' milieu, whose
fundamental components, Augustinism and Cartesianism, allow for many indi
vidual variations. Though they fought together for years against common en
emies, Jesuit or Protestant, Nicole and Arnauld are far from having identical
convictions.
Next, let us pass in review the essential traits of the thought of Antoine Arnauld,
chief among the protagonists. On the theological level, he holds strict, Augus
tinian positions on the separation between nature and grace ( on account of original
sin there is no room for pagan virtue), predestination (both universal actual
grace, i.e., grace moving the intellect and the will, and 'sufficient grace' are
rejected), and human guilt.
On the philosophical level, whose autonomy he upholds, Arnauld puts con
siderable distance between himself and the Platonism of Augustine and Jansenius,
which he thinks he can dissociate from their theological positions. Depending
heavily on Thomas Aquinas, whom he studied very closely, he pleads for the
functional independence of the human intellect (it is the human individual who
thinks), i.e., for the individual's capacity to produce ideas (instead of receiving
or discovering them). Accordingly, he asserts its limitation: We can know God
and the archetypes only indirectly. We have no vision of God or in God at all.
At the same time, Arnauld comes out as a defender of pure Cartesian doc
trine. Some of these positions are in agreement with Thomistic ones, such as the
individuality and independence of thought. Others are new, such as the identifi
cation of thought with consciousness and the transparency of the soul to itself,
but do not seem to Arnauld irreconcilable with Thomism. In his eyes they are
decisive, complementary gains. Hence it is that, in spite of his adherence to new
Arnau Id versus Nicole 141
ideas, Amauld was able to keep Thomistic theses alive and functioning in the
midst of the seventeenth century. 50 This observation gives us reason to revise the
oversimplified view of that century as divided between innovators, on the one
hand, and diehards, on the other: There existed intermediate positions.
Notes
I wish to express my warm thanks to Professor Thomas De Koninck of Laval

University, Quebec, and Professor Elmar Kremer, for their contribution to the
translation of this paper. Extant inaccuracies are mine.
Cf. OA, the ' Prefaces historiques et critiques' to vol. 10 (xix ff.), and to vols. 19,
38, and 39; the preface to Nicole's Traite de la grace generate ( 1715). See also
J. Laporte, La Doctrine de Port-Royal, vol. 2: La Doctrine de la grace chez
Arnauld (Paris: Presses Universitaires de France 1922), 2 14-23; G. Lewis, Le
Probleme de / 'inconscient et le Cartesianisme (Paris: Presses Universitaires de
France 1950), 200---18, and A.-R. Ndiaye, La Philosophie d 'Antoine Arnauld
(Paris: Vrin 199 1), 344 ff.
2 Traite de la grace generale I, 50 1-2.
3 Ibid., I, 2 10.
4 Ibid., L 83-4.
5 Ibid.
6 This is what Nicole calls 'to know a truth in its principle' (ibid., 83).
7 Ibid., I, 87.
8 Cf. Ennar. in Ps., 6 1, 2 1: ' Ubi, inquam, vides hoc justum, quo viso reprehendis
injustum? Unde illud nescio quid, quod aspergitur anima tua ex multis partibus in
calignine constituta, nescio quid hod quod coruscat menti tuae ... vade illuc ubi
semel locutus est Deus, et ibi invenies fontem justitiae . . . ' Also, De Trin., XIV, 15 :
' Ubi sunt istae regulae scriptae, ubi quid sit justum et injustum agnoscit, ubi ergo
scriptae sunt in libro lucis quae Veritas dicitur, unde omnis lex describitur? ' Many
other passages are quoted by Nicole in II, 7 1 ff. He also finds this opinion in other
Fathers of the church, and assimilates it to the 'sufficient grace' of modern
Thomists.
9 Traite de la grace generate, I, 9 1.
IO Ibid., II, 85. 'Those reasons are known; but few know that they know them. ' (I,
88). Arnauld will contest this very claim: that one can know without being aware
that one knows.
11 Ibid. , I, 463-4. Conversely, 'what is called feeling is only a lesser perceptibility. '
12 Ibid. Concerning the importance of the notion of feeling in seventeenth-century
Augustinian thought, see Ch. Chesneau (alias Julien-Eymard d'Angers), Le P. Yves

de Paris et son temps, vol. II (Paris: Societe d'histoire ecclesiastique de la France
I 946), 33 ff.
13 Traite de la grace generate, I, 88.
14 The Dictionnaire de Theologie catholique (Paris: Librairie Letouzey et Ane 1922),
vol 7, col. 35 1, describes him as a friend of Arnauld and Quesnel. He was born in
163 1 and died in 1702. As a professor of philosophy and theology at Adrian VI
college in Louvain, his scholastic career (in particular his admission to the 'strict
faculty of theology,' in which resided the controlling authority in the university)
was compromised by his Jansenist opinions, which he defended against M.
Steyaert, one of Arnauld's opponents, and against the Jesuit fathers G. Bolek and
De Vos (especially on freedom and grace). The Inquisition in Toledo condemned
his treatise on penitence. and two collections of his theses were put on the Index in
1685 and 169 1. See Biographie nationale de Belgique ( Bruxelles 1886--7), vol. 9,
col. 729--46. Arnauld tried to support Huygens's candidature to the strict faculty in
a letter to Neercassel (25 June 1684, OA, 42: Supplement aux lettres, p. 46) and
voices his disappointment at Du Vaucel (letter of 15 October 1684, OA, 42, 48-9).
He mentions the matter again in his letter of 28 December 169 1 (OA, 3, 42 1) and
undertakes Huygens's defence against a libel in his letter to Du Vaucel of 2 1
December 169 1 (OA, 3, 4 14- 16).
15 As can be seen from his letter to Arnauld of 5 August 1693 ( OA 3. 669), Lamy had
read the Dissertatio a year and a half earlier; therefore. this work was completed
as early as February 1692. So the date of 1693 given in OA. 40, and in the
chronology of OA, 42. xvi, is false (as it is for the Reg/es de bon sens: see below. )
16 A document of the 'archives de Port-Royal' in Amersfoort (#32 17, 15, quoted by
G. Lewis, Le Probleme de / 'inconscient . . . , 204, n. I ) specifies that Huygens was
'tormented by the bull of Pius V which condemns the proposition that all actions
of the faithless are sins.'
17 ' Veritas principaliter est in intellectu, et secondario in rebus. ' Dissertatio, Art. I ,
quoted from Summa theologiae, Ia, 16, I , c. Huygens first tried to draw support
from St Thomas, but Arnauld corrects his interpretation.
18 Likewise, the so-called eternal truths are eternal inasmuch as they are contained
in the divine intellect, because it is itself eternal; in so far as they are in our
intellect, they are not eternal. ' Veritas propositionum, quae dicuntur aeternae
veritatis, non est aeterna, et ubique, proprie loquendo, sicut Deus est aeternus, et
ubique, sed tantum improprie, quia non est alligata cero tempori, et loco'
(Dissertatio, Coroll. VI).
19 Veritas enuntiabilium non est aliud quam veritas intellectus' : Thomas Aquinas,
Summa, I, 16, 7, c.
20 Dissertatio, Art. II, Coroll. II and III, 1 19-20.
2 1 Ibid., Coroll. IV.

22 See E. Gilson, ' Pourquoi saint Thomas a critique saint Augustin,' in Archives
d 'Histoire doctrinale et litteraire du Mayen Age I ( 1926). In the ad primum of
Summa theologiae, I, 16, 6, which Huygens quoted on his own behalf, St Thomas,
as Amauld remarks judiciously, only wants 'infringere auctoritatem Augustini
doctrinae suae oppositam, quam pro sua in Augustinum observantiam, perrumpere
nee audebat, nee volebat': ·s. Thomas hie obscurior est, quia occultare maluit
quam prodere suam ab Augustino dissensionem' (Art. III, 125).
23 Summa, I. 1 17, 1, c. Cf. I, 84, 5, c. and 88, 3 ad 1 and ad 2.
24 Dissertatio, Art. I I I, 130.
25 Ibid., Art. IV, ratio II, 134.
26 Ibid.
27 Here Amauld refers to Cartesian innate ideas, in order to prove that ideas are not
transcendent: 'opus tantum habet [mens nostra] ut attendat ad ideas claras et
simplices, quas in se ipsa reperit, ex quarum connexione ilia judicia effonnata
sunt' (ibid.)
28 Ibid., 134-5.
29 Ibid., Art. V, reply to Objection I, 136--7.
30 Avertissement of the editors of the Reg/es du bon sens, OA 40, 1 15. Cf. Lettres de
M. Nicole ( 17 18), Vol. 2, no. 47, 13 January 1693, to P. Quesnel.
3 1 See the Dictionnaire de theologie catholique, vol. 8, col. 2552-5; Dictionnaire de
spiritualite, vol. 9, col. 174-7; and Dom Jean Zehnder, ' Un Representant de la vie
intellectuelle fran�aise entre 1680, et 17 10. Dom Fr. Lamy Moine benedictin et
religieux de la Congregation de Saint-Maur 1636--17 1 1. Essai d'introduction a sa
vie et a son oeuvre' (thesis, University of Fribourg, Zoug; Imprimerie E. Kalt
Zehnder 1944). He was born in 1636 and died in 17 1 1, was a Benedictine monk
and professor of philosophy in Soissons in 1672-3. Himself a student of Rohault,
he introduced Cartesianism into the congregation of Saint-Maur. This is perhaps
why he was deprived of his office of Prior by royal order in 1687. His spiritual
works seem to be influenced by Jansenism. However, he championed Fenelon's
cause against Bossuet and Malebranche. (The latter answered him with the Traite
de I 'amour de DieuJ
32 It remained unpublished, and is considered as lost by J. Zehnder, ' Un
Representant de Ia vie intellectuelle fran�ais . . . ,' 200, but it is reported under the
title Reflexions sur le traite de la grace generale by Lewis in Le Probleme de
/ 'inconscient .. . , 276.
33 See J. Zehdner, ' Un Representant de la vie intellectuelle fran�ais . .. ,' 9 1 and 200,
and Lewis, Le Probleme de / 'inconscient . .. , 2 12, n. 5, and 276. It was composed
of fifty pages or so (cf. Lewis, ibid.) and was divided into remarks, as it appears
from some passages of Arnauld's retort to Lamy (Reg/es du bon sens, OA 40, 157,
158, etc. ) It was meant by Lamy for Nicole alone, but the latter committed the
indiscretion of communicating it to Arnauld. When this came to Lamy's
knowledge, he wrote to Amauld to apologize for the sharp tone he used in his
reply (letter of 5 August 1693, QA 3, 670). But Amauld had already showed that
he did not feel offended in his letter of 22 April 1693, to an anonymous
correspondent (QA 3, 624), and confirmed it to Lamy on 12 September 1693 (QA
3, 676), after the latter had written to him on 3 1 August 1693 (QA 3, 673--4) to
thank him for his leniency. (He had seen the letter to the anonymous
correspondent.) Thus, despite their disagreement, Arnau Id and Lamy remained
good friends.
34 First published in 17 15, in the Recueil des ecrits sur la grace generate, I, by
Fouillou and Petitpied, the work was completed in May 1693, as Amauld
informed Du Vaucel (letter of 22 May 1693, QA 3, 1693). It was reprinted in QA,
40. In what follows, I refer directly to page numbers in that edition.
35 In 167 1, when Louis XIV forbade the teaching of the new philosophy, Amauld
addressed to the Parliament a memoir entitled Plusieurs raisons pour empecher la
censure ou la condamnation de la philosophie de M. Descartes. There he
defended the freedom of thought (on subjects irrelevant to Catholic dogmas) and
proved that it is dangerous to involve ecclesiastic and civil authority in such
matters. He protested also against the censors who put Descartes on the Index, and
not Gassendi. Against some of his friends of Port-Royal, who thought that
philosophy was a waste of time, he was convinced that Cartesianism (thanks
especially to its sharp distinction between soul and body) was able to stop
materialism and atheism. Again in the name of Cartesianism, he opposed the
theories of Malebranche and of Nicole, as we shall see later.
36 Cf. Thomas Aquinas, quoted in Dissertatio bipartita, Art. IV: ' Illud propter quod
aliud cognoscitur, erit magis notum, ut principia conclusionibus' (Summa, Ia, 87,
2, ad 3m).
37 Cf. Thomas Aquinas, quoted in Dissertatio bipartita, Art. I I I : 'Dicendum est quod
aliquid in aliquo dicitur cognosci dupliciter. Uno modo sicut in objecto cognito,
sicut aliquis videt in speculo ea, quorum imagines in speculo resultant: sic in
rationibus aetemis cognoscunt omnia beati, qui Deum vident, et omnia in ipso.
Alio modo dicitur aliquid cognosci in aliquo, sicut in cognitionis principio: sicut
dicimus quod in sole videntur ea quae videntur per solem: et sic necesse est dicere
quod anima omnia cognoscat in rationibus aetemis, per quarum participationem
omnia cognoscimus. Ipsum enim lumen intellectuale quod est in nobis, nihil est
aliud quam quaedam participata similitudo luminis increati, in quo continentur
rationes aeternae ... ' (Summa, Ia, 84, 5, resp.).
38 To speak is to explain one's thoughts by signs invented by man for that intention
(A. Amauld and C. Lancelot, Grammaire generate et raisonnee, QA 4 1, 5). The
Amauld versus Nicole 1 45
discussion has to do with the definition d 'usage of the term 'truth ' and the real
definition (qua res) of 'truth. ' In the latter sort of definition, as the Logic explains,
' on Iaisse au terme qu'on definit, comme homme ou temps, son idee ordinaire,
dans laquelle on pretend que sont contenues d'autres idees, comme animal
raisonnable ou mesure du temps; au lieu que dans la definition de nom, comme
nous avons deja dit, on ne regarde que le son, et ensuite on determine ce son a etre
signe d ' une idee que l ' on designe par d'autres mots' (OA 4 1 , 1 7 1 ). The nominal
definition is free, for the signification of a sound is arbitrary, and one can set it ad
libitum. But one is not free to enforce it in the discussion, instead of the definition
d 'usage: ' comme Jes hommes ne sont maitres que de leur langage, et non pas de
celui des autres, chacun a bien le droit de faire un Dictionnaire pour soi, mais on
n ' a pas le droit d 'en faire pour les autres, ni d 'expliquer Ieurs paroles par les
significations particulieres qu' on aura attachees aux mots. C' est pourquoi quand
on n' a pas dessein de faire connaitre simplement en quel sens on prend un mot,
mais qu' on pretend expliquer celui auquel ii est communement pris, les
definitions qu' on en donne ne sont nullement arbitraires; mais ell es sont liees et
astreintes a representer, non la verite des choses, mais Ia verite de I ' usage; et on
Jes doit estimer fausses, si ell es n' expriment pas veritablement cet usage; c' est-a
d ire, si elles ne joignent pas aux sons Jes memes idees qui y sont jointes par
) ' usage ordinaire de ceux qui s'en servent' (OA 4 1 , 1 78).
39 Cf. Descartes, Principles IV, 205.
40 ' Potui . . . pro certo affirmare nihil in me, cujus nullo modo sim conscious, esse
posse' (Meditations VI, AT 7, 8 1 ).
41 Using the same Cartesian principle (that unconscious thought is impossible),
J.-P. Sartre will contest the Freudian theory of a psychological unconscious ( or,
more accurately, the description of the unconscious as an entity able to censure,
disguise, use craft, etc. - in short, of acting as a genuine, but unconscious,
consciousness).
42 However, Amauld himself had objected to the Meditations: 'Quis non videt multa
in mente posse esse, quorum mens conscia non sit' (Fourth Objections, AT 7, 2 1 4;
9, 1 67). Descartes answered: ·nous voyons fort bien qu'il n'y a rien en lui
[I ' esprit], Iorsqu' on le considere de la sorte [comme une chose qui pense ], qui ne
soit une pensee ou qui ne depende entierement de la pensee: autrement cela
n 'appartiendrait pas a l 'esprit en tant qu'il est une chose qui pense; et ii ne peut y
avoir en nous aucune pen see, de Iaquelle, dans le meme moment qu' elle est en
nous, nous n 'ayons une actuelle connaissance' (AT 9, 1 90; 7, 246).
43 On discussions among Cartesians concerning the value of the testimony of
consciousness, see Lewis, Le Probleme de l 'Jnconscient . . . , ch. II.
44 In Descartes, thought is defined by consciousness, though this does not always
entail express reflexion. So he can say that we have been actually thinking since
146 Jean-Luc Solere
the beginning of our existence, even if we have no memory of it.

45 Cf. Des Vraies et des Fausses ldees, ch. II, repr. in Corpus des oeuvres de
philosophie en langue fran9aise (Paris: Fayard 1986), 23.
46 Fr. Lamy, De la Connaissance de soi-meme (Paris: 1694-8), vol. III, 36 1.
47 Traite de la grace generate, I, 122-3.
48 Ibid.
49 Amauld quotes the Premier Etablissement de la foi dans la Nouvelle France of
Chretien Le Clercq.
50 This appears also from his controversy with Malebranche and Bayle on the
question of pleasure and happiness. See my contribution to the Antoine Amauld
Conference in Paris, October 1994: 'Tout plaisir rend-ii heureux? Une polemique
entre Amauld, Malebranche et Bayle,' Chroniques de Port-Royal no. 44 (Paris:
Bibliotheque Mazarine 1995).
10
' Tange mantes et fumigabunt' :

Arnauld on the Theodicies of
Malebranche and Leibniz
STEVEN NADLER
In a century full of brilliance and great personalities, Malebranche, Arnauld,

and Leibniz deserve credit for shining particularly brightly. Arguably three of
the seventeenth century's most important thinkers after Descartes, they none
the less were gifted in different ways: Leibniz was blessed with sheer natural
brilliance and a vision both broad and penetrating; Arnauld's talent lay in keen
analytical skills and was perhaps the most acute critical mind of his time; and
Malebranche. while slow at times, was the great and pious synthesizer (of
Augustine and Descartes) and system builder. What makes these three philoso
phers particularly fascinating as a group is their mutual relations. The Arnauld
Malebranche debate is one of the intellectual events of the century; the Arnauld
Leibniz correspondence proves crucial in the development of Leibniz's mature
metaphysics; and Leibniz's admiration for Malebranche is evident even when
the two are deeply engaged in polemic. Sometimes the fray was bloody, and
tangling with Arnauld could be a dangerous affair. Leibniz, for one, knew what
he was getting into, but was none the less taken aback when Arnauld trashed the
outline of the Discours de la metaphysique (1686; hereinafter Discours) that
Leibniz had sent for his perusal. To paraphrase Leibniz: 'It's no wonder you
don't have any friends left.' 1
The philosophical issue I focus on here, as a way of bringing the three think
ers together, is the theodicy problem - that is, the problem of justifying God's
ways in the face of the apparent imperfections and evil in the world, and of the
apparent unfairness in the distribution of grace. What is initially remarkable
(but not surprising) is a similarity: in this case, it concerns the general theodicean
strategy adopted by both Malebranche and Leibniz. But what I will show is not
1 48 Steven Nadler
that there are no differences between Leibniz's and Malebranche's theodicy. In

fact, in the course of my argument, I take note of some important differences
between their respective accounts, differences which are recognized both by
recent scholars and by Malebranche himself. Rather, I argue that, for Arnauld,
whatever differences there may in fact be between the two accounts would be
outweighed by some crucial similarities, not just in general theodicean strategy,
but also in the mechanics of the divine modus operandi in the realms of nature
and grace. And using Arnauld's perspective in this way aids us in identifying
those more particular but less obvious similarities.
In 1 680, Malebranche published his Traite de la nature et de la grace (herein

after, TraiM), the work that ignited Arnauld's wrath and that caused Arnauld to
reconsider his earlier mild judgment of Malebranche' s De la rechere he de la
verite ( 1 674-5). The Traite elaborates on themes previously discussed in the
Recherche and the Conversations chretiennes ( 1 676). Malebranche explains
how God's omnipotence, benevolence, and perfection can be reconciled with
the persistence of evil and imperfections in the world (including the balance of
human pleasure and pain) and with what appears to be an unjust, and even
haphazard, distribution of the grace required for everlasting happiness.
Our concept of God tells us that God is infinitely wise, good, powerful, and
perfect. God's knowledge and benevolence are without bounds, and God's will
and power are necessarily efficacious. 2 Moreover, God wishes, with respect to
the realm of nature, to make the most beautiful and perfect world possible, the
better to express and honour the qualities and perfections possessed by God
h imself. None the less, the world which God has, in fact, created certainly ap
pears to us to be quite imperfect in its details and full of disorders of every
variety. Monsters and deformed creatures are born; there is sin and misery among
human beings; even the rain regularly falls on the sea, where it is not needed,
rather than on fertile soil. 3 As Theodore, Malebranche' s spokesman in the
Entretiens sur la metaphysique, exclaims, ' The Universe then is the most per
fect that God can make? But really! So many monsters, so many disorders, the
great number of impious men - does all this contribute to the perfection of the
universe?'4 Aristes, his interlocuter, is led thereby to wonder either about the
efficacy of God's will or the benevolence of God's intentions: ' God wishes to
make a work that is the most perfect possible. For, the more perfect it is, the
more it will honor Him. This appears evident to me. Yet I understand that the
work would be more accomplished if it were free of thousands upon thousands
of defects which disfigure it. Here is a contradiction that stops me quite short. It
Amauld on the Theodicies of Malebranche and Leibniz 1 49
seems that God did not accomplish His plan, or that He did not adopt the plan
most worthy of His attributes. ' 5
The resolution of this conundrum, as presented in both the Entretiens and the
Traite, is to be found in the consideration, not just of the details of the visible
universe, not just of the particular effect wrought by God, but also of the means
undertaken to achieve and sustain this product. According to Malebranche, God,
when creating, looks not only to the final result of the creative act (that is, to the
goodness and perfection of the world per se). God is honoured, not just by the
creatatum, but also by his work or ways. And the activity or means most expres
sive of God' s nature are of maximum simplicity, uniformity, fecundity, and uni
versality.6 God does not accomplish by complex means that which can be ac
complished by simple means; and God does not execute with many particular
volitions that which can be excecuted by a few general volitions. This holds true
even if it means that the world created by God could be spared some imperfec
tions were God to compromise the simplicity and generality of his operations.
That is, the perfection of the world in its details as a product is completely
relative to the mode of activity that is most worthy of God. God might increase
the absolute perfection of the world, perhaps by decreasing the number of de
fects or evils therein. But this would entail greater complexity in the divine
ways. God might, for example, keep the rain from falling on anything but fertile
and inseminated soil. But this would involve departing from the general laws of
nature established at creation.
In a sense, it can be said that God wants all his creatures to be perfect; that he does
not want infants to perish in their mothers' wombs; that he does not like monsters,
and that he has not made the laws of nature in order to engender them; and that if
he had been able, by equally simple ways, to create and conserve a more perfect
world, he would not have established laws from which so many monsters necessar
ily result. But it would have been unworthy of his wisdom to multiply his volitions
in order to prevent certain particular disorders. 7
Thus, the world that God has created is the one out of the infinitely many
possible worlds that best reconciles perfection of design with simplicity and
generality of means of production and conservation. 8 By a number of 'particular
volitions [volontes particulieres]' - volitions that are ad hoc and not occasioned
by some prior event in accordance with some law of nature - God could correct
deformities of birth, keep fruit from rotting on trees, prevent physical disasters
about to occur by the regular course of the laws of nature, and forestall sin and
wickedness. But, Malebranche insists, 'we must be careful not to require con
stant miracles from God, nor to attribute them to him at every moment. ' 9 God,
1 50 Steven Nadler
in other words, only acts by 'general volitions [ volontes genera/es] ' - volitions
that are in accordance with some law and whose operation is occasioned by a
prior event, as dictated by that law - and the most simple ways (!es voies !es
plus simples), and never by particular volitions (Dieu n 'agit point par des volontez
particulieres). The solution to the problem of evil, then, is found in the simplic
ity and uniformity of God's causal conduct in the world, in the generality of the
divine will.
Similar considerations apply to the problem of grace. A benevolent God wills,
with what Malebranche calls a 'simple volition,' that sinners convert and that
all humans should be saved. But clearly not all humans are saved; many are
lost. And not all those who receive grace appear to be ready to receive it, or even
worthy of salvation. The anomaly is again explained by the generality of God's
volitions. The distribution of grace is, like the events of nature, governed by
certain general laws willed by God, and the occasional causes responsible for
the actual distribution of grace in accordance with those laws are the thoughts
and desires in the human soul of Jesus Christ. Because Jesus qua human is not
omniscient, at any given time he never knows with full awareness all the rel
evant facts about the agents upon whom grace is to be bestowed - for example,
whether they are ready to make proper use of it. Thus, as with the distribution of
evil and imperfection in the natural world, God allows grace to be disbursed
unevenly, and even inequitably, by the laws of grace in combination with the
occasional causes that activate them. 1 0
Leibniz was generally impressed by Malebranche' s insights, 1 1 and this is
reflected in Leibniz's own solution to the problem of evil. God, Leibniz claims,
does everything in the most desirable way, and cannot have made things any
better than they are. This is, he insists, the best of all possible worlds. But what
makes it best is obviously not the total absence of pain and other apparent evils.
Nor is it that these are simply outweighed by a great abundance of pleasures and
other good things. Rather, God has chosen the one out of the infinitely many
possible worlds that best combines simplicity of laws or 'hypotheses' and rich
ness of phenomena. The world must contain the greatest amount of possibility
or essence, but be governed by laws which are of maximum simplicity. 1 2 God is
like a good architect who ' utilizes his location and the funds destined for the
building in the most advantageous manner, ' or like an ' excellent Geometer who
knows how to find the best construction for a problem.' 13 The world we experi
ence may not seem perfect to us in its details; it may be full of apparent irregu
larities and suffering. But, Leibniz says, ' I do not believe that that which is best
and most regular is always convenient [commode] at the same time for all crea
tures. ' 14 These unpleasant features of the world belong to the course of nature as
determined by its laws, laws which are themselves few, simple, universal, and
fertile . Evil and misfortune are permitted by God, but not positively willed,
Arnauld on the Theodicies of Malebranche and Leibniz 1 51
since they occur 'by means of the laws of nature that he has established'; 1 5 they
are a part of the rich variety of phenomena that, following from and in combina
tion with those laws, constitute the best overall result.
One cannot help but notice striking similarities here with Malebranche's ac
count. For both Leibniz and Malebranche, God in creation chooses from an
infinity of possible worlds, and pays particular attention, not just to the created
theatre itself, but especially to its relationship with the laws of nature and grace,
laws that must be of maximum simplicity. Malebranche and Leibniz agree that
evil and sin occur because God allows them to occur as a result of the ordinary
course of nature as governed by the laws God has chosen (for Leibniz, there is a
sense in which God actually wills or intends the evil, but only with what he calls
a 'permissive will' ). And they agree that God could (theoretically) diminish, or
even eliminate, the apparent imperfections in the world - the quantity of pain
and unhappiness or inconvenience - but only by interfering with the laws, and
thus violating the simplicity of the divine ways (as Malebranche would put it) or
by detracting from the overall and maximum metaphysical goodness or perfec
tion of the world (as Leibniz would say). These points of agreement have been
well catalogued by others, most notably Catherine Wilson 16 and Robert C. Sleigh,
Jr. 1 7 As general theodicean strategies, then, the Malebranchian and Leibnizian
accounts share some important substantive features. On both accounts, crea
tures are allowed to suffer, the needs and desires of many are unsatisfied, there
is sin, and not all humans are saved, because God must satisfy some higher
value - for Malebranche, that higher value is simplicity oflaw/means; for Leibniz,
it is a best overall world that is simplest in laws and richest in effects.
Does this mean that Leibniz's theodicy and Malebranche's theodicy are, in
all practical respects, equivalent? One recent scholar suggests as much, saying
that 'even if they do not agree on all points, one is led to believe that they are to
all intents saying the same thing.' 1 8 Catherine Wilson presents a more nuanced
view. While noting that 'Leibniz' s original rationale for optimism was
Malebranchian, ' and that 'he continued to think of the justification of evil in
Malebranchian terms, ' none the less she concludes that 'he never succeeded in
integrating the Malebranchian solution with his own theory of individual sub
stance. ' 1 9 And Sleigh insists that there is a fundamental difference between
Leibniz and Malebranche on the ranking of values that God must coordinate in
the choice of the world, and concludes that they 'appear to differ on details,
albeit details of substance. ' 20
Donald Rutherford takes the analysis even deeper. He recognizes a funda
mental difference in the way in which Malebranche and Leibniz conceive of
God's wisdom and of how it gets played out in God's aim in creation. For
Malebranche, God acts only for the sake of his own glory, 2 1 which he finds
reflected in a work that is sanctified by a divine person (Jesus Christ) and that is
152 Steven Nadler
created and sustained by sufficiently worthy means. Leibniz's God, on the other
hand, aims to produce the maximum amount of goodness possible, and this
goodness is an inherent feature of the world itself. Rutherford notes that, 'within
[Leibniz's] scheme, the issue of the worthiness of the created world vis a vis
God receives a completely different treatment than in Malebranche. Quite sim
ply, we can say that the only possible world worthy of God is that world which
contains, in and of itself, the greatest perfection or reality. ' 22 God's wisdom for
Leibniz, then, consists in the 'knowledge of the good,' and it is wisdom so
understood that guides God in choosing the best. For Malebranche, on the other
hand, God's wisdom is expressed in the simplicity of his ways. As Rutherford
concludes, Malebranche's position is thus fundamentally at odds with the main
tendencies of Leibniz's thought. 23
But now consider Leibniz's own assessment of the situation:
The ways of God are the most simple and uniform : for he chooses rules that least
restrict one another. They are also the most productive in proportion to the simplic
ity of ways and means ... one may, indeed, reduce these two conditions, simplicity
and productivity, to a single advantage, which is to produce as much perfection as
is possible: thus, Father Malebranche's system in this point amounts to the same as
mine. 24
Upon reading this in 1711, Malebranche demurs, and suggests to Leibniz that
there is an important difference in their respective views:
I am persuaded, as you are, that God has produced for his creatures al l the good that
he can produce for them, acting nonetheless as he must act, that is to say, acting
according to his law which can only be the immutable order of his divine perfec
tions, which he invincibly loves and which he can neither violate nor neglect. And
thus his work [son ouvrage] is the most perfect it can be, not absolutely, however,
but in relation to the means in accordance with which it is executed. For God is
honored not only by the excellence of his work, but also by the simplicity and the
fecundity, by the Wisdom of his ways. 2 5
A month later, Leibniz in effect acknowledges that he may have oversimplified

things a bit, and explains to Malebranche the ground of the difference
Malebranche is describing:
In effect, when I consider the work [l 'ouvrage] of God, I consider his ways as a part
of the work, and the simpl icity joined to the fecundity of the ways form a part of the
excellence of the work, for in the whole the means form a part of the end [and the
simpl icity of means form a part of the excellence of the work) . 26
Amauld on the Theodicies of Malebranche and Leibniz 153
But towards the end of the letter, Leibniz goes on to recap some important points
of agreement, and I wonder if he is not suggesting that the difference between
them is really just a nominal one: Leibniz says that God creates the best of all
possible worlds; Malebranche says that God does not create the best of all possi
ble worlds - but this is only because Malebranche uses the term 'world' (or
ouvrage) more narrowly than Leibniz does to refer only to the created product,
exclusive of the laws governing it. This is what allows Malebranche to declare
that the goodness of the world is only relative to God's means, or the laws.
Leibniz' s point could be that, terminological differences aside, they agree that
the composite of product design and laws is, taken as a whole, the best overall
and the most worthy of God's wisdom and benevolence, even if the product
design itself, exclusive of the laws, seems fraught with imperfections. 27
And yet, in an addition to the Traite made in the fourth edition ( 1684 ), it
becomes clear that, in Malebranche's eyes, at least, the differences must be more
substantial than Leibniz seems willing to allow. For Malebranche's God is not
concerned with producing as much good as possible, as is Leibniz's God.
Malebranche's God is not trying to balance or even maximize two values to
produce the best outcome. Rather, for Malebranche, the simplicity of God's ways
takes precedence: ' His wisdom in a sense renders him impotent; for since it
obliges him to act by the most simple ways, it is not possible for all humans to be
saved [as God would like], because of the simplicity of his ways ... God loves his
wisdom more than his work ... because his wisdom prescribes means which
most bear the character of his attributes.'28 One can characterize the difference
as that between Malebranche's God the deontologist, for whom a particular
value must be pursued no matter what the consequences, and Leibniz's God the
consequentialist, who chooses means in order to produce as much overall good
as possible. 29
II
There are important differences, then - noted both by recent scholars and by
Malebranche himself - between the theodicies of Malebranche and Leibniz,
especially regarding how God's wisdom is conceived and the goals and details
of his operations. And these differences are not inconsistent with the general
(but not abstractly trivial) similarities in theodicean strategy that I (partly tak
ing my cue from Leibniz) describe above.
I would now like to approach the issue from another perspective, and here is
where Amauld comes in. In 1686, when commenting on Leibniz's outline of
the Discours, Amauld has just finished his masterful and massive critique of
Malebranche' s Traite, the Reflexions philosophiques et theologiques sur le
nouveau systeme de la nature et de la grace (1685; hereinafter, Reflexions).
154 Steven Nadler
Arnauld never directly comments on Leibniz' s theodicy strategy, especially

proposition five of Leibniz's summary, whose terms are certainly reminiscent of
Malebranche' s account: 'In what the principles of the perfection of the divine
conduct consist and that the simplicity of the means is in balance with the rich
ness of the effects. ' 30 And yet we can ask the following speculative but relevant
question: Is Leibniz's theodicy subject to the same criticisms that Amauld lev
els at Malebranche? Must it, in Amauld' s eyes, suffer from the same flaws; or,
on the other hand, are there substantial-enough differences between Leibniz' s
account and Malebranche's account for Leibniz to escape Amauld' s charges?
These are important questions. While I do not intend to claim that Amauld' s
perspective on the similarities and differences between Leibniz and Malebranche
is in any sense a privileged one, none the less in the seventeenth-century context
of Leibniz' s and Malebranche' s discussions oftheodicy, Amauld' s point of view
is rather crucial. First, Amauld is one of the period's leading thinkers on the
issue of grace, and the theodicy problem arises in the seventeenth century mainly
over a problem of grace: how is it that not all humans are saved when Scripture
tells us that God wants to save all?3 1 Second, Leibniz is himself trying to earn
Amauld' s approval and thus win him over to his ecumenical proj ect32 - and in
this he will certainly not succeed unless he can, in Amauld's eyes, distinguish
himself from Malebranche. I argue that, through the lens of Amauld' s critique
of Malebranche, we can discern some further differences in detail between the
two theodicean accounts. Yet in the end, through that same lens, and even at the
level of detail - especially with respect to the way God's omnipotence is treated,
a crucial consideration for Amauld in assessing a theodicy - the differences
tum out to be less important than some essential similarities.
Prima facie, it looks as though Leibniz does escape some of Amauld' s
criticisms. Consider Amauld's complaint that Malebranche humanizes God's
mode of acting. He accuses Malebranche of limiting God's omnipotence by in
sisting that God must choose the simplest means, and therefore must proportion
his work to fit those means. Thus, God chose the world which he could produce
and conserve by the simplest laws. Now it may be true, Amauld says, that hu
man beings often have to tailor their work to fit their means. An architect might
not be able to make the building he really wants to make, the absolutely best
building of its design, if the funds available are limited. But, he argues, this
kind of reasoning cannot possibly apply to an omnipotent God, for whom all
means are equally easy and available. 'All means for executing his designs are
equally easy for God ... and his power so renders him master of all things and so
independent of the need for help from others. that it suffices that he will for his
volitions to be executed. ' 3 3
Leibniz, when composing the Discours de la metaphysique, certainly had
Amauld on the Theodicies of Malebranche and Leibniz 1 55
Amauld' s Reflexions in mind. 34 And it may be precisely this criticism that

Amauld directs at Malebranche which occasioned Leibniz to say in the Discours
that 'nothing costs God anything, just as there is no cost for a philosopher who
makes hypotheses in constructing his imaginary world, because God has only to
make decrees in order that a real world comes into being. '3 5 This kind of remark
should satisfy Amauld's demand that any theodicy respect God's omnipotence,
and perhaps Leibniz is on better ground here than Malebranche. After all,
Leibniz's God aims to create as much perfection as possible, or a world that is
absolutely the best, and necessarily succeeds in doing so; whereas Malebranche' s
God would like to create the best world but cannot execute that desire. More
over, unlike Malebranche's God, who must sacrifice perfection of effect for the
sake of simplicity of means and universality of law, Leibniz's God apparently
gets to maximize all his values: simplicity of laws and richness of effect. As
Malebranche notes, simple laws have unfortunate consequences (La simplicite
des lois . . . a, necessairement, des suites fdcheuses), 3 6 whereas, for Leibniz, the
most simple laws are also the ones that produce the richest effects. The point of
Leibniz's theodicy is simply to explain that a world that is simplest in laws and
richest in effects - that is, the best of all possible worlds - will not necessarily be
a world that is the most convenient for all the particulars it contains; it will not
necessarily be a world in which each individual being enjoys its highest perfec
tion and all humans are saved.
Finally, both Arnauld and Leibniz appear to agree on the moral character of
God's creation. Arnauld accuses Malebranche of granting to the enemies of
religion that the world really is imperfect, and claims that now Malebranche' s
project is simply to explain this as a necessary and unavoidable consequence of
the simplicity and generality of the laws. 3 7 This entails, as well, that, for
Malebranche, the irregular and imperfect aspects of the world which are sup
posed to detract from its beauty and goodness are parts of creation not 'posi
tively and directly' willed by God. The true view, Arnauld insists, is that of St
Augustine and St Thomas: the imperfection of the world is only in appearance.
God desired to create a world which is on the whole entirely good and beautiful.
And each thing in that world was positively and directly chosen by God because
it contributes to the world's goodness and beauty. To be sure, there are events we
regard as unfortunate, and beings whom we consider monstrous. But we must
not allow these to adversely affect our assessment of the whole, for our view of
the whole is necessarily restricted and incomplete. Even monsters contribute to
the beauty of the world.
Nothing seems more contrary to music than dissonances, or what we otherwise call
false agreement; and yet a dissonance, mixed with many consonances, is what
1 56 Steven Nadler
makes for a most excel lent hannony. And do we then say that, this dissonance
being a false agreement, it was not positively and directly put in the composition by
the composer? Similarly, a monstrous animal is a kind of dissonance in the har
mony of the universe; but it does not fail to contribute to the overall hannony. 3 8
There are no faults in God's work, Amauld concludes (ii n y a aucun defaut
dans /es ouvrages de Dieu). It is all a matter of perspective. If somehow our
vision could be expanded to take in the whole - and by 'whole' Amauld means
both the visible elements of the world and its principles or laws - we would see
the beauty of the universe and the contribution each part makes to that beauty.
Now, if this is Amauld's view, how could he fail to appreciate Leibniz's al
most verbatim approach? When arguing that the world is both metaphysically
and morally the most perfect, Leibniz insists, to those who would deny this by
pointing to the misfortunes and confusions that seem to predominate, that 'it is
truly unjust to render a judgment without having studied the whole ... Great
composers very often mix dissonance with harmonious chords to stimulate the
hearer [so that he may be pleased]. ' 3 9 In the Theodicee, too, Leibniz says that
the imperfection of the world is only apparent, and a function of our limited
insight.
God, by a wonderful art, turns all the errors of these little worlds to the greater
adornment of his great world. It is as in those devices of perspective, where certain
beautiful designs look like mere confusion until one restores them to the right
angle of v ision or one views them by means of a certain glass or mirror. It is by
placing and using them properly that one makes them serve as adornment for a
room. Thus the apparent deformities of our little worlds combine to become beau
ties in the great world, and have nothing in them which is opposed to the oneness of
an infinitely perfect universal principle: on the contrary, they increase our wonder
at the wisdom of him who makes evil serve the greater good. 40
One is tempted to conclude that, in fact, it is Amauld and Leibniz who share the
same theodicean strategy, and not Malebranche and Leibniz.
It would appear, then, that Leibniz's theodicy should, from Amauld' s per
spective, fare better than Malebranche's. And yet, 1· suspect that such a conclu
sion would be based on a fairly superficial comparison, and that, ultimately, at a
deeper level, Amauld should be no more pleased with Leibniz' s system than
with Malebranche' s.
Let us go back to Amauld' s critique of Malebranche. As we saw, what par
ticularly bothers Amauld is the way in which Malebranche' s theodicy dimin
ishes God's omnipotence and freedom. Malebranche claims that God's wisdom
and simplicity constrain his power: although God would have liked to create a
world that is absolutely the best, he cannot. God, for Malebranche, necessarily
follows the principles of Order; that is, God's power is regulated and limited by
his wisdom, and this wisdom prescribes to God certain means. As Malebranche
puts it, ' the immutable Order of Justice is the essential rule governing the will
of God. ' 4 1 But, for Amauld, God's will has no rule other than itself: this is the
very meaning of omnipotence and divine freedom.42 According to Amauld, on
Malebranche's view ' God is impotent in that he cannot choose means that are
unworthy of his wisdom and that, instead of bearing the character of his wis
dom, goodness, constancy, and immutability, bear a character of less intelli
gence, of malevolence, inconstancy, and lightness of mind. ' 43
We can put all this another way by focusing on a particular case of the theodicy
problem in the realm of grace. Malebranche takes seriously and literally the
words of St Paul: God wants all humans to be saved.44 But, of course, not all
humans are saved; many souls are lost. This is because God wills all humans to
be saved only with what Malebranche calls a 'simple volition' - a volition or
preference that does not necessarily get carried out, for some reason or another.
Just because God has a simple volition for something does not mean that that
thing or state of affairs obtains. In this case, the reason lies in God's wisdom,
that is, immutable order and the simple laws of grace. In other words, not all of
God's volitions are efficacious: 'When it is claimed that all the volitions of God
are efficacious, this refers only to practical volitions [ volontes pratiques]. For
there are things God wants but does not do. ' 4 5
Amauld has two serious problems with this account, both of which have their
ultimate ground in his Jansenism. First, he cannot accept the claim that God
literally wills all individual humans be saved. Nothing could be farther from
Jansenist doctrine, which stipulates in no uncertain terms that God in his mercy
has decided to save some humans and deliver them to glory by infallible means,
while deliberately leaving others to perdition.46 The phrase 'God wills that all
humans be saved' is Scripture, to be sure, but Amauld insists that it needs to be
properly interpreted.4 7 Malebranche, however, takes the phrase in a strong lit
eral sense, and this, to the Jansenist Amauld, is per se unacceptable. Second,
the Jansenist doctrine of predestination aside, there is an obvious tension be
tween the claim that God wills the salvation of all and the evident fact that not
all are saved. If God's volition to save all is taken literally and is a real volition
(as Malebranche would have it), then this implies, as Malebranche willingly
grants, that not all of God's volitions are efficacious. And this, too, Amauld (or
any Jansenist) cannot accept. It is absurd to say that a divine volition, the voli
tion of an omnipotent being, is not accomplished. As Amauld states, 'if God
willed that all humans be saved, then they would be. ' 48 Or, more generally, 'it
suffices that [God] will in order for his volitions to be carried out. ' 49 This issue
is, of course, intimately bound up with the Jansenist doctrine of efficacious grace.
158 Steven Nadler
Similarly, if, as Malebranche insists, God wills that the world be absolutely the
best but this volition cannot be carried out, then not all of God's volitions are
efficacious, and God's omnipotence is undermined - by his own wisdom, to be
sure, but undermined none the less.
Turning to Leibniz, we find that in Arnauld's eyes his theodicy must appear
to undermine God's omnipotence in similar ways. Leibniz's God, it is true,
necessarily accomplishes his plan to create the best. But it is also the case that,
for Leibniz God's will to create the best, and thus his power to do so or other
wise, is determined (albeit with a moral but not a metaphysical necessity) by his
wisdom and goodness: 'Moral necessity ... constrains the wisest to choose the
best. ' 50 Or, in other words, Leibniz's God does not will the world 'freely and
indifferently,' as Arnauld would have it, but is, like Malebranche' s God. con
strained by reason and principle: 'The highest liberty [is] to act in perfection
according to the sovereign reason. ' 5 1 Once again, wisdom wins out over power,
sagesse over toute-puissance. I agree entirely with Sleigh's judgment here: 'Had
Arnauld grasped the full scope of the principle of sufficient reason in Leibniz's
philosophy, in particular its application to God's will in every single act of that
will, even creation, Arnauld would have been convinced that Leibniz's scheme
fared no better than Malebranche's with respect to a proper account of God's
freedom in creation. The fact is that Arnauld saw .. . the idea that there must be
some reason for God's decision to create, other than simple appeal to his will, as
the real culprit.' 52 The deep concern for safeguarding God's freedom that Amauld
demonstrates in his correspondence with Leibniz over metaphysics would, I am
certain, simply carry over to the theodicy question as well. 5 3
Amauld would not be pleased, then, with Leibniz's use of the distinction
(adopted from St Thomas) between an antecedent will and a consequent will,
which is basically equivalent to Malebranche's distinction between a simple
volition and a practical volition. 54 God wills only good for each and every thing,
but only with an antecedent will. These antecedent wills do not necessarily get
carried out, for the sake of the consequent will to create the best overall.
Taking it in the general sense, one may say that will consists in the inclination to do
something in proportion to the good it contains. This will is called antecedent when
it is detached, and considers each good separately in the capacity of a good. In this
sense it may be said that God tends to all good, as good, ad perfectionem simpliciter
simplicem, to speak like the Schoolmen, and that by an antecedent will. He is
earnestly disposed to sanctify and to save all men, to exclude sin, and to prevent
damnation. It may even be said that this will is efficacious of itself (per se), that is,
in such sort that the effect would ensue if there were not some stronger reason to
prevent it: for this will does not pass into final exercise (ad summum conatum),
else it would never fail to produce its full effect, God being the master of all things.
Success entire and infal lible belongs only to the consequent will, as it is called.
This it is which is complete� and in regard to it this rule obtains, that one never fails
to do what one wills, when one has the power. 55
Note that Leibniz has God will the salvation of all humans, albeit only with an
antecedent and inefficacious volition. (Apparently the best of all possible worlds
must allow for the eternal loss of many.) It was exactly this that brought Amauld's
wrath down upon poor Malebranche, and I see no reason to think that it wouldn't
equally bring it down upon Leibniz as well. There is suffering and sin, and
humans are lost, because God is, in some sense, impotent to carry out his (ante
cedent/simple) volitions which are geared to the individual good of each crea
ture and would effectively save all humans. Amauld may have had a fine mind
for fine distinctions, but even the 'moral' and non-metaphysical necessity that
constrains Leibniz's (and Malebranche's) God would be too much for him. It is
still a rule higher than God's power, and one that renders some of his volitions
inefficacious.
III
By putting Leibniz' s theodicy in the context of the Amauld-Malebranche

debate, we gain an illuminating perspective on what may be the similarities and
differences between his and Malebranche's account. We see that in addition to
the general strategical similarities between the Malebranchian and Leibnizian
theodicies, there are also, for Amauld, deep and substantive similarities of de
tail regarding the nature of God's operations in the realms of nature and grace,
similarities that emerge in the light of Amauld's criticisms of Malebranche. I
do not at all intend to minimize the important differences between the two
theodicies that recent scholars have brought to our attention. Yet I have argued
that from Arnau Id's perspective - a not insignificant perspective when it comes
to issues of grace and God's modus operandi - these differences, however real,
are outweighed by some essential similarities. The overriding issue for Amauld
is safeguarding God' s omnipotence, which, to his mind, both Leibniz and
Malebranche fail to do in similar ways. 56
Notes
See Leibniz' s letter to Count Ernst von Hessen-Rheinfels, 1 2 April 1 686: 'Je ne
scay que dire a la lettre de M.A., et je n'aurois jamais cru qu'une personne dont la
160 Steven Nadler
reputation est si grande et si veritable et dont nous avons de si belles Relexions de

Morale et de Logique, iroit si viste dans ses jugemens; apres cela je ne m 'etonne
plus, si quelques uns se sont emportes contre luy' : G II, I 6. See also the letter of
the same date, where Leibniz says that ' I am not surprised now that he has so
easily fallen out with Father Malebranche and others who used to be his fast
friends': G II, 22.
2 Traite, Discourse I, article 12, in OC 5, 27.
3 Ibid., articles 13 and 14, OC 5, 28-30.
4 Entretiens sur la metaphysique (hereinafter, Entretiens), IX.9, OC 12, 2 1 1. The
translation is from Dialogues on Metaphysics, Willis Doney, tr. (New York: Abaris
Books 1980), 2 1 1 (hereinafter, D).
5 Entretiens, IX.9, OC 12, 2 1 1; D 2 1 1.
6 Traite 1. 13, OC 5, 28.
7 Ibid., 1.22, OC 5, 35 .
8 Ibid., 1. 13, OC 5, 28.
9 Ibid., 1.2 1, OC 5, 34.
I O For the discussion of God's ways in the realm of grace, see ibid., discourses II and
III.
1 1 See, for example, Leibniz's Theodicee, Essay I I,. § 204, G IV, 238; and his letter to
Malebranche, 22 June/2 July 1679, G I, 337.
12 See Discours de la metaphysique, § § 3-7; and 'On the Radical Origination of
Things, ' G VII, 303-4.
13 Discours, § 5 .
1 4 Theodicee, I I, §2 1 1, G VI, 244.
15 Discours, §7.
16 ' Leibnizian Optimism, ' Journal of Philosophy 80 ( 1983), 765-83; and Leibniz 's
Metaphysics: A Historical and Comparative Study (Princeton, NJ: Princeton
University Press 1989), ch. 8.
17 Leibniz and Arnauld: A Commentary on Their Correspondence (New Haven, CT:
Yale University Press 1990), 43-7.
18 A.-R. Ndiaye, La Philosophie d 'A ntoine Arnauld (Paris: J. Vrin 199 1), 22 1.
19 ' Leibnizian Optimism, ' 774--6.
20 Leibniz and Arnauld, 45.
21 See Traite, I. I, OC 5, 12-15.
22 Rutherford, 'Natures, Laws, and Miracles, ' in Steven Nadler, ed., Causation in
Early Modern Philosophy (University Park: Pennsylvania State University Press
1993), 156.
23 Ibid.
24 Theodicee, II, §208, G VI, 24 1.
25 Malebranche to Leibniz, 14 December 17 1 1, in Andre Robinet, ed., Malebranche
et Leibniz: Relations persone/les (Paris: J. Vrin 1955), 4 17.
26 Leibniz to Malebranche, January 1 7 1 2, in ibid., 4 1 8.

27 Then again, as Donald Rutherford has suggested to me, it could be that Leibniz
has particular motives for exaggerating their agreement - for example, to win
Malebranche' s approval.
28 Traite 1.3 8-9, additions, OC 5, 47. See also Traite 1. 1 3, OC 5, 28. As far as I
know, Sleigh is the only commentator who picks up on this significant difference;
see Leibniz and Arnauld, 45.
29 For further discussion of this, see Charles Lannore, Modernite et morale (Paris:
Presses Universitaires de France 1 993 ), ch. 5.
30 Leibniz to Count Ernst von Hessen-Rheinfels, 1 / 1 1 February 1 686, G II, 1 2.
3 1 See St Paul, I Timothy 2: 5-6. Jansenius himself gets things going when he attacks
the Pelagian and semi-Pelagian (i.e., Molinist) way of reconciling these two
claims, and Arnauld then spends the rest of his life defending Jansenius.
32 This is why Leibniz had Arnauld sent a summary of his Discours de la
metaphysique in the first place.
33 Reflexions 1.2, OA 39, 1 89-90.
3 4 It is often assumed that Leibniz's Discours was intended for Arnauld as a part of
Leibniz's ecumenical project. For an amplification of this, see Leroy E. Loemker,
'A Note on the Origin and Problem of Leibniz's Discourse of 1 686, ' Journal of
the History of Ideas 8 ( 1 94 7), 449-66.
35 Discours § 5 .
36 Meditations chretiennes VII. 1 3 ; Traite 1.43 .
3 7 Malebranche does, in fact, say that ' I do not agree that there is evil only in
appearance. I think that there is evil, that God pennits it ... , ' Entretien d 'un
philosophe chretien et d 'un philosophe chinois, Avis, OC 1 5, 53 .
3 8 Reflexions 1.2, OA 39, 205 .
3 9 'On the Radical Origination of Things, ' G VII, 306-7.
40 Theodicee, II, § 1 47, G VI, 1 97-8. As Catherine Wilson has pointed out, the
broader perspective needed here is not simply a larger experience of the visible
world, its past and present, but some insight into its hidden dimensions, 'an aspect
of the world which is not directly available to perception' (i.e., the laws governing
it); see her 'Leibnizian Optimism. '
4 1 Traite 1.20, O C 5 , 33.
42 See Reflexions II.26, O A 39, 599: 'The divine will detennines itself, freely and
indifferently, to all things to which it does not have a necessary relation; that is, to
all things outside of God. '
43 Ibid., 1.3, OA 39, 2 1 2. And Malebranche seems to accept this characterization:
' God ' s wisdom renders him impotent, so to speak, since his wisdom obliges him to
act by the simplest means ... ' ( Traite 1.38, addition, OC 5, 47). For a discussion of
this, see Ndiaye, La Philosophie d 'Antoine Arnauld, 230.
44 Traite 1.38, OC 5, 47. See I Tim. 2: 5-6.
162 Steven Nadler
45 Reponse aux Reflexions philosophiques et theologiques, I. I .iv, OC 8, 655. See also

Reponse a une dissertation de Mr. Arnau/d contre un Eclaircissement du Traite de
la nature et de la grace, OC 7, 527: 'Not all of God's volitions are practical.' For a
good discussion of the distinction between simple and practical volitions, see
Sleigh, Leibniz and Arnauld, 154ff.
46 See the article ' Jansenisme' by J. Carreyre in the Dictionnaire de theologie
catho/ique, vol. 8, part I, 3 18-529; and Jean Laporte, La Doctrine de Port-Royal,
vol. 2: Exposition de la doctrine (d 'apres Arnau/d): Les verites de la grace (Paris:
Presses Universitaires de France 1923).
47 Amauld's preferred interpretation is that 'all humans' refers to all kinds of human
beings: kings and peasants, old and young, wise and ignorant, etc. See Reflexions
11.23, OA 39, 572-3.
48 Apologie pour /es Saints Peres, OA 18, 108. Amauld (following Thomas) does
grant the distinction in God between antecedent volitions and consequent
volitions, where antecedent volitions are basically what Malebranche calls 'simple
volitions'; see Reflexions I I.23, OA 39, 572ff. Thus it looks as though even
Amauld recognizes non-efficacious volitions in God. Yet he also insists that such
antecedent volitions are not real volitions; rather, he claims that they are volitions
only metaphorically, being mere ve//eites; see Apologie pour /es Saints Peres, OA
18, 1 1 1.
49 Reflexions 1.2, OA 39, 190: ' II suffit que [Dieu] veuille afin que ses volontes soient
executes.'
50 Theodicee II, §367, G VI, 333.
5 1 Discours §3. See Ndiaye, La Phi/osophie d 'A ntoine Arnau/d, 239-40.
52 Leibniz and Arnau/d, 46-7.
53 In fact, the specific concern for safeguarding God's freedom that Amauld
demonstrates in criticizing Leibniz's account of substance in their correspondence
is framed in almost exactly the same terms as it appears a year earlier in his attack
on Malebranche's theodicy. In the Reflexions (11.26, OA 39, 599), Amauld notes
that, for Malebranche, ' God must produce the most perfect [world]. And the
design being formed, he is no longer free to choose the means by which he will
execute it: for he necessarily chooses general ways, which are the most worthy of
his wisdom, his greatness, and his goodness. Thus, the design being taken, and the
means fixed, he [Malebranche] often calls what follows from this "the necessary
consequences of the general laws". God is thus free only in having willed to create
something, but everything else has been the effect of a more than Stoic fatality,
with the exception of the miracles he brings about by means of particular
volitions.' As Sleigh notes (Leibniz and Arnau/d, 46), Amauld wrote this passage
just before receiving the outline of the Discours from Leibniz. Compare it with the
following from Amauld's letter to Leibniz of 13 May 1686: "'The individual
concept of each person involves once and for all, all that will ever happen to him"
. . . Whence I thought that we could infer that God was free, in so far as the creating
or not creating of Adam, but supposing that he had wished to create him, all that
has since happened to the human race has come and must come by a fatal
necessity . . . ' (G II, 27). Sleigh ' s comment seems apt: 'The necessity has gone from
stoical to fatal, its source is different. but from Arnauld's point of view the same
unacceptable consequence is involved in Leibniz's scheme as in Malebranche's'
(Leibni= and Arnauld. 46).
54 See note 49. above.
55 Theodicee I, §22. G VI, 1 1 5- 1 6, emphasis added.
56 I am very grateful to Vincent Carraud, Charles Larmore, Don Rutherford, and
Catherine Wilson for their helpful comments on an earlier draft of this paper; and
to audiences at the University of Toronto and the Sorbonne (both commemorative
gatherings on the occasion of the three-hundredth anniversary of Arnauld's death)
for their questions and suggestions.
11
Arnauld on Efficacious Grace
and Free Choice
RO B E RT C. S L E I G H, J R
In a letter written on 21 June 1692, 1 Amauld advised those interested i n assess

ing his views on freedom that they ought to examine what he had written on the
subject in the last seven or eight years (i.e., since 1684 ), rather than what he
wrote earlier, in his apologies for Jansen. With respect to the latter, he noted
that he was then obliged to defend Jansen. In the same letter, Amauld suggested
that his later views on freedom resulted from a careful examination of St Thomas's
writings relevant to freedom, and, in particular, from his having noted the re
spects in which the views expressed in Thomas's later writings are superior to
those expressed in his earlier writings.2 In concluding the letter of 21 June 1692,
Amauld listed five advantages of the account of freedom to be found in St
Thomas's later works, one of which is that the account offered makes it easy to
reconcile the efficaciousness of grace with freedom. 3 A major aim of this essay
is to specify what forms of compatibilism are consistent with Amauld's mature
account of freedom. This attempt involves an examination of Amauld's mature
thought about freedom and the nature of efficacious grace. In order to keep track
of the forms of compatibilism Amauld accepted, I contrast Amauld's position
with a Catholic position that is as incompatibilist as the faith allows. Actually,
Malebranche's position would do, but his central metaphysical tenet - that God
is the only true cause - produces static on the line. So I use the Molinist position
for contrast.
In preparing these remarks, I have had the advantage of studying two items
from the secondary literature that are models of good work in the history of
philosophy: Jean Laporte's book on Amauld on grace, and Elmar Kremer's
recent article 'Grace and Free Will in Amauld.'4 My remarks are really no more
Amauld on Efficacious Grace and Free Choice 165
than footnotes to those two studies. With Laporte and Kremer as predecessors,
definitive results are a reasonable expectation; but none is forthcoming. In their
place, I record various doubts and hesitations, accompanied by divagations on
metahistory, in this case on why it is so hard to reach firm conclusions about
what an author, now dead three hundred years, meant by what he said, even
when that author wrote clearly and employed those clear writings to express the
thoughts of one of the sharpest minds of a period in the history of philosophy
noted for sharp minds.
In section I, I record various theses concerning efficacious grace and free
choice to which Amauld was committed in his mature period. Then, I highlight
the controversial aspects of the theses recorded, in order to prepare the way for
discussion of the primary topics of the essay. In section II, I discuss the basic
change in Amauld's conception of efficacious grace. And, in section III, I con
sider features of the Thomistic conception of freedom to which Amauld com
mitted himself after 1684, and their relation to various forms of compatibilism.
On the score of grace, Amauld seems to have held the following in his mature
period:
a) That fallen man is incapable of performing supernaturally meritorious

actions unaided by grace;
b) That efficacious grace is efficacious in virtue of its own intrinsic nature -
' par elle-meme,' as Amauld often put the matter;
c) That an action is supernaturally meritorious only if it is free;
d) That an action� s being done under the influence of efficacious grace is
consistent with its being done freely;
e) That the freedom with which efficacious grace is consistent, and which is
required for supernatural merit, requires, in tum, that the agent could
have done otherwise; and
t) That no operative actual grace need be posited other than efficacious grace,
even in fallen man, i.e., no merely sufficient grace need be posited, even
though fallen man lacks the capacity to perform supernaturally meritori
ous actions unaided by grace. 5
Much of Amauld's writing on grace is devoted to sustaining (f) against, for

example, the neo-Thomists and Nicole. 6 I plan to discuss Amauld on thesis (f)
in a separate paper; here it is not considered. I believe that Amauld would claim
166 Robert C. Sleigh, Jr
that theses (a), (c), (d), and (e) are defide for Catholics and, hence, non-negoti
able. Thus, according to Arnauld, the only thesis from (a) through (e) that is up
for serious philosophical discussion is (b). 7 Here is a remark on ( e) that has
obvious applications with respect to the other theses: what the agent could have
done otherwise is elicit a volition. There is no effort here to purchase freedom
on the cheap by arguing that freedom fundamentally applies to first-order bod
ily actions that the agent could have done otherwise, meaning no more than that
the agent would have done otherwise had the agent so chosen. 8
(b) is where the seas of philosophical theology rise high. I approach it with
considerable trepidation. There is a primitive notion that is helpful here -
the idea that each actual (as opposed to habitual) grace that occurs is aimed at
contributing to bringing about some volition v in some agent a. The admittedly
imprecise locution 'is aimed at' is employed rather than, for example, 'is
intended as, ' because what God intends absolutely (all things considered) comes
about. So, by 'is aimed at' I mean something like 'is intended to contribute
to . . . , other things being equal ( which they may not be). '
Consider the following two propositions:
I) Necessarily, for any grace g, volition v, and agent a, if grace g aims

at volition v in agent a, and grace g occurs and is efficacious, then a
elicits v.
2) For any grace g, volition v, and agent a, if grace g aims at volition v in
agent a, and grace g is efficacious, then necessarily, if g occurs and agent
a exists, then a elicits v.
The first is a proposition that all parties to these disputes would accept; its
necessity is a simple consequence of the rather humdrum fact that 'efficacious'
is a success-adjective. The second is where the action is. Molinists ( e.g.
Malebranche) reject it; Thomists, neo-Thomists, and Arnauld in every one of
his phases accept it. I focus the debate somewhat by noting one thing that is not
at stake. All sides would agree that, if grace g aims at v in a, then, necessarily, g
aims at v in a. So the issue boils down to the truth-value of the following:
3) If grace g is efficacious, then, necessarily, grace g is efficacious.
Molinists reject (3), the others mentioned accept it, and, then, conjoining it with
other items not in dispute, derive (2). This third proposition is a way of captur
ing the claim (b), i.e., that efficacious grace is efficacious in virtue of its own
intrinsic nature - 'par elle-meme' - as Arnauld was wont to put it.
It is important to have a grasp of what Arnauld took to be the main enemy
here. It is the Molinist claim that whether a grace supplied by God to agent a,
Arnauld on Efficacious Grace and Free Choice 167
aimed at volition v, is or is not efficacious, is ultimately up to a and not to God.

If a freely elicits v, then g is efficacious; otherwise not. Furthermore, whether a
freely elicits v under the influence of g is ultimately up to a in the sense that
there is nothing God can do that is either metaphysically or causally or naturally
sufficient for a's freely eliciting v. Of course, it is within God's power to bring
it about that a elicits v, but it is a's freely eliciting v that matters here. And, of
course, God has it in his power to choose circumstances, both external and
internal to a, that will contribute towards a' s freely eliciting v when under the
influence of g. None the less, it is crucial to see how radical the Molinist posi
tion is. It has these consequences: in accounting for the ultimate sufficient con
ditions of the events that occur in a world that contains creatures who elicit free
choices, you must mention those choices as well as God's. Put another way:
there are possible worlds that God can not create. There are mainline, tradi
tional conceptions of theological determinism, theological compatibilism, and
predestination with which these consequences are incompatible.
Authors often claim that the relevant conceptions of theological determin
ism, theological compatibilism, and predestination are de fide for Catholics.
But that is rash. The Congregatio de auxiliis ended with a formal decree pro
claiming, among other things, that the relevant aspects of Molinism are not
incompatible with anything that is de fide for Catholics. 9 I believe that Arnauld
had these points in mind in contrasting his position with Malebranche's: he
wrote:
The grace that I maintain as the foundation of gratuitous predestination, and that I
claim has the consequence that the merits of the Saints are the gifts of God, is not
a grace of the sort that some theologians imagine it to be, which has, or lacks, effect
according to whether it is agreeable to the will; rather, it is grace that is efficacious
par el/e-meme; i . e., that does not rely on our willing, but that brings it about that
we will. 1 0
Molinism may offer a less than robust notion of predestination, but it clearly
has advantages, in virtue of its denial of (3), with respect to establishing the
consistency of (a) through (e). By contrast the affirmation of (3) generates a
problem for Arnauld and his ilk. Consider the following, which is a conse
quence of (2):
4) Let g be some efficacious grace that aims at volition v in agent a; it is not

possible that g occurs and a does not elicit v.
Isn't (4) inconsistent with (e)? Of course, as a chorus, those non-Molinists who
accept (4) say no. To the rescue comes the notorious distinction between the
composed and the divided senses. It is sometimes explained in the following

way. Consider:
5) Let g be some efficacious grace that aims at v in agent a and let g occur;
nonetheless, it is possible that a does not elicit v.
According to the explanation I am now considering, (4) affinns of a certain

compound state of affairs that it can not obtain - the composed sense; whereas
(5) says that, even if one component of this compound state of affairs does
obtain, it remains possible that the other component does as well - the divided
sense. The explanation grants that in virtue of affinning (3) the non-Molinist
is saddled with (2), and, hence, ( 4), but claims that disaster does not ensue
because (5) is all that is required in order to establish the consistency of (a)
through (e).
It is primarily a defence of (e) that requires this convoluted escape mecha
nism. In turn, (e) is required by the following from the Council of Trent: 'If
someone says that the free choice of man, moved and excited by God, in no way
cooperates with the exciting call from God, by an assent in virtue of which man
disposes and prepares himself to obtain the grace of justification; and that man
can not dissent, even if he wills to, .. . let him be anathema. ' 1 1 Suppose we take
the 'possible' in (5) as standing for metaphysical possibility. Then, surely, it
crosses the mind that (5) is simply no match for this robust canon from the
Council of Trent.
I believe that Arnauld did not rely on anything so tepid as (5); that, in place of
(5), he offered the following elongation of (4):
6) Let g be some efficacious grace that aims at volition v in agent a; it is not

possible that g occurs and a does not elicit v. But it is possible that g
occurs and a retains the power not to elicit v and also the power to elicit
some volition v' other than v.
Focusing on what (6) says is impossible and what (6) says is possible yields a
more robust and useful contrast between the composed and the divided senses.
It also introduces us to Arnauld' s convoluted thought on the topic of power, a
topic beyond the scope of this essay.
II
In a letter to Nicole (28 August 1685) Arnauld noted that there are three
non-Molinist, but Catholic candidates for the role of efficacious grace: the phy
sical predeterminations (predetermination physique) of the neo-Thomists ;
the victorious pleasure (delectation victorieuse) of the Jansenists; and the

inspiration of love (inspiration d 'amour) of Estius, which Arnauld attributed to
Thomas, and which Arnauld favoured in his later phase. 1 2 For ease of reference,
I identify each candidate via a champion of that candidate noted as such by
Arnauld: physical predetermination, with Alvarez; victorious pleasure, with
Jansen; and the divine inspiration of love, with Estius.
In a letter of July 1693, to Bossuet, Arnauld claimed that there are two ac
counts of actual grace, one of which identifies actual efficacious grace with the
combination of God's mercy in the form of a divine volition concerning creature
a and a resulting form inherent in the soul of a; the other of which identifies
actual efficacious grace solely with God's mercy in the form of a divine volition
concerning creature a. 1 3 In the letter to Bossuet, Arnauld referred to an early
work of his, the Dissertatio Theologica, published in 1656, which contains a
discussion of these issues. 1 4
I think that the difference in count here - are there two theories or three? - is
of no consequence. By Amauld's lights there are two fundamental distinctions
to note. All the relevant theories of grace claim that a given case of actual effica
cious grace involves at least these two items: first, a volition on God's part
concern ing a free motion (volition) of some creature's will; second, the free
motion (volition) of that creature's will. The first fundamental distinction is
between those theories that posit some intermediary entity in a's soul (intellect
or will) that serves to execute God's volition and those that do not. In their
theories, Alvarez, Jansen, and even Molina postulate such an intermediary en
tity. In the cases of Molina and Jansen, the alleged intermediary entities func
tion like any other motive for choice in a human agent, i.e., they make a causal
contribution to choice along with, and often in competition with, other second
ary causal contributors. The second fundamental distinction concerns the type
of determination (or lack thereof) that holds in the theory between the item
identified with grace and the resulting free choice. In Jansen's theory, the deter
mination involved seems close to a causal determination; in the theories of
Alvarez and Estius, the determination seems to be metaphysical but not causal;
and, in Molina's theory, there is no determination in the relevant sense at all.
Amauld had a variety of reasons for preferring the theory of Estius to its
competitors. 1 5 I take note of a pair of those reasons. Amauld claimed that, on
Estius's view, ' it is much easier to explain the efficaciousness of grace and to
reconcile this efficaciousness with freedom.' 16 The reconciliatory point is two
fold: (i) prima facie, there is a difficulty in explaining how a created entity,
e.g., a victorious pleasure, operationg on the will without benefit of the agent's
rational deliberation, can bring it about that the agent elicits a choice without
harming the will ' s freedom; and (ii) prima facie, there is a difficulty in
explaining how the occurrence of a created entity in the will, distinct from the
1 70 Robert C. Sleigh, Jr
elicitied choice of the agent, can metaphysically detennine that the agent elicits
a specific choice. The first difficulty applies to Jansen's account; the second, to
Alvarez' s account.
On the positive side, there is a clear advantage to Estius's view with respect
to (ii). It is metaphysically detennined that if God wills p with a consequent
will, then p. So, if each occurrence of efficacious grace is identified with some
consequent willing of God, there is no problem in explaining the resulting
metaphysical detennination. But does this account not exacerbate Arnauld's
problem with respect to (i)? True, on Estius's view, efficacious grace does not
involve a created entity operating on the will without benefit of the agent's
rational deliberation. Still, it involves an uncreated entity, God's will, which is
not subject to the agent's deliberative powers, apparently metaphysically deter
mining that the agent elicit a specific choice.
What Arnauld wrote on this point is standard fare. For example, in his In
struction sur I 'accord de la grace avec la /iberte, 1 7 Amauld argued as follows:
(a) we know, in virtue of God's omnipotence, that he has the power to bring it
about through an exercise of his consequent will that a created agent freely
elicits a specific choice; (b) we know that it is metaphysically necessary that
what God consequently wills occurs; and ( c) we know that we are free. This line
of reasoning is weak. We can skip (b) and (c); (a) is the whole ballgame. The
issue is: is the state of affairs consisting in God's bringing it about that an agent
freely elicitis a specific choice possible? Molina thought not; Amauld simply
asserted that it is in this text.
Perhaps I am being unfair. Amauld often wrote about God's will determining
a created agent's will, not about God's will bringing it about that the agent
elicits a specific choice. Elmar Kremer, in the essay mentioned above, notes that
there are divine detenninations that metaphysically detennine an agent's choice
without threatening freedom, e.g., divine foreknowledge. We need another in
tuitive notion here. Philosophers in the seventeenth century recognized various
varieties of detennination: semantical, in virture of a true prediction; epistemo
logical, in virture of someone' s foreknowledge; and, of course, causal determi
nation, among others. Some of these were seen as bringing about the item they
determined; others were not. I call the former 'the category of quasi-causal de
termination.' The question to be answered is this: Did Amauld have a rational
basis for claiming that, although divine efficacious grace metaphysically deter
mines the state of affairs it aims at, it does not involve quasi-causal determina
tion? I do not know the answer to that question; if Amauld had such a rational
basis, he was a master at hiding it.
I have some confidence on one point here: If Amauld had such a basis, it
turned essentially on features peculiar to grace and free choice. It was not a
general denial such as the following principle involves:
7) For any x and y, if x metaphysical ly necessitates y, then it is false that x

causally necessitates y.
Suppose Arnauld had been attracted to (7). He knew that Malebranche was
attached to the following:
8) For any x and y, if x causally necessitates y, then x metaphysical ly

necessitates y.
Now from (7) and (8) we can deduce:
9) It is false that there is an x and y such that x causally necessitates y.
So had Arnauld accepted (7), he would have had the tools to establish that
Malebranche's fundamental thesis concerning causality, i.e., (8), has the conse
quence that there are no causal connections. Had argumentative, abrasive Arnauld
possessed those tools, he would have applied them. But, to the best ofmy knowl
edge, there is no such argument in Arnauld's lengthy exchange with Malebranche.
But that must be because Arnauld rejected (7).
III
I have noted that Arnauld accepted a version of theological compatibilism that

Molina rejected. In this section I consider a fonn of compatibilism - infallible
determination compatibilism; Arnauld accepted it and Molina rejected it. I then
consider, utterly indecisively, whether infallible-detennination compatibilism
and causal compatibilism amount to the same thing.
As noted previously, for Arnauld, the eliciting of a volition by a human agent
is the central item in a proper conception of human freedom. Hence, a relevant
form of causal compatibilism would be the following:
I 0) There is some human agent a, volition v, and time t such that: (i) a
elicits v at t; and (ii) there are conditions k, ... k" that obtain at or prior
to t and that are causally sufficient for a's eliciting v at t; and (iii) a is
free in eliciting v at t.
So, if Amauld rejected causal compatibilism, he denied that ( I 0) is so much as

possible. Well, did he? I do not know. Part of the point of this section is to
provide some excuse for my ignorance.
My concern is with the relationships among three concepts of determination
- natural determination (and, correspondingly, natural necessity); infallible
determination; at last, causal determination (and, correspondingly, causal

necessity). In his mature phase, Amauld held that natural determination (natu
ral necessity) is incompatible with freedom, whereas infallible determination is
compatible with freedom. Taking 'causal detennination' ('causal necessity') to
refer to our concept of causal determination (causal necessity), my problem is
how natural determination, infallible determination, and causal determination
are related, according to Amauld. My claims are these: (i) the texts make clear
that if Amauld's natural determination is our causal determination, then Amauld
denied the possibility of ( IO); and (ii) the texts do not settle the matter as to
whether Amauld's natural determination is our causal determination.
Here is a useful text for commencing discussion on this topic. In his Instruc
tions sur I 'accord de la grace avec la liberte, Amauld, intending to help the
reader understand that an act can be free and yet done under the influence of
efficacious grace, wrote: 'One has only to conceive properly what freedom is,
and to get rid of the false idea that many have of it, who imagine that a person
can only freely will something when that person is not determined to will it' ... 1 8
Amauld then went on to note various cases of what sound like psychological
causation that involve free choices, e.g., a prostitute who is infallibly deter
mined to commit a sin freely in virtue of aspects of her reaction to the sum
offered for her services. 1 9 After a number of such examples, the interlocutor asks
if there is any necessity and any determination that is incompatible with free
dom. Amauld answered that there is indeed one such case: 'It is the case of
natural necessity, or determination to one thing by nature. For our soul would
not be free, if it were naturally determined' ...20 He went on to add that, in the
case of natural determination, the agent is not master of his own action, and the
agent lacks potentia ad opposita, the power to do (choose) otherwise.2 1
Here are some remarks on the two passages quoted above. In the first passage,
Amauld claimed that, in order to conceive properly what freedom is, we must
rid ourselves of a widely held, but false idea - namely, that a person can elicit a
choice freely only when no variety of determination yielding the aforementioned
eliciting, i.e., no quasi-causal determination, is present. I believe that the posi
tion Amauld asked us to forgo here is the Molinist position. According to the
Molinist position, various conditions, both internal and external to the eliciting
agent, are germane to what choice the agent elicits, and, in fact, control over
those conditions is, according to Molina, exactly how divine providence is exer
cised with respect to the free choice of creatures. This germaneness may be
construed as a kind of counter-factual sufficiency that both Amauld and Molina
would view as something less than a determination - at any rate, something less
than a quasi-causal determination. In the examples accompanying the first
passage quoted, Amauld assumed that what we might call 'infallible sufficiency�
is a kind of quasi-causal determination, and, hence, is a stronger connection
Arnauld on Efficacious Grace and Free Choice 1 73
than mere counter-factural sufficiency and is compatible with freedom. Let

( 1 O') be the result of replacing ' causally sufficient' with 'infallibly sufficient' in
( I 0). Thus, ( I O') formulates a version of compatibilism that Arnauld accepted
and that he believed Molina rejected. Noting this fact is what first led me to
consider the hypothesis that in his mature phase Arnauld may have accepted
( I 0) after all.
Let ( 1 O") be the result of replacing 'causally sufficient' in ( 1 0) with 'naturally
sufficient. ' The second passage quoted above shows that Arnauld rejected ( I O"),
and, of course, we know that Molina got off the compatibilism bus long before
stop ( I O"). In general, a philosopher gets off the compatibilism bus as soon as
that philosopher reaches a level of determination that he or she views as depriv
ing the agent of potentia ad opposita, the power to choose otherwise. Molina's
view, at least as Arnauld construed it, was that any quasi-causal determination
leads to the relevant deprivation; Arnauld believed that infallible determination
leaves the agent in possession of potentia ad opposita, whereas a natural deter
mination does not.
Arnauld, intending to follow St Thomas here, took natural necessity (or the
lack thereof) to apply to the exercise of a power ( active or passive) by an agent in
specified circumstances. There is considerable plausibility to the following claim:
in the case of non-intelligent agents, the assertion that such an exercise occurs
as a natural necessity is tantamount to the assertion that the state of affairs that
obtains as a result of that exercise is causally necessary in the specified circum
stances. My problem concerns how to understand Arnauld's talk about natural
necessity in the case of exercises of powers peculiar to intelligent agents, i.e.,
agents with intellects and rational wills. According to Arnauld, again intending
to follow St Thomas, when an intellect exercises its powers and comes to know
a self-evident proposition, it does so as a natural necessity; in every other case,
exercise of an intellectual power does not occur as a natural necessity. And
when an intelligent agent wills something (its own happiness, as it turns out)
that would utterly satisfy the agent, were it possessed, then, and only then, does
the agent will as a natural necessity. My radical proposal is this: what we are
getting here are extensions of the notion of natural necessity from its home base
in non-intelligent agents; these extensions are in some measure stipulative. Hence,
to say that a given volition v elicited by agent a does not occur as a result of a
natural necessity just amounts to saying that even were v fulfilled, agent a would
not be utterly satisfied. But that seems consistent with saying that, in the cir
cumstances then obtaining, it was causally necessary that a elicit v - that, for
example, a' s high moral character, plus his beliefs about the circumstances then
prevailing, causally necessitated that he elicit v. And that, according to my radi
cal proposal, is exactly the content of the remark Arnauld might have made
about a's situation, namely, that a was infallibly determined to elicit v.
My situation is this: I cannot convince myself that this radical proposal is

false. There are numerous objections to it; I close by noting one. Elmar Kremer
has drawn my attention to passages in the Reflexions in which Amauld, in his
mature period, stated that the thesis (which he attributed to Malebranche) that
all human choices are 'a necessary consequence of natural laws ... ' involves the
mistake of Wyclif, since it deprives humans of freedom. 22 This surely sounds
like a denial of causal compatibilism. It is, if Amauld means by 'natural law'
what we mean by it. However, there is the possibility that careful examination
would convince us that in Amauld's system the more basic notion is natural
necessity, with natural law the derived notion, so that Amauld understood a
natural law as a universally quantified conditional that holds as a natural neces
sity. In that case we would be back to square one. And, in fact, the nearest
preceding passage in the Reflexions in which Amauld commented in detail on
the character of laws of nature is this:
General laws of nature include only a certain measure of motion, imprinted on all
matter, and the rules of the communication of motion, by which a body which
collides with another, detennines a different motion in it, either by communicating
a new motion to it, or by preventing it from continuing with the motion it had. 23
Only a materialist would suppose that choices fell under laws of nature, so con
strued; only the strictest sort of epiphenomenalism would yield determinations
of choices under laws of nature, so construed.
To this point, my investigation suggests that Amauld was committed to at
least two forms of compatibilism rejected by Molina - theological compatibilism
and infallible-determination compatibilism. In fact there is some reason to
suppose that Amauld would insist that there is only one form of compatibilism
here, under two descriptions. In describing Amauld' s views about the relation
of efficacious grace to the will under the influence, I have allowed myself the
woolly term ' metaphysical determination. ' But when we examine Amauld' s
efforts to explain the consistency of determination of the will by efficacious
grace with freedom of the will so determined, the determination begins to
look like infallible determination. Still, the really interesting question - how
infallible - determination compatibilism and causal compatibilism are related -
remains beyond my grasp.
Notes
I OA 3, 497-8.
2 Amauld did not have access to De Malo, with its important question on free
choice.
Amauld on Efficacious Grace and Free Choice 175
3 The other alleged advantages are as follows: St Thomas's mature view of freedom
is perfectly consistent: the authority of Thomas renders it above suspicion; it
explains why freedom from necessity, and not just freedom from coersion, is
required for merit and demerit: and, lastly, it explains just what is right, and what
is wrong, about St Bernard's famous remark 'Ubi voluntas, ubi libertas. '
4 Jean Laporte, La Doctine de Port-Royal. Tome deuxieme: Exposition de la doctrine
(d 'apres Arnauld). I - Les Verites de la grace (Paris: Presses Universitaires de
France 1 923) : and Elmar Kremer, ' Grace and Free Will in Arnauld,' in The Great
Arnauld and Some of His Philosophical Correspondents, Elmar Kremer, ed.
(Toronto: University of Toronto, Press 1 994), 2 1 9-39.
5 See for example Instruction sur la grace, OA 1 0, 40 1-34.
6 See for example Ecrit du pouvoir physique, OA 1 0, 48 1 -530.
7 See for example OA 39, 9 1 -2.
8 See for example " Humanae Libertatis Notio, ' in Causa Arnaldina, Pasquier
Quesnel, ed. ( apud Hoyouxi: Leodici Eburonium 1 699), 99-1 1 1 . There is a French
translation of this work by Quesnel in OA 1 0, 6 1 4-24. Elmar Kremer kindly
supplied me with a copy of the Latin version.
9 I take it that at OA 39, 9 1-2, Arnauld agrees.
1 0 OA 39, 68.
1 1 Council of Trent, Session VI, ch. 4. In this quotation, the expression 'exciting call'
refers to actual (prevenient) grace, which is the grace that concerns us, the grace
that aids us in perfonning supernaturally meritorious actions. The grace of
justification is an item that occurs, if at all, much later in the process, the end
result of which, when things work out, is salvation.
1 2 OA 2, 5 58-9.
1 3 OA 3, 664.
1 4 OA 20, 232-39.
1 5 See OA 2, 5 58-9; OA 3, 573-4, 578, 636, 664; OA 1 0, 6 1 6, 620: OA 20, 232-9.
1 6 OA 3, 664.
1 7 OA 1 0, 436.
18 Ibid.
1 9 Instruction sur la grace, OA 1 0, 437.
20 Ibid., OA 1 0, 438.
2 1 Ibid.
22 OA 39, 3 1 6 (cf. OA 39, 30 1 ).
23 OA 39, 258.
Notes on Contributors
Jill Vance Buroker is Professor of Philosophy at California State University

at San Bernardino, and Adjunct Professor of Philosophy at Claremont Graduate
School. Her publications include Space and Incongruence: The Origin of Kant 's
Idealism (Dordrecht and Boston: D. Reidel 1981) and a translation of the Port
Royal Logic (Cambridge: Cambridge University Press 1996) as well as articles
on Kant and on Cartesian philosophy, including 'Judgment and Predication in
the Port-Royal Logic,' in The Great Arnauld and Some of His Philosophical
Correspondents, Elmar J. Kremer, ed. (Toronto: University of Toronto Press
1994).
Vincent Carraud is Maitre de conferences en philosophie at the Universite de

Caen. His publications include Pascal et la philosophie (Paris: PUF, coll.
'Epimethee, ' 1992) and (in collaboration with F. de Buzon) Descartes et /es
'Principia ' II: corps et mouvement (Paris: PUF, coll. 'Philosophie,' 1994), as
well as many articles in modem philosophy, including 'Amauld from Ockhamism
to Cartesianism,' in Descartes and His Contemporaries, R. Ariew and M. Grene,
eds. (Chicago: University of Chicago Press 1995).
G raeme Hunter is Associate Professor of Philosophy at the University of

Ottawa. His publications include the Leibniz-Lexicon (Olms: Hildesheim 1988)
and 'The Phantom of Jansenism in the Amauld-Leibniz Correspondence,' in
The Great A rnauld ...
178 Notes on Contributors
Elmar J. Kremer is Associate Professor of Philosophy at the University of

Toronto. His publications include a translation of Amauld's On True and False
Ideas and New Objections to Descartes ' Meditations with Descartes ' Replies
(Queenston: Edwin Mellen 1990); and two articles in The Great Arnauld ...
Thomas M. Lennon is Professor of Philosophy at the University of Western

Ontario. His publications include a translation (with Paul J. Olscamp) and philo
sophical commentary of Malebranche's The Search after Truth (Columbus: Ohio
State University Press 1980) and The Battle of Gods and Giants: The Legacies
of Descartes and Gassendi, 1 665-1 715 (Princeton, NJ: Princeton University
Press 1993).
Steven Nadler is Associate Professor of Philosophy at the University of

Wisconsin (Madison). His publications include Arnauld and the Cartesian Phi
losophy of Ideas (Princeton, NJ: Princeton University Press 1989); Malebranche
and Ideas (Oxford: Oxford University Press 1992); and many articles on early
modem philosophy, including 'Malebranche's Theory of Perception,' in The
Great Arnauld ... He is currently working on a biography of Spinoza.
Aloyse-Raymond Ndiaye is Dean of the Faculty of Letters and Human Sci

ences at the University of Dakar. He is currently serving as the director of the
Fonds International de Cooperation Universitaires, Montreal. He is a member
of the steering committee of the International Federation of the Societies of
Philosophy. His publications include La Philosophie d 'Antoine Arnauld (Paris:
J. Vrin 1991).
Alan Nelson is Associate Professor of Philosophy at the University of Cali

fornia, Irvine. His publications include ' Are Economic Kinds Natural?' in
Minnesota Studies in the Philosophy of Science 14 (1991): 'Social Science and
the Mental,' in Midwest Studies in Philosophy 15 (1991) ; and 'Cartesian
Actualism in the Leibniz-Arnauld Correspondence,' in Canadian Journal of
Philosophy 23 (1993).
Peter A. Schouls, formerly Professor of Philosophy at the University of Alberta,

is Head of Massey University, New Zealand. His publications include Reasoned
Freedom: John Locke and Enlightenment (Ithaca, NY: Cornell University Press
1992) and Descartes and the Enlightenment (Edinburgh: Edinburgh University
Press; and Montreal and Kingston: McGill-Queen's University Press 1989), as
well as many articles on modem philosophy.
Notes on Contributors 179
Robert C. Sleigh, Jr, is Professor of Philosophy at the University of

Massachusetts (Amherst). His publications include Leibniz and Arnauld:
A Commentary on Their Correspondence (New Haven, CT: Yale University
Press 1990) and an edition and translation of Leibniz forthcoming from Yale
University Press.
Jean-Luc Solere is a member of the Centre National de la Recherche Scientifique,

Paris. His publications include the sections on Guillaume d' Auxerre and Durand
de Saint-Poun;ain in La Puissance et son ombre, de Pierre Lombard a Luther,
sous la direction d' Olivier Boulnois (Paris: Aubier 1994).
Index
Abelard, Peter, 9 1, 92, 105 philosophiques et theologiques sur le

abstraction, 4, 23, 26 nouveau systeme de la nature et de la
Alexander VII, Pope, 53, 58, 59 grace, 54, 62, 94-6, 98, 105, 106,
Alvarez, Diego, 169, 170 107, 108, 1 10, 143, 15 1, 153, 162,
Aquinas, Thomas, 38, 46, 7 1, 83, 84, 87, 174
89, 90, 92, 98- 100, 105, 107, 108, Augustine, St, 35-4 1, 43, 44, 46, 47, 48,
1 10, 1 1 1- 13, 124, 125, 127, 130, 49, 53, 58, 66, 70, 73, 82, 83-4, 9 1,
132-5, 140, 142, 143, 144, 155, 158, 96, 104, 107, 1 19-20, 122, 127, 129,
162, 164, 169, 173, 175 130, 132, 133, 139, 140, 147, 155
Arnauld, Antoine: Apologie pour /es
Saints Peres, 162; De l 'autorite des Bayle, Pierre, 5 1-6, 6 1, 62, 63, 146
miracles, 125, 170; Dissertatio Bellannine, Robert-Francis-Romulus, 53
Bipartita, 90, 1 10, 129, 13 1, 132, Bolek, Gerard, 142
144; Dissertatio Theologica, 169; Bossuet, Jacques-Benigne, 92, 104, 105,
Ecrit du pouvoir physique, 90; Fourth 106, 132, 143, 169
Objections, 33, 36, 65, 66, 7 1-4, 75, Burnett, Thomas, 1 13
76, 78, 87, 94-6, 145 ; Humanae
Libertatis Notio, 88, 108, 175; Calvin, John, 58, 1 12, 124
Instruction sur I 'accord de la grace Caterus, 7 1
avec la liberte, 170; Nouveaux Essays clear and distinct ideas, 16, 3 1, 5 1, 129,
de geometrie, 34; On True and False 137
Ideas, 29, 32, 83, 88, 134; Port-Royal compatibilism, 164-74
Logic (The Art of Thinking), 3, 7-9, Condoret, Marie-Jean-Antoine-Nicolas
1 1, 34, 35, 4 7, 48, 50, 59, 62, 63, 1 17, Caritat, Marquis de, 45
1 19, 120, 122, 125, 134; Reg/es du confused ideas, 23, 24, 32
bon sens, 90, 132, 143; Reflexions Congregatio de auxiliis, 167
182 Index
Diderot, Denis, 45 immortality, 66, 77

dualism, 77 imperceptible thoughts, 129, 136--9
Dummett, Michael, 10 incomprehensibility of God, 92--6
Index of Forbidden Books, 44, 142, 144
efficacious grace, 58, 128, 157, 164-74 Innocent X, Pope, 52, 53, 58
Estius, 96, 169, 170 intentionality, 13, 20, 22-3, 28, 54
eternal truths, 5, 7, 45, 64-75, 9 1-1 10, Iroquois, 139
142
Eucharist, 35-7, 58, 77-83, 86, 89, 95 Jansen, Cornelius, 35, 47, 49, 5 1,
Eutyches, l 06 52--63, 1 16, 125, 126, 129, 139, 140,
externalism, 4-5 142, 16 1, 164, 169, 170
Jansenism, 127, 128, 143, 157
Fenelon, Franc;ois de Salignac de la judgment, 3-12, 13-32, 59, 62, 63, 76,
Mothe, 92, 93, 104, 105, 106, 107, 88, 1 19, 130, 13 1, 135, 140
132, 143
free choice, 69, 164-74 Kant, Immanuel, 10, 44, 50; Critique of
freedom of indifference, 96--100, l 04, Pure Reason, 10; Foundations of the
107 Metaphysics of Morals, 50; What Is
Frege, Gottlob, 3-12; On Sense and Enlightenment?, 50
Reference, 1 1; The Thought, 3 Kepler, Johannes, 9 1
Galilei, Galileo, 33, 9 1 Lamy, Franc;ois, 1 10, 127, 13 1-8, 142,

Gassendi, Pierre, 45, 52, 53, 90, 144 143, 144, 146
general grace, 127-38 Leibniz, Gottfried, 34, 46, 64-75,
general volitions, 149-50 9 1- 1 10, 1 13, 124, 137, 147--63
Genesis, Book of, 93 LeMoine, Dean of the Chapter of Vitre.
God as causa sui, 96, 100 80--3, 95, 96
Gouhier, Henri, 74, 94-5, 104, 105, Lombard, Peter, 92, 105
106 Luther, Martin, 90, 105, 1 12, 124
Hessen-Rheinfels, Ernst von, 109, 159, materially false ideas, 13-32, 35, 36
16 1 matter, essence of, 77-83, 86
Hobbes, Thomas, 17, 33, l 13 Malebranche, Nicolas, 5 1, 54, 55, 56,
holy thorn, miracle of, 1 14-17, 120, 62, 64, 66--70, 74, 75, 8 1-3, 86, 89,
12 1-2 92-1 10, 1 13, 122--4, 132, 133, 134,
Hume, David, 12 1-2, 126 136, 137, 143, 144, 146, 147--63, 164,
Huygens, Gommaire, 1 10, 127-3 1, 132, 166, 167, 17 1, 174
139, 142, 143 Mersenne, Marin, 30, 33, 34, 37, 45, 49,
65--6, 72-3, 74, 75, 90, 9 1, 107
illocutionary force, 3-12 Mesland, Denis, 30, 85, 89, 90
Index 1 83
miracles, 82, 108, 1 1 1-26, 149, 160, vacuum, possibility of, 79-80
162 Voltaire, Fran�ois-Marie Arouet de, 45
Molina, Luis de, 169-74 Vos, Philippe de, 142
Molinism, 167
natural determination, 17 1-3

Nestorius, 106
Newton, Isaac, 1 13
Nicole, Pierre, 1 1, 50, 52, 53, 6 1, 89,
1 17, 127-46, 165, 168
objective reality, 13-32, 7 1

obscure ideas, 29
omnipotence, 77, 84-7, 90-104, 148,
154-9, 170
Pascal, Blaise, 35, 43, 49, 50, 5 1, 52,

58--60, 62, 95, 1 14, 1 15, 123, 124,
125, 129
potentia ad opposita, 172, 173
Quesnel, Pasquier, 88, 108, 142, 143,

175
Racine, Jean, 125

Regius, Henri le Roy, 6, 44
representation, 13-32, 10 1
Sartre, J.-P., 145

Scripture, 92-3, 1 12, 123, 135, 154, 157
sensory ideas, 13-32, 47, 90
sign-representation, 20, 22, 23, 27, 3 1
Socinians, 95
secondary causes, 1 1 1- 13, 120, 123, 124
Spinoza, Benedict, 92, 94, 105, 1 13
Steyaert, Martin, 142
Suarez, Francisco, 92
Trent, Council of, 79, 80, 88, 89, 168,

175

Kremer - Interpreting Arnauld (1990)

Uploaded by

Copyright:

Available Formats

Kremer - Interpreting Arnauld (1990)

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Kremer - Interpreting Arnauld (1990)

Uploaded by

Copyright:

Available Formats

Interpreting Arnauld

UNIVERSITY OF TORONTO PRESS

Printed on acid-free paper

Toronto Studies in Philosophy

Canadian Cataloguing in Publication Data

Main entry under title:

(Toronto studies in philosophy)

1. Amauld, Antoine, 1612-1694 - Congresses. I. Kremer, Elmar J. II. Series.

Bl 824.A864l57 1996 194 C95-933240-5

University of Toronto Press acknowledges the financial assistance to its publishing

1 Amauld on Judging and the Will 3

8 Arnau Id's Defence of Miracles and Its Context 111

an earlier version of which appeared in that journal.

The following abbreviations are used throughout this volume.

OA Oeuvres de Messire Antoine Arnauld, edited by G. Du Pac de

AT Oeuvres de Descartes, 12 vols, edited by Charles Adam and Paul

Alquie Oeuvres philosophiques de Descartes, edited by Ferdinand Alquie,

CSMK The Philosophical Writings of Descartes, vols 1 and 2 edited and

oc Oeuvres completes de Malebranche, under the direction of Andre

G G. W Leibniz: Die philosophischen Schriften, edited by C. I. Gerhardt,

JILL VANCE BUROKER

Judgment and Illocutionary Force

its force through recognizing the possibility of using declarative sentences

Judgment in the Port-Royal Logic

Arnauld's conception ofjudgment in the Port-Royal Logic differs from Descartes's

gives it a central focus in the Port-Royal Logic's semantics. What results is a

a word whose principal function is to signify an affirmation, that is, to indicate

affirmation only in so far as it has become an object of thought by mental reflection.

A clear explanation of this aspect of Frege' s theory is available in Michael

University Press 1 98 1 ). See especially ch. 1 0, ' Assertion,' 295-363 .

Arnauld and Some of His Philosophical Correspondents, Elmar J. Kremer, ed.

Since Arnauld's set of Objections to the Meditations, many commentators have

materially false ideas 'present themselves,' in a phenomenological sense, as

Preliminaries: Material Falsity and Objective Reality

So an idea is materially false if it provides 'subject-matter' for error, for false

Undoubtedly, the ideas which represent substances to me amount to something

'Representation' and the Idea of Cold

(hypothetical) fact, coldness is the absence of heat, a privation. 1 1 In this case, we

In order to distinguish what is clear in this connection from what is obscure, we

Bad Science and the Idea of Cold

we supposed that this abstraction had an extramental existence. This possibility

Obscure and Confused Ideas

The Idea of Cold Again

Arnauld and Descartes

Standardized terminology is essential here, but neither Descartes nor subsequent

mechanical reflex actions as reaching to soothe the offended part.

universalizes method, making it apply to all objects of thought - a method which

first, through insightfully bringing to bear traditional wisdom on the most

truth and goodness are to be discovered and developed by the individual on

Amauld's objection is inevitable, coming as it does from a person who at this

conclusions which can be recalled when we are no longer attending to the

resemblance is again superfici al. F or Augustine, th is doubt underm ines

I am indebted to my colleagues Vivien Bosley and Martin Tweedale for helpful

(New York and Oxford: Oxford University Press I 993 ), ch. 3.

24 All quotations in th is paragraph are from Arnauld ' s reproduction of Augustine at

Arnauld and Scepticism :

Amauld was viewed in his time as an extraordinary genius of great learning

Soon after its appearance in 1685, Arnauld's Reflexions philosophiques et

Arnauld' s response to the church' s attempted condemnation of Jansenius' s

sense perception is passive. So it is clear that this was an important modifica