garber descartes embodied

http://www.cambridge.org/0521783534

DESCARTES EMBODIED

This volume collects some of the seminal essays on Descartes by DanielGarber, one of the preeminent scholars of early-modern philosophy. A central theme unifying the volume is the interconnection betweenDescartes’ philosophical and scientific interests, and the extent towhich these two sides of the Cartesian program illuminate each other,a question rarely treated in the existing literature.

Among the specific topics discussed in the essays are Descartes’ celebrated method, his demand for certainty in the sciences, hisaccount of the relation of mind and body, and his conception of God’sactivity on the physical world.

This collection will be a mandatory purchase for any serious studentof or professional working in seventeenth-century philosophy, historyof science, or history of ideas.

Daniel Garber is Lawrence Kimpton Distinguished Service Professor in Philosophy and in the Committee on the Conceptual Foundationsof Science at the University of Chicago. He is inter alia the author ofDescartes’ Metaphysical Physics and coeditor of The Cambridge History of Seventeenth-Century Philosophy.

DESCARTES EMBODIEDReading Cartesian Philosophy through

Cartesian Science

DANIEL GARBER

University of Chicago

CONTENTS

Acknowledgments page ix

Abbreviations, Citations, and Translations xi

Introduction 1

Part I. Historiographical Preliminaries

1 Does History Have a Future? Some Reflections on Bennett and Doing Philosophy Historically 13

Part II. Method, Order, and Certainty

2 Descartes and Method in 1637 33

3 A Point of Order: Analysis, Synthesis, and Descartes’ Principles (with Lesley Cohen) 52

4 J.-B. Morin and the Second Objections 64

5 Descartes and Experiment in the Discourse and Essays 85

6 Descartes on Knowledge and Certainty: From the Discours to the Principia 111

Part III. Mind, Body, and the Laws of Nature

7 Mind, Body, and the Laws of Nature in Descartes and Leibniz 133

8 Understanding Interaction: What Descartes Should Have Told Elisabeth 168

9 How God Causes Motion: Descartes, Divine Sustenance, and Occasionalism 189

10 Descartes and Occasionalism 203

vii

11 Semel in vita: The Scientific Background to Descartes’ Meditations 221

12 Forms and Qualities in the Sixth Replies 257

Part IV. Larger Visions

13 Descartes, or the Cultivation of the Intellect 277

14 Experiment, Community, and the Constitution of Nature in the Seventeenth Century 296

Sources 329

Index 333

viii contents

ABBREVIATIONS, CITATIONS, ANDTRANSLATIONS

Although these essays were originally published at different times, indifferent places, and using different abbreviations and conventions ofcitation, I have tried to bring a certain amount of consistency to thecollective whole, at least when dealing with the writings of Descartes.In the essays that follow, I have used the following abbreviations:

AT Descartes, René, Oeuvres de Descartes, ed., Charles Adam and Paul Tannery, new edition. (11 vols.) Paris: CNRS/Vrin,1964–74. References by volume number and page (e.g., AT VII 74).

CSM Descartes, René, The Philosophical Writings of Descartes, ed.and trans., John Cottingham, Robert Stoothoff, and DugaldMurdoch (2 vols.). Cambridge: Cambridge University Press,1984–85. References by volume number and page (e.g.,CSM II 74).

CSMK Descartes, René, The Philosophical Writings of Descartes VolumeIII: The Correspondence, ed. and trans., John Cottingham,Robert Stoothoff, Dugald Murdoch, and Anthony Kenny.Cambridge: Cambridge University Press, 1991. Referencesto page numbers (e.g., CSMK 146).

AT remains the standard original-language text, and CSM and CSMKhave become the standard English translations. In some essays, thereare references to both AT and an English translation; more often, not. Since CSM and CSMK key their texts directly to AT, it should be

xi

easy enough to move from the AT citations that I usually give to thosetranslations. Though I do not always cite them, I often do borrow fromthem in essays written after they became available. In the earlier essays,I made some use of earlier translations that they replaced. In par-ticular, some of the earlier pieces in this collection borrow from theonce standard translations of Haldane and Ross1 and the volume ofDescartes’ letters edited and translated by Anthony Kenny2 (whichmetamorphasized into CSMK), as well as the translations of Anscombeand Geach,3 Laurence J. Lafleur,4 Paul J. Olscamp5 (for the Dioptrics andMeteors), and others that are lost in the sands of time and on the shelvesof my library. To these helpful crutches go all the praise and none ofthe blame: If I have borrowed their mistakes in translation (or, evenworse, made original mistakes of my own), it’s my own damned fault.In any case, direct references to outdated translations in the originalessays have been eliminated.

I have not tried to revise essays or footnotes in any extensive way.When I found that I no longer agreed with a view expressed in an essayI published some years ago, I was more inclined to omit it from thisvolume than try to correct it. Also, I have made no attempt to updatethe notes and references. Changes are limited to making the system ofreferences more consistent from one essay to the next, adding somecross references to other essays in this volume, and, in the case of oneessay, translating the quotations from Latin and French into English. I also tried to omit some overlapping passages. However, these essayswere written to be independent and free-standing, and given the inter-connected themes, some amount of overlap is inevitable.

xii abbreviations, citations, translations

1 Descartes, The Philosophical Works of Descartes, ed. and trans. E. S. Haldane and G. R. T. Ross(2 vols.). Cambridge: Cambridge University Press, 1911, and often reprinted.

2 Descartes, Philosophical Letters, ed. and trans. Anthony Kenny. Oxford: Oxford UniversityPress, 1970, later reprinted, Minneapolis: University of Minnesota Press, 1981.

3 Descartes, Philosophical Writings, ed. and trans. E. Anscombe and P. Geach. Edinburgh: T. Nelson, 1954, and often reprinted.

4 Descartes, Discourse on Method and Meditations, ed. and trans. Laurence J. Lafleur. Indianapolis: Bobbs-Merrill, 1960. Descartes, Rules for the Direction of the Mind, ed. and trans. Laurence J. Lafleur. Indianapolis: Bobbs-Merrill, 1961.

5 Descartes, Discourse on Method, Optics, Geometry, and Meteorology, trans. Paul J. Olscamp. Indianapolis: Bobbs-Merrill, 1965.

INTRODUCTION

My interest in Descartes was originally piqued when, as a graduatestudent, I had to assist in an introduction to philosophy. The DescartesI was asked to teach the students didn’t make much sense to me; I couldn’t figure out his point of view, why he was asking the kinds ofquestions he was asking, and why he was giving the kinds of answers hewas giving. Something about his larger intellectual context seemed tobe missing. But even then I knew that Descartes was deeply involved inthe physical sciences of his day, and even without knowing exactly whatCartesian science meant, I had a deep suspicion that it was somehowconnected with the philosophical writings I was teaching my under-graduates, the Meditations and the Discourse on the Method. At the time Iwas also very interested in the latest currents in contemporary philos-ophy, particularly the philosophy of Quine. Quine’s enormously influ-ential “Epistemology Naturalized” had just appeared, and everyone wastalking about a more general naturalization of philosophy and the inti-mate connection between philosophy and the sciences.1 That gave meall the more reason to turn to Descartes and his contemporaries, who,in a sense, took it for granted that there was a continuum between whatwe call philosophy and what we consider the sciences.

And so I undertook a serious study of Descartes’ science, as well asthat of his contemporaries. This led me to a number of interestingobservations. I came to see that Descartes’ thought must be understoodin the context of the attempt to reject Aristotelian physics, and replace

1

1 See W. V. O. Quine, Ontological Relativity and Other Essays. New York: Columbia UniversityPress, 1969, pp. 69–90.

it with a different kind of physics, one grounded in a mechanistic con-ception of nature. For an Aristotelian physicist, natural philosophy isultimately grounded in the irreducible tendencies bodies have tobehave one way or another, as embodied in their substantial forms.Some bodies naturally fall, and others naturally rise; some are naturallycold, and others are naturally hot; some are naturally dry, and othersare naturally wet. For the mechanist, though, the world is a machine,all the way down. According to the mechanical philosophy, of whichDescartes was a founder, I would argue, everything in the physical worldmust be explained in the way in which we explain machines, throughthe size, shape, and motion of their parts. Descartes was not the onlythinker of the period to hold such a view. Though there are some interesting and important differences among them, differences thatDescartes himself emphasized in many cases, one must also includehere contemporary figures such as Galileo, Mersenne, Gassendi,Hobbes, Roberval, and Beeckman, later Boyle, Locke, and many others. Nor was the mechanical philosophy the only alternative to Aris-totelianism; there were also alchemical, astrological, hermetic, Platonic,and other alternatives in the mix. One must understand Descartes’ philosophy as a part of this larger program to replace the Aristotelianphilosophy with a new and better alternative.

But there is a particular way in which Descartes approached the taskof replacing the Aristotelian philosophy with a mechanical philosophy.Although Descartes was interested in what we would call mathematicaland scientific questions, it was important for him to ground his view ofthe make-up of bodies and the laws that they observe in what he calleda metaphysics. In a celebrated passage from the preface to the Frenchedition of the Principia, Descartes writes that “all philosophy is like atree, whose roots are metaphysics, whose trunk is physics, and whosebranches, which grow from this trunk, are all of the other sciences,namely medicine, mechanics, and morals.”2 In the philosophical liter-ature, particularly that written by Anglo-American historians of philos-ophy, almost all the attention has been to the metaphysical roots. Ithought that it would be very useful to turn my attention to the part ofthe tree above ground, the trunk and the branches which were, if any-thing, more visible to Descartes’ contemporaries than the metaphysicalroots.

2 introduction

2 AT IXB 14. See the note on abbreviations and translations for the conventions used inciting Descartes’ writings.

One of the fruits of this work was my book, Descartes’ MetaphysicalPhysics.3 In this book, I tried to give a critical exposition of Descartes’physical thought, and discuss the arguments and positions that Descartesoffered in his writings on physics, mainly Le Monde (1633) and the Prin-cipia Philosophiae (1644), paying special attention to the way in which theyare grounded in metaphysics. But, at the same time, I was also workingon some of the more traditional questions in Descartes’ thought, ques-tions about knowledge, method, mind, and matter, exploring the way in which understanding Descartes’ scientific thought might illuminatethose more familiar aspects of Descartes’ philosophy. Many of the essaysin this collection are part of this effort. In taking the approach I do inthese essays, I do not mean to argue that it is the only approach that onecan take, that the only way one can understand Descartes is through hisscientific writings. Descartes was a multifaceted character, and there area number of approaches that one can take to illuminate his thought. AllI mean to assert is that this is one of them.

I should also say something about the historiographical ideas that liebehind these essays. The last twenty or thirty years have seen enormouschanges in the way in which the history of philosophy is written, at leastin English. When I first began working in the field in the mid-1970s, the dominant trend in Cartesian studies was to give careful attention toDescartes’ arguments and positions, and scrutinize them in accordancewith the current philosophical standards and doctrines. What it alsomeant, often enough, was a Cartesian philosophy pulled out of its intel-lectual context, with any historical considerations explicitly marginal-ized. I can remember in the late 1960s one of my undergraduate teacherswondering, in all seriousness, whether Descartes wrote before or afterNewton! Furthermore, the texts were almost always studied in transla-tion, with no need to know either the original language texts or any ofthe literature outside of English. Things have changed considerablysince then; the history of philosophy, at least in the early-modern period,is more and more genuinely historical. It is getting less and less possibleto do history of philosophy in translation alone, with no attention to his-torical context, and I am proud to have had some small part in thischange of standards. This historiographical theme is also reflected in theessays collected here. For me, understanding Descartes historicallymeans first and foremost situating him in the context of the larger

introduction 3

3 Chicago: University of Chicago Press, 1992. It has recently appeared in French as Laphysique métaphysique de Descartes, Paris: Presses Universitaires de France, 1999.

intellectual trends. However, it should also involve the attempt to under-stand Descartes as a living, breathing human being, who learns (andforgets) things, whose views develop and change over time, even if hehimself is not always aware of that dimension of his thought.

My historical temperament should not be taken to mean that I amuninterested in philosophy, and that I am abandoning a genuinelyphilosophical history of philosophy for a contextual history of ideas or an intellectual biography. Like many philosophical historians of philosophy, I believe in engaging historical figures, such as Descartes,in critical discourse, and even in rationally reconstructing their positions. However, as a historian of philosophy, I want as much as possible to do so on their own terms. Insofar as my job is to illuminatethe thought of a Descartes or a Leibniz or a Locke, I would prefer todo so by using terms and doctrines that they would find intelligible, to debate with them in their own language. Again, I acknowledge that this is not the only valid way of approaching the subject: It isimportant for us now to understand why a Cartesian account of thephysical world is no longer acceptable, and to do this involves engag-ing Descartes in a discussion with modern philosophy of science andeven modern physics. But unless we understand Descartes’ projects on their own terms, in the terms in which they were conceived, wecannot really understand what exactly his views really were, how theyreally relate to current conceptions, and what their true philosophicalsignificance is.

It is for reasons like this that I want to downplay (or perhaps evenblur) the distinction between history of philosophy and history of ideas.As Bernard Williams characterizes the distinction in his classic book,Descartes: The Project of Pure Enquiry,4 “history of ideas is history beforeit is philosophy, while with the history of philosophy it is the other wayround.” When dealing with an historical text, the history of ideas,according to Williams, focuses on the question “what did it mean” forits contemporaries, whereas the history of philosophy focuses on thequestion of its philosophical content. Williams writes: “The history ofphilosophy of course has to constitute its object, the work, in genuinelyhistorical terms, yet there is a cut-off point, where authenticity isreplaced as the objective by the aim of articulating philosophical ideas.”Williams casts his lot with history of philosophy understood in this way, and offers a self-consciously twentieth-century reconstruction of

4 introduction

4 New York: Penguin Books, 1978. All the quotations are taken from pp. 9–10.

Descartes’ thought. But can we really make the kind of separation thatWilliams (and many, many others) postulate? I can certainly understandthose who want to ignore history, and attack philosophical questionsdirectly; this, in a way, is the Cartesian spirit. However, if one choosesto write about Descartes (or Spinoza, or Locke, or . . .), then, it seems,this entails a kind of commitment to understand what they are tryingto say; a history of philosophy based on myths and partially understoodtexts is neither good history nor good philosophy, substituting forDescartes’ authentic thought a pale reflection of the contemporaryviews of interest to us. If we are to learn philosophy from Descartes, asopposed to using him as a mere foil for our contemporary views, thenwe must try to reach genuine understanding of what he thinks. Andgenuinely understanding an historical figure requires significant his-torical work, often going beyond the texts themselves and into the con-temporary culture to understand their presuppositions. Similarly, onecannot approach good history of ideas (in Williams’ sense) withoutunderstanding the philosophy as philosophy, as arguments and dis-tinctions and attempts at addressing systematically what are taken to beimportant problems. I don’t think that one should have to choosebetween the one and the other, between philosophical interest and historical sophistication. One needs both. Period.

Though the essays in this collection are all attempts at recovering agenuinely historical Descartes, in reading them over again, I am struckby how far scholarship has come in the last years. When I originallywrote them, and when they were originally published, many of theseessays were then on the outer edge of what was acceptable in the historyof philosophy; it is only through the kindness of editors who invited meto contribute to collections or special issues of journals that many ofthem found their way into print. But looking back at them now, theyseem, in a sense, rather old-fashioned. The essays are based on a carefulreading of the texts, all the texts, and not just the few generally read inphilosophy classes. Also, I try very hard to put those texts in the contextof other texts then in circulation, particularly late scholastic texts.However, two main things are missing. Although there is a smatteringof names unfamiliar to historians of philosophy, there are not enoughof them. In part this defect is addressed in the Cambridge History of Seventeenth-Century Philosophy,5 which I co-edited with Michael Ayers.There we made sure that less familiar names such as Sir Kenelm Digby,

introduction 5

5 Cambridge: Cambridge University Press, 1998.

Henry More, Louis de La Forge, and many others were reintegrated inthe story. But what is missing even there is the social context. Ideas existin people, and people exist in societies. As a consequence, social factorscan sometimes play a nonnegligible role in philosophy. Although thisis a commonplace now in the history of science (indeed, probably over-done), it is, I suspect, still a heresy in the history of philosophy. WhileI was doing my best to be heretical in some of the essays published inthis volume, the social historical approach was a kind of heresy that Ihadn’t yet come to appreciate. It will be better represented in somework currently in progress, a general study of the rise of the “new philosophy” in Paris in the 1620s and beyond.

It may be helpful to the reader to provide a brief guide to the con-tents of the book, and point out some themes and connections thatmight not be evident at first reading.

Part I of the book (“Historiographical Preliminaries”) is a generalhistoriographical essay, (1) “Does History Have a Future?” In this essay, I treat the general question of how one ought to do the historyof philosophy, and why one ought to do it. I argue, most centrallyagainst Jonathan Bennett, but also against many who share his con-ception of the history of philosophy, that the history of philosophyshould be done in a historically responsible way, and that the only wayto recover the true philosophical significance of historical figures is tounderstand them in their proper historical context. I further try to showwhat the history of philosophy done in this way can contribute to theenterprise of philosophy, how it can be used to challenge assumptionsthat we take for granted by exhibiting philosophical programs with perspectives very different from ours. This essay serves to present the methodology that I follow in the remainder of the essays in the collection.

Part II of the collection (“Method, Order and Certainty”) is con-cerned with methodological and epistemological issues in Descartes’philosophy. In (2) “Descartes and Method in 1637,” I treat the methodas articulated in Descartes’ Rules for the Direction of the Mind (1620–1628(?)) and the Discourse on the Method (1637). It is generally assumed thatthe method that Descartes articulates in those earlier works follows him throughout his career. In opposition to that, I argue that in animportant sense, the official method is abandoned in Descartes’ laterwritings, both scientific and philosophical. In the following two essays,(3) “A Point of Order: Analysis, Synthesis, and Descartes’s Principles”

6 introduction

(written jointly with Lesley Cohen) and (4) “J. B. Morin and the SecondObjections,” I treat the question of geometrical method in Descartes’writings. There is a standard reading of Descartes in accordance withwhich the Meditations (1637) are written in the analytic style, suppos-edly following the method of discovery of the Rules and the Discourse,whereas the more scientifically oriented Principles of Philosophy (1644)was written in the synthetic style characteristic of Euclidean geometry.This distinction has shaped a number of readings of Descartes’ philos-ophy, including most visibly the influential reading of Martial Guer-oult.6 In “A Point of Order” I argue against this dogma of Cartesianscholarship and suggest how to understand the different styles of thesetwo central works in Descartes’ corpus. In “J. B. Morin and the SecondObjections” I extend the argument by showing that one of the texts thatsupposedly grounds this interpretation, the end of the Second Replies tothe Meditations, was originally written not to endorse the syntheticmethod in any way, but as a reaction against another philosopher (andwell-known Aristotelian, anti-Copernican, and astrologer of his day),Jean-Baptiste Morin, who wrote a short tract on the existence of Godin the style of a Euclidean geometry text, a tract from which Descartesclearly wanted to dissociate himself. The last two essays in this partconcern Descartes’ actual method of conducting experimentalinquiries in his earlier and later works. In (5) “Descartes and Experi-ment in the Discourse and Essays” I show how, Descartes’ method from the Rules and the Discourse was used in the practice of experi-mental science by examining his analysis of the rainbow as given in theMeteors, published with the Discourse in 1637. In that essay, I try to showhow, for Descartes in this period, experiment is fully consistent with certainty. In (6) “Descartes on Knowledge and Certainty,” I show howthe problems of experimental philosophy ultimately move Descartes to abandon the claim that he can have certain knowledge of themicrostructure of matter, something that I think he had earlier believedhe could have.

Part III of the collection (“Mind, Body, and the Laws of Nature”) isconcerned with a number of central metaphysical and scientific ques-tions in Descartes’ philosophy. In (7) “Mind, Body, and the Laws of

introduction 7

6 See Martial Gueroult, Descartes’ Philosophy Interpreted According to the Order of Reasons, trans.Roger Ariew. (2 vols.) Minneapolis: University of Minnesota Press, 1984; and Nouvelles réflexions sur la preuve ontologique. Paris: Vrin, 1955.

Nature in Descartes and Leibniz,” I discuss the relation between voluntary activity and the laws of nature. It has been a standard view ofDescartes that he had wanted to make all the physical behavior of thehuman being consistent with his law of the conservation of quantity ofmotion. On that reading, Descartes is supposed to have held that thehuman will can change the direction of the motion of a body, but notits speed. Since Descartes’ conservation law governs only speed and notdirection, it was thought that this account allowed Descartes to renderhuman voluntary activity consistent with his conservation law. However,Leibniz showed that Descartes’ conservation law is incorrect, and thatthe correct conservation laws constrain direction as much as they dospeed. And so, Leibniz argued, that ploy won’t work. By carefully exam-ining Descartes’ conception of the laws of nature and how they derivefrom God, I argue that Descartes never intended human beings to begoverned by his laws of nature. I also show how Leibniz’s metaphysicsdiffers profoundly from Descartes’ in this regard, and why for him, thehuman being cannot stand outside of nature, as it can for Descartes.The following essay, (8) “Understanding Interaction: What DescartesShould Have Told Elisabeth,” also concerns mind and body inDescartes. It argues that Descartes’ famous letters to Elisabeth in 1643,explaining mind-body and body-body interaction, are importantly mis-leading. In those letters, Descartes claims that mind-body interactionand body-body interaction are each understood through their own sep-arate primitive notions. This, I claim, is inconsistent with some ofDescartes’ most basic commitments elsewhere. Rather, I argue, body-body interaction, the interaction between inanimate physical objects,must be understood ultimately through God, whose activity determinesthe laws of motion. The activity of God, in turn, must be understoodthrough our own experience of how we act on our own bodies. In thisway, mind-body interaction is the ultimate model in terms of which weunderstand all physical interaction for Descartes. The analysis of thephysical interaction among bodies is continued in the next piece, (9)“How God Causes Motion: Descartes, Divine Sustenance, and Occa-sionalism,” where I discuss how the dependence of the laws of natureon God gives rise to accusations of occasionalism in Descartes, andexplicit arguments for occasionalism in some of his followers. I arguethat the way in which Descartes conceives of divine activity leads himto reject a full occasionalism, where God is the only genuine causalagent. However, differences in the way some of his followers conceiveof divine activity lead them in another direction, to the occasionalism

8 introduction

characteristic of the later Cartesian tradition. In the following essay,(10) “Descartes and Occasionalism,” the question of Descartes’ occa-sionalism is examined in a more general way, where it is argued thatcontrary to much of the critical literature, Descartes was not a genuineoccasionalist. The last two essays in this section, (11) “Semel in vita: TheScientific Background to Descartes’ Meditations” and (12) “Forms andQualities in the Sixth Replies” both deal more directly with the relationbetween Descartes’ metaphysics and his physics. “Semel in vita” gives ageneral overview of the way in which Descartes’ metaphysics and epis-temology undermine Aristotelian science and ground the new physicsthat he is presenting in his works. “Forms and Qualities” discusses morespecifically the issue of Descartes’ rejection of Aristotelian forms andqualities, particularly as it is treated in a crucial passage at the end ofthe Sixth Replies.

In Part IV of the collection (“Larger Visions”), I include two essaysthat give larger views of Descartes’ philosophy. In (13) “Descartes, orthe Cultivation of the Intellect,” I present a view of Descartes’ concep-tion of the educated person, and how his conception of the humanbeing and the natural world led to a revolutionary conception of education, rejecting the authority of the book and the teacher for the authority of the intellect. Finally, in (14) “Experiment, Community,and the Constitution of Nature in the Seventeenth Century,” I putDescartes’ epistemology in the context of larger movements in seven-teenth-century thought, and show how Descartes’ radically individual-istic epistemology eventually gave way to a more social conception ofknowledge and scientific inquiry, as institutions such as the RoyalSociety of London and the Académie des Sciences in Paris entered thescene, and redefined the scientific world.

The careful reader may have noticed an oddity in the subtitle of thiscollection, “Reading Cartesian Philosophy through Cartesian Science.”Strictly speaking, this title makes little sense for the seventeenthcentury. At that time, neither philosophy nor science as we now knowthem could properly be said to exist as distinct domains of knowledge:What we call philosophy and what we call science were part of a singledomain of inquiry, which went under the rubric of philosophy. Butwithin Descartes’ thought there certainly was a distinction between thefoundational disciplines of philosophy, what he called “first philosophy”or sometimes “metaphysics,” and natural philosophy, between the rootsof his tree of philosophy and the trunk. It is this distinction that I have

introduction 9

in mind when I am talking about reading the philosophy through thescience. What I am attempting to do is put some of the Cartesian metaphysical, epistemological, and methodological doctrines on whichphilosophers have concentrated in recent years into the perspective ofDescartes’ larger system.

10 introduction

PART I

HISTORIOGRAPHICAL PRELIMINARIES

1

DOES HISTORY HAVE A FUTURE?

Some Reflections on Bennett and Doing Philosophy Historically

The history of philosophy seems to play a very significant role in theactual practice of philosophy; historical figures come up again andagain in the courses we had to take, both as undergraduates and as graduate students, and historical figures continue to come up again andagain in the papers we read, the courses we teach, the conferences weattend. Philosophy seems to be a subject that is obsessed with its past,but it is more than just an obsession. Most of us would agree that under-standing the history of philosophy is somehow important to doing phi-losophy, that we are better philosophers for knowing the history of oursubject. I think that this is true. As philosophers, we have an obligationto ourselves to reflect on this fact: why is history important to our enter-prise, and how is history important to our enterprise?

This is what I would like to do in this short essay, make some obser-vations about the ways in which history of philosophy can and doesinfluence the practice of philosophy. I shall begin by discussing the viewof history found in Jonathan Bennett’s recent and already enormouslyinfluential book, A Study of Spinoza’s Ethics. I have chosen to talk aboutthat book in good part because it is, I think, the best representative ofa certain genre of writing in the history of philosophy; Bennett nicelyarticulates a view of the history of philosophy that is widespread amongwriters on the subject, particularly those writing in English. Bennett’sview, widely shared, is that history is important because studying his-torical figures can teach us philosophy; in the history of philosophy wehave a storehouse of arguments and positions worth taking seriously asphilosophy, worth discussing and debating in the same way the work of a very good contemporary philosopher is worth discussing and

13

debating. I shall not really criticize Bennett’s view of the matter. Thereis a sense in which he and the multitude of other philosophers and his-torians of philosophy who share his view are absolutely correct. But, Ishall argue, Bennett makes use of only a portion of the riches thathistory has to offer. In the second part of this essay I shall try to sketchand illustrate a somewhat different conception of the use of history inphilosophy that complements the conception Bennett offers.

History as Storehouse

I would like to begin my discussion by outlining what I take to beJonathan Bennett’s attitude toward history in his recent book, A Studyof Spinoza’s Ethics. My interest in the book will be largely metaphilo-sophical (or, perhaps, metahistorical); though I have some disagree-ments with Bennett on matters of substance, I shall do my best not dodrag them in here and muddy the waters.

Early in the book, Bennett gives the reader ample indication of thenature of his interest in Spinoza. “I am not writing biography,” he notes.“I want to understand the pages of the Ethics in a way that will let melearn philosophy from them.”1 A bit later in the book, Bennett indi-cates that his interest is “not with Spinoza’s mental biography but withgetting his help in discovering philosophical truth.”2 At the end of thebook Bennett writes:

The courtly deference which pretends that Spinoza is always or usually right,under some rescuing interpretation, is one thing; it is quite another to lookto him, as I have throughout this book, as a teacher, one who can help us tosee things which we might not have seen for ourselves. That is showing hima deeper respect, but also holding him to a more demanding standard.3

Bennett’s interest here is clear: it is finding philosophical truth andavoiding philosophical falsehood that he is after, and the study ofSpinoza is supposed to help us in this search. What he says aboutSpinoza presumably holds more generally for the study of other figuresin the history of philosophy. So conceived, the history of philosophy isa kind of storehouse of positions and arguments, positions and argu-ments that we can use as guides or inspirations to the positions weshould take, or illustrations of dead ends that we should avoid.

14 historiographical preliminaries

1 Jonathan Bennett, A Study of Spinoza’s Ethics (Indianapolis, Ind.: Hackett PublishingCompany, 1984), p. 15.

2 Ibid., p. 35. 3 Ibid., p. 372.

This last provision is important. The point is not that Spinoza (or anyother historical figure) will simply hand us philosophical truth on aplatter, arguments or positions that we can immediately adopt withoutchange. Bennett’s Spinoza often makes mistakes, and bad ones; hardlyan argument in the Ethics can stand without some correction. Yet we canlearn from Spinoza even when he is wrong (or, at least we usually can;Bennett seems unsure about whether anything can be learned from themistakes Spinoza makes when discussing the eternity of the mind).4

Bennett writes:

I do say that Spinoza’s total naturalistic program fails at both ends and in themiddle; as though he undertook to build a sturdy mansion all out of wood,and achieved only a rickety shack using bricks, as well as wood. But hisattempt was a work of genius; and a thorough, candid study of it can be won-derfully instructive. The failures have at least as much to teach as the suc-cesses, if one attends not only to where Spinoza fails but why.5

Bennett completes the thought a few pages later:

I spoke of how much we can learn from Spinoza’s successes and, especially,his failures. It is his minimalism that makes his work so instructive. If you seta mechanical genius to build an automobile engine out of a Meccano set,you won’t get a working engine from him, but as you watch him fail you willlearn a lot about automobile engines.6

(In giving these quotations I don’t mean to imply that they are trans-parently intelligible or true on their face, but I would like to postponethose questions for the moment.)

What does the history of philosophy look like from Bennett’s point ofview? We begin by trying to reconstruct the arguments the philosopherwe are studying gave, trying to follow the train of thought he followed.But our ultimate goal is philosophical truth, and it is with that in mindthat we must approach our reconstruction; we must carefully examinethe truth of the premises, the validity of the inferential steps, and with acold and unsentimental eye, judge the truth or falsity of the conclusionand the adequacy of the means by which the conclusion was reached. Ifappropriate, we might make some attempt to patch up the argument,adding new premises, substituting better premises for worse, trying a newpath to the conclusion in question, or whatever. This is, I think, a fairrepresentation of what Bennett is doing in the Spinoza book.

does history have a future? 15

4 Ibid., pp. 372, 357. 5 Ibid., p. 38. 6 Ibid., p. 41.

All of this is interesting and, in an important sense, valuable activity.But if we are to follow Bennett and hold that the history of philosophyis valuable primarily insofar as it helps us to find philosophical truth,in some more or less direct sense, then there are certain consequenceswe must accept.

First of all, if we insist on philosophical truth as the only motivationfor studying history, then a great deal of the history of philosophy mayturn out to be marginal to the philosopher. Bennett would agree thatfew historical figures have any large store of doctrines or argumentsthat we would now consider live candidates for truth or even approxi-mate truth. There are those who study Aristotle or Saint Thomas, Kantor Marx, because they think that at least some of what they wrote isclose to being true, and because they believe that attention to their writ-ings can help lead us directly to insights we would not otherwise have.But how many study Descartes or Leibniz or Spinoza for this reason?The noble attempts of the past, one might argue, are instructive in theirfailures. But while failures can be instructive, a few can go a long way.The student architect can learn to fit the building to the available materials and know the strengths and weaknesses of both from thebuilding that collapsed. But one learns to design successful buildingsby studying successful buildings, not just failures. Having had a deprivedchildhood, I’m not sure I know exactly what a Meccano set is, but if itis what I think it is, I doubt that I could learn much about automobileengines by watching someone try to build one from a Meccano set, nomatter how talented one might be. Similarly, the philosopher mustlearn to recognize a bad argument and must be trained to avoid themistakes people make. This is only a small portion of one’s philosoph-ical education, which, I think, should focus on positions and argumentsthat people think are live candidates for the truth, at least insofar asone is being trained to seek philosophical truth. Bennett may overesti-mate what we can learn directly from failed arguments and programs.Insofar as the great majority of historical arguments, positions, and pro-grams are failures when judged against the high standard of philo-sophical truth (as we see it), the study of the history of philosophy mayhave less to contribute to philosophy than Bennett seems to think, andless than we historians would like.

There is another feature of Bennett’s position worth drawing out.Bennett’s position has the danger of distorting the history of philosophy.First of all, insofar as we regard history of philosophy as contributing tothe discovery of philosophical truth, we are led to emphasize those por-tions of a philosopher’s work that speak to our interests, that address our


conception of where philosophical truth is to be found, leaving otheraspects of the work aside, thereby mutilating what may be a unified andsystematic point of view. Bennett has not done any such thing to Spinoza,but one can call to mind the numerous commentaries on Descartes andanthologies of his writing that barely mention his work in mathematics,physics, or biology; the accounts of Pascal that focus on the wager argu-ment without indicating its larger context; books like Anthony Kenny’slittle book Aquinas, in the Past Masters series, or John Mackie’s Problemsfrom Locke, which quite self-consciously use standards of contemporaryrelevance to choose what to discuss and what to ignore. In each case, thefocus on philosophical truth distorts our historical understanding of thefigure and his position. But there is another way in which historical dis-tortions may enter. If our interest is philosophical truth, then the pointof the historical enterprise is to capture whatever philosophical truth orinteresting philosophical falsehood there may be in some philosopher’swritings. What this has often meant in practice is what has been dubbedrational reconstruction, taking the argument or position as given andmaking sense of it in terms that make sense of it to our philosophicalsensibilities, whether or not the reformulation captures anything thephilosopher himself would have acknowledged. Examples of this include Bernard Williams’s reconstruction of the argument of Descartes’Meditations using modern epistemological concepts, or Benson Mates’sreconstruction of Leibniz’s doctrine of possible worlds using contem-porary modal logic. Bennett is tempted in this direction as well. In apassage, part of which we have already quoted, he writes:

I want to understand the pages of the Ethics in a way that will let me learnphilosophy from them. For that, I need to consider what Spinoza had inmind, for readings of the text which are faithful to his intentions are likelyto teach me more than ones which are not – or so I believe, as I think himto be a great philosopher. And one can be helped to discover his intentionsby knowing what he had been reading, whose problems he had been chal-lenged by, and so on. But this delving into backgrounds is subject to a law ofdiminishing returns: while some fact about Maimonides or Averroes mightprovide the key to an obscure passage in Spinoza, we are more likely to gethis text straight by wrestling with it directly, given just a fair grasp of hisimmediate background. I am sure to make mistakes because of my inatten-tion to Spinoza’s philosophical ancestry; but I will pay that price for the ben-efits which accrue from putting most of one’s energies into philosophicallyinterrogating Spinoza’s own text.7


7 Ibid., pp. 15–16.

Indeed, many benefits come from directly interrogating a historicaltext, leaving aside nice worries about historical context, but there is adanger of misunderstanding. (In Bennett’s Spinoza book this comesout most clearly in his discussion of space and his attribution of a “fieldmetaphysic” to Spinoza in chapter 4, a lovely philosophical position,but one that I do not think occurred to Spinoza.)

This may sound like a criticism of the approach Bennett takes tohistory, but I assure him, it is not. If our only goal is philosophical truth,then history of philosophy may turn out to be marginal, if not alto-gether expendable; If our goal is simply philosophical truth, we mustface up to the facts in an unsentimental way. And, if our goal is philo-sophical truth, then historical veracity can have only an instrumentalvalue at best; it is of value only insofar as it helps us attain our princi-pal goal. The point of interpretation, on this view, is to make the phi-losophy breathe, to make it available to us, and historical veracity isimportant only insofar as it serves this end.

In calling for us to focus on the truth and falsity of Spinoza’s claims,the adequacy and inadequacies of his arguments, Bennett is implicitlycontrasting the approach that he takes with other more disinterestedlyhistorical and, in one sense, less philosophical approaches that onemight take to the material. In one place Bennett contrasts his approachwith that of “intellectual biography,” with “mental biography” inanother, and with that “which pretends that Spinoza is always or usuallyright, under some rescuing interpretation” in a third passage.8 Now, itseems to me that the disinterested historian shouldn’t always assumethat Spinoza is right. But insofar as we agree with Bennett that Spinozawas “a great philosopher,” we should at very least subscribe to theworking hypothesis that what Spinoza is up to is sensible, the sort of thingthat a smart person might believe in a particular historical context,given what he had learned, what others around him believed, theassumptions taken for granted, and so on. (This is just a special case ofwhat has been called the principle of charity or, in variant, the princi-ple of humanity in the theory of interpretation in the philosophy oflanguage.) This is not to say that we should not expect to find lapses ofreasoning and judgment, even when the whole context is open to us,or that this kind of historical inquiry will clear up all our puzzlements.It is important to remember that Spinoza, for example, was a puzzle tohis contemporaries as well, and they had more access to his context


8 Ibid., pp. 15, 35, 372.

than any of us can ever hope to have. In its way, this kind of rationalityis no less demanding a standard to hold Spinoza to than philosophicaltruth is.

Unlike philosophical truth, which judges Spinoza by what is true, orby what we have come to think is true, this standard is internal. Thealternative to the sort of history Bennett advocates is an historical recon-struction of Spinoza’s views, the attempt to understand Spinoza’s posi-tions and arguments in terms that he or a well-informed contemporaryof his may have understood. It involves coming to understand whatSpinoza or a contemporary of his would have considered unproblem-atic background beliefs, what they would have had trouble with, and inthe light of that and other similar contexts, coming to understand whatSpinoza’s conception of his project was, how he thought he had estab-lished the conclusions he had reached, and what he thought was impor-tant about those conclusions, all under the assumption that, by andlarge, Spinoza’s project is the work of a smart person working within aparticular historical context. This sort of investigation is not biographyof any sort, neither intellectual nor mental; it is, quite simply, the historyof philosophical ideas.

In practice, the kind of history I was sketching may come out lookingvery little different from the history Bennett prefers. As Bennett haspointed out, if it is the lessons of history for philosophical truth thatinterest us, then the lessons are likely to be more interesting the closerwe come to a genuine representation of Spinoza’s (or whoever’s)thought. The only conspicuous difference may be the relative lack ofjudgments of truth and falsity in the sort of disinterested history Ipropose. If our interest is in historical reconstruction, the question ofthe ultimate truth or falsity of the doctrines is simply not at issue; theonly thing that is important is whether or not our account has madethe beliefs intelligible. Sometimes this will call for a judgment that onhis own terms, some premise or inference a philosopher uses may not beavailable to him, properly speaking. If we are interested in historicalreconstruction, then, for example, the falsity of a premise then univer-sally accepted is not a relevant part of the story.

Bennett would certainly have to agree that there is a real projecthere, whether or not he himself is interested in carrying it out. I thinkthat he would also have to agree that there is no reason why one mustchoose one conception of the history of philosophy over the other.While in practice a single scholar may find it difficult to pull off bothsorts of history at the same time, within the confines of a single essay


or book, they are not competing programs in the sense in which, say,deontological programs for ethics compete with teleological programs.One can find the history of philosophy richer for having bothapproaches represented in the literature, one can find both interestingand never be put into the position of having to choose one over theother. In this sense the two approaches to history of philosophy arecomplementary rather than competing.

A question remains, a central question. On Bennett’s view, thehistory of philosophy is important to philosophy in an obvious way; onhis conception, history of philosophy actually contributes to theunearthing of philosophical truths. Now, as I noted, the sort of disin-terested historical reconstruction I have sketched can contributeto Bennett’s enterprise, but taken by itself, does it have any philosoph-ical interest at all? Leaving aside the question of the philosophical truth it may help to uncover, is the purely historical study of philo-sophical ideas of more than antiquarian interest? Is there any reasonfor philosophers qua philosophers to take an interest in such dis-interested history?

In Defense of Disinterested History

In arguing for the philosophical significance of disinterested history, Iwould like to proceed historically and begin with a consideration of theviews of a philosopher whose opinion on the matter is in many waysattractive to me. The philosopher I have in mind here is Descartes. AsBennett proposes we learn from Spinoza, I propose that there is muchwe can learn from Descartes.

Descartes may seem at first glance an odd character to turn to in thisconnection. Descartes was conspicuously unsympathetic to the study ofbooks, old or new. In the Discours, Descartes wrote:

[A]s soon as I was old enough to emerge from the control of my teachers, Ientirely abandoned the study of letters. Resolving to seek no knowledge otherthan that which could be found in myself or else in the great book of theworld, I spent the rest of my youth in traveling. . . . For it seemed to me thatmuch more truth could be found in the reasoning which a man makes con-cerning matters that concern him than in those which some scholar makesin his study about speculative matters. For the consequences of the formerwill soon punish the man if he judges wrongly, whereas the latter have nopractical consequences and no importance for the scholar except thatperhaps the further they are from common sense the more pride he will take


in them, since he will have had to use so much more skill and ingenuity intrying to render them plausible.9

This attitude also comes out nicely in a letter from 1638. Unfortunately,the recipient of the letter is unknown, as is the book Descartes is com-menting on in the letter, but his point is clear:

[The author’s] plan of collecting into a single book all that is useful in everyother book would be a very good one if it were practicable; but I think thatit is not. It is often very difficult to judge accurately what others have written,and to draw the good out of them without taking the bad too. Moreover, theparticular truths which are scattered in books are so detached and so inde-pendent of each other, that I think one would need more talent and energyto assemble them into a well-proportioned and ordered collection . . . thanto make up such a collection out of one’s own discoveries. I do not meanthat one should neglect other people’s discoveries when one encountersuseful ones; but I do not think one should spend the greater part of one’stime in collecting them. If a man were capable of finding the foundation ofthe sciences, he would be wrong to waste his life in finding scraps of knowl-edge hidden in the corners of libraries; and if he were no good for anythingelse but that, he would not be capable of choosing and ordering what hefound.10

Descartes does not mince words here. If it is truth we are after, bookswill not help us to find it. He does not seem to think that we can learnmuch from other people’s mistakes, unlike Bennett; mistakes justengender other mistakes. The truths we find in books are so rare andso scattered that anyone who has the ability to recognize them and seekthem out would be better off simply looking for them on his own,directly, without the help of these paper-and-ink teachers. If it is philo-sophical truth you are after, Descartes tells Bennett (and anyone elsewho will listen), then don’t look to the philosophers of the past. (It issomewhat disquieting to the historian when one of his or her subjectstalks back in such a rude way.)

Descartes, in general, has little truck with scholarship, with the studyof the past, but Descartes was not altogether dismissive of history.Though he thought it inappropriate to look for philosophical truth inhistory, he did not think that reading the authors of the past is altogetherwithout value. In the Discours he wrote:


9 Descartes, Discours de la methode, I, AT VI 9–10; CSM I 115.10 AT II 346–47; CSMK 119.

I knew . . . that reading good books is like having a conversation with the mostdistinguished men of past ages – indeed, a rehearsed conversation in whichthese authors reveal to us only the best of their thoughts.11

This conversation is valuable to us for an interesting reason. Accordingto Descartes:

[C]onversing with those of past centuries is much the same as traveling. It isgood to know something of the customs of various peoples, so that we mayjudge our own more soundly and not think that everything contrary to ourown ways is irrational, as those who have seen nothing of the world ordinar-ily do.12

Through such experience in books and in the world Descartes claimsthat he learned that there are “many things which, although seemingvery extravagant and ridiculous to us, are nevertheless commonlyaccepted and approved in other great nations; and so I learned not tobelieve too firmly anything of which I had been persuaded only byexample and custom.”13

The idea is an interesting one. We can learn from the past in some-thing of the same way we can learn from travel. By traveling we can geta certain kind of perspective on our lives and the way we lead them, thethings we do and the things we believe. We go to other countries, learntheir languages, observe their customs, eat their foods (or, at least,observe the kinds of foods they eat), discuss their beliefs about theworld. This, Descartes thinks, can give us a certain perspective on ourown lives. It can, among other things, free us of the belief that the waywe see things is the way things have to be, that X is fit for human con-sumption but Y is not, that weeks must have seven days, that childrenmust be raised by their own parents, etc. Descartes’ point is not rela-tivistic here; he would be among the last to say that anything goes. Eventhough we observe others eating a certain food we do not, we may stillshun it and continue to hold the belief that it is unhealthy or improperfor us to eat. Seeing what others do may at least get us to raise the ques-tion for ourselves why we have the beliefs and customs we do and,perhaps, lead us to see what is arbitrary and what is well grounded inour beliefs and behavior.

A similar case can be made with respect to the study of the past ingeneral, and the study of past philosophy in particular, Descartes sug-


11 Discours I, AT VI 5; CSM I 113. 12 Discours I, AT VI 6; CSM I 113–14.13 Discours I, AT VI 10; CSM I 115–16.

gests. Many of the philosophical beliefs we now take for granted are notshared by figures in the past. By studying the past, taking the past seri-ously, we are led to reflect on our beliefs, in just the same way as we areled by travel to reflect upon our customs. Such reflection need not leadto a change in our beliefs. The fact that some past geographers thoughtthe earth flat, or past physicists thought that there is such a thing aselemental fire that by its nature rises, these historical observationsshould not move us to give up our present conceptions of geographyor combustion. Reflection on some of the things people have believedshould at least cause us to ask ourselves why we believe the things wedo, and whether our grounds are sufficient to support the explicit orimplicit beliefs we have and assumptions we make.

Is such reflection important for us as philosophers? It does notdirectly contribute to the discovery of philosophical truth in the way inwhich discovering a good argument (or an interesting false one) in thework of a historical figure perhaps might, in the way in which Bennettconceives of history contributing to the practice of philosophy. The sortof contribution Descartes saw was of a different, and more subtle,though no less important kind. Historical investigation conceived inthis alternative way gives us a kind of perspective on the beliefs we haveand the assumptions we make. It helps us sort the good from the bad,the arbitrary from the well grounded, insofar as it challenges us toreflect on why we believe what we do. While it may not help lead usdirectly to new arguments and new philosophical truths, it leads usdirectly to something just as valuable: philosophical questions.

All of this is very abstract and cries out for some concrete examples,specific assumptions and beliefs we make that are illuminated by suchhistorical reflection. Before I present such an example, I would like tocontinue a bit longer in this abstract vein.

Descartes has suggested a philosophical use for the history of phi-losophy different from the one Bennett suggests; the suggestion, as Ihave developed it, is that the history of philosophy can be importantnot because it leads to philosophical truths, but because it leads to philo-sophical questions. But what sort of history is relevant here? To learnfrom history in the way Descartes suggests we can involves trying tounderstand historical figures on their own terms. If I travel to Tokyo orNairobi, look for what is familiar to me in the alien setting, and seek itout, I may acquire a nice camera cheaply, or learn one way not to makea pizza. I may indeed have a lovely vacation, but I will not learn what I might. Similarly, if what I am looking for in history is a guide to


philosophical truth, if I look for things recognizable to me as interest-ing philosophical problems and promising (if possibly flawed) philo-sophical arguments, as Bennett seems to suggest we should, then I maymiss features of philosophy as it has been that might raise interestingquestions about philosophy as it is. To learn from history in the wayDescartes suggests, we should – we must, I think – undertake the kindof disinterested historical investigation I suggested earlier as an alter-native to the sort to which Bennett’s views lead him. If it is an histori-cal perspective on our beliefs and assumptions we are interested in, thenthe truth or the falsity of past views is simply irrelevant. It matters not atall whether Descartes’ or Aristotle’s or Kant’s views are true or false forthis use of history. What is important is that we understand what theirviews were, and that we understand how it is that smart people couldhave regarded them as true. It is not their truth, much less their falsity,that causes us to reflect on our own beliefs; it is the fact that smartpeople took seriously views often very different from ours that is important here.

This, I think, answers the question posed at the end of the lastsection. The sort of disinterested historical reconstruction I proposedas a complement to Bennett’s philosophically informed investigation ofthe history of philosophy is philosophically significant, a worthwhileactivity for philosophers to engage in, though for a reason somewhatdifferent from what Bennett suggests for his program. Bennett’s historyseeks philosophical truth, answers to philosophical questions; mineseeks the questions themselves.

Raising Questions: Science and Philosophy

I have been sketching out a way of doing philosophy historically, usinga disinterested historical reconstruction of past thought as a way ofraising important philosophical questions that might otherwise escapeour notice. A brief example illustrates the approach I have been advocating.

Bennett makes an interesting statement in the course of his com-mentary on Spinoza. He writes: “Much of the Ethics is philosophicalrather than scientific, i.e., is answerable to conceptual analysis ratherthan to empirical observation”14 The claim is not central to Bennett’sreading of Spinoza, and in raising questions about it I don’t mean to


14 Bennett, Spinoza, p. 24.

cast doubt on Bennett’s larger interpretation (though I do think thaton at least one occasion it does lead him a bit astray). The quotationappeals to a certain widely held conception of philosophy: that it is anactivity pretty largely distinct from scientific activity, and that philoso-phy makes use of conceptual analysis, whereas science makes use ofobservation and experience. This conception of philosophy and its rela-tion to science is worth some historical examination.

We might begin by noting that in Spinoza’s day, things were not soneatly partitioned. It is now generally recognized that the words “phi-losophy” and “science” didn’t have distinct and separate meanings inthe seventeenth century. Whereas “philosophy” was sometimes usednarrowly, in perhaps something of the way we use it now,15 it was alsoused more broadly to include knowledge in general, including what wenow call science, as in the title to Descartes’ Principia Philosophiae, three-fourths of which is scientific by our standards. Similarly, whereas“science” was sometimes used as we do now,16 it often took on a meaningderived from its Latin origin, scientia, knowledge. This, of course, is onlya matter of terminology. The important question is not what things werecalled, but whether Spinoza and his contemporaries drew an interest-ing distinction between what we call philosophy and what we areinclined to call science, between a certain collection of foundationalquestions, investigated through argument and conceptual analysis, anda different set of questions about the natural world, investigatedthrough observation and experience.

Here, I think we can say that while we can certainly find differentquestions studied by different thinkers using different modes of inves-tigation, there is no radical distinction between what we call philo-sophical and what we call scientific.

It is quite widely known that arguments that are in general termsphilosophical play a major role in seventeenth-century science. A niceexample is the derivation Descartes gave for his laws of motion.Descartes started from two main premises. The first was an analysis ofthe “nature of time,” which, Descartes claims, is “such that its parts arenot mutually dependent,” and from which he argued that God isrequired to keep everything in existence at every moment.17 The


15 See, e.g., Discours I, AT VI 6, 8–9; CSM I 113, 114–15.16 See, e.g., the preface to the French translation of the Principia Philosophiae, AT IXB 14,

CSM I 186.17 Principia I 21, AT VIIIA 13; CSM I 200.

second premise was that God is immutable by his nature and operates“in a manner that is always constant and immutable.”18 From thesepremises Descartes argued that a constant quantity of motion is main-tained in the world, and that bodies in uniform rectilinear motion willtend to remain in uniform rectilinear motion.19 These conclusions, con-clusions that spring from Descartes’ metaphysical foundations, wereenormously influential on later seventeenth-century physicists. Thoughnot altogether correct in detail, Descartes’ conclusions constituted thefirst published statement of a conservation principle and the first clearversion of what Newton was later to call the principle of inertia. WhenNewton presented his version of these laws in his Principia almost fifty years later, the metaphysical argument was gone. But it wasn’t dead. Leibniz, Newton’s great and greatly maligned contemporary, a physicist and mathematician whose only clear better was Newtonhimself, made free use of metaphysical arguments in his physics. LikeDescartes, Leibniz chose to derive the laws of motion from God, thoughin a different way: from God the creator of the best (and so, mostorderly) of all possible worlds, not God the moment-by-moment sustainer of all. God, Leibniz reasoned, would want to create the worldin such a way that whatever power, whatever ability there is to do workin a complete cause, must be found intact in its full effect. Using thisas his main premise, Leibniz established two of the main principles ofclassical mechanics, the law of conservation of what we call kineticenergy (mv2, vis viva), and the conservation of what we call momentum(mv).20

These arguments establish what we would call scientific conclusionsby way of what we would call philosophical premises. There are alsoinstances in which what we would call (and Bennett has called) con-ceptual analysis taken more narrowly is used in the service of science.What I have in mind is Descartes’ celebrated arguments for the identi-fication of space and body, and his conclusion that there is no emptyspace, no vacuum. In one representative version, noted by Spinoza inhis Principles of Descartes’ Philosophy, quoted and discussed by Bennett inhis commentary, the claim reads:


18 Principia II 36, AT VIIIA 61; CSM I 240.19 Principia II 36–39, AT VIIIA 61–65; CSM I 240–42.20 For an account of Leibniz’s work here, see, e.g., Martial Gueroult, Leibniz: dynamique et

metaphysique (Paris: Aubier-Montaigne, 1967), chapter 3.

Space and body do not really differ [because] body and extension do notreally differ, and space and extension do not really differ. It involves a contradiction that there should be a vacuum [i.e.] extension without bodilysubstance.21

Bennett claims that this position is a purely philosophical one, and thatneither Descartes nor, following him, Spinoza should confuse this withdoing science. He writes: “[W]hen he [Descartes] says that there is novacuum, he is not predicting what you will find if you ransack the phys-ical universe. His point is a conceptual one.”22 Bennett furthermoreregrets “that he words this possible philosophical truth so that it soundslike a scientific falsehood” and goes on to chastise Descartes andSpinoza for their occasional lapses into thinking that this philosophi-cal argument has empirical consequences for physics.23 Bennett is toocharitable here, and in his charity, he misses the point of the argument,both in Descartes and in Spinoza. Descartes’ point was precisely to estab-lish that there is no vacuum in the physical world, and I know of noreason to believe that Spinoza read the argument any differently.Whether or not there is a philosophical truth in the claim, it was whatwe have come to recognize as a scientific falsehood that interestedDescartes and his contemporaries; the denial of a vacuum not only in philosophy but also in rerum natura was an important feature of Cartesian physics, one that grounds Cartesian cosmology, the vortextheory of planetary motion.24

The examples so far are of cases in which philosophical argument,conceptual analysis, leads to what we would consider scientific conclu-


21 C. Gebhardt, ed., Spinoza Opera (Heidelberg: Carl Winter, 1925), 1:187–88, as para-phrased in Bennett, Spinoza, p. 100. Spinoza refers the reader here to Descartes’ PrincipiaII 17–18, AT VIIIA 49–50; CSM I 230–31.

22 Bennett, Spinoza, p. 101. 23 Ibid.24 Descartes’ view was that the present state of the world can be explained if we imagine an

initial state of disorder, which sorts itself out into swirls of fluid by way of the laws of motionalone. These swirls of fluid, vortices, are what Descartes identifies with planetary systems,a sun at the center of each, and planets circling about the sun. Essential to this accountis the assumption that all motion produces circular motion, which Descartes derives fromthe doctrine of the plenum. It is because all space is full, he argues, that all motion mustultimately be circular, one hunk of material substance moving to make room for a givenmoving body, a third hunk moving to make room for the second, and so on until a finalhunk moves to take the place left by the original moving body. In this way, Descartes’whole cosmology depends on the denial of the vacuum. For the account of motion as circular, see Principia II 33 (AT VIIIA 58–59; CSM I 237–39) and for the derivation of thecosmos from an initial state, see Principia III 46ff. (AT VIIIA 100ff.; CSM I 256ff.).

sions. There are a few interesting and, to the modern mind, verystrange instances in which seventeenth-century philosophers usedempirical claims to support conclusions that we would consider philo-sophical. The case is strange, and I’m not entirely sure I have it right,but Leibniz seems to have taken such a position. Leibniz held (or, atleast, he often held) that animals are genuine substances, corporeal sub-stances. As substances, Leibniz argued, they cannot arise throughnatural means, nor can they perish by natural means. This is a conclu-sion Leibniz often establishes by pure philosophical argument; it is aconclusion of the celebrated predicate-in-notion argument of Discourseon Metaphysics, §8,25 and, Leibniz sometimes argues, of the no-less-philosophical principle of continuity.26 Leibniz also called on the exciting discoveries of microscopists like Leeuwenhoek and Malpighifor support. For example, he wrote to Queen Sophie Charlotte in May1704 concerning an important consequence of his view of corporeal substance:

Speaking with metaphysical rigor, there is neither generation nor death, butonly the development and enfolding of the same animal. . . . Experience con-firms these transformations in some animals, where nature herself has givenus a small glimpse of what it hides elsewhere. Observations made by the mostindustrious observers also lead us to judge that the generation of animals isnothing but growth joined with transformation.27

Microscopic examiners are being called upon to support one of thebasic propositions of Leibniz’s metaphysics, the natural ungenerabilityand incorruptibility of substance.

If this strikes us as being a bit strange, stranger still is Henry More,who calls upon the world of ghosts and goblins as empirical support forhis belief in the existence of incorporeal souls. In his Immortality of theSoul (1662 edition) More calls our attention to

such extraordinary Effects as we cannot well imagine any natural, but mustneeds conceive some free or spontaneous Agent to be the Cause thereof,whereas yet it is clear that they are from neither Man nor Beast. Such are


25 See C. I. Gerhardt, ed., Leibniz: Philosophische Schriften (Berlin: Weidmannsche Buchhand-lung, 1875–1890), 4:432–33, translated in Leroy Loemker, ed. and trans., Leibniz: Philo-sophical Papers and Letters (Dordrecht: D. Reidel, 1969), pp. 307–8. See also the letter toArnauld, 28 November/8 December 1686, Gerhardt 2:76.

26 See Leibniz’s letter to Queen Sophie Charlotte, 8 May 1704, Gerhardt 3:345.27 Ibid. See the discussion of this and the references cited in Michel Serres, Le systeme de

Leibniz (Paris: Presses Universitaires de France, 1968), 1:354ff.

speakings, knockings, opening of doors when they were fast shut, . . . shapesof Men and several sorts of Brutes, that after speech and converse have suddenly disappeared.28

That there are such happenings is, for More if not for us, an empiricalfact. For More these apparitions speak strongly in favor of souls distinctfrom body: “Those and like extraordinary Effects . . . seem to me to bean undeniable Argument that there be such things as Spirits or Incor-poreal Substances in the world.”29 More may have been deluded in think-ing that there are ghosts and obscure about how the phenomenon inquestion is supposed to support his conclusion, but he certainly seemedto think that the question of the existence of incorporeal substance, ametaphysical question par excellence, could be settled by a trip to ahaunted house. In this he was not alone. Hobbes, no advocate of imma-terial substance, made a special point of denying the reality of ghostsas part of his case against incorporeal souls.30 Although he did notsupport the view More was pressing, Hobbes certainly seemed to thinkthat empirical evidence was germane to the question.

Why are these historical observations interesting? For one, they dopertain to the proper interpretation of Spinoza and his contemporaries;they suggest that we should be careful about attributing our distinctionbetween philosophy and science to earlier thinkers. There is a philo-sophical lesson to be learned as well. My point is not that we should lookfor philosophical truth in the sorts of arguments I was discussing; thelaws of motion shouldn’t be derived from God, nor should the ques-tion of the vacuum be settled by an appeal to our intuitions about spaceand extension. Nor do I think that metaphysical issues about the natureof substance can be settled by looking into microscopes, nor should weconsider seriously the ontological status of ghosts and goblins. Muchthat was live in seventeenth-century thought is now dead, and I don’tintend to revive it. The examples I have given do raise an interestingquestion: Why is it that we tend to see such a radical break between phi-losophy and science, and, more important, should we? The question canbe raised directly, without the need for history, as Quine has done. Buthistory brings the point home in an especially clear way: It shows us anassumption we take for granted by pointing out that it is not an assump-tion everyone makes.


28 Henry More, The Immortality of the Soul, p. 50, in A Collection of Several Philosophical Writingsof Dr Henry More (London: William Morden, 1662).

29 Ibid. 30 Hobbes, Leviathan, chapter 46; cf. chapter 2.

Conclusion

Some years ago, an anthropologist friend told me something of what itis like to do field work. When one enters a new community, she said, itis all very alien, an alien language, alien customs, alien traditions. Aftera while things change; the language and customs become familiar, andone is inclined to think that the differences are only superficial, thatthe once-alien community is just like home. The final stage comes whenthe similarities and differences come into focus, when one recognizeswhat one’s subjects share with us, while at the same time appreciatingthe genuine differences there are between them and us. The case issimilar for the history of philosophy. We cannot ignore the ways inwhich past thinkers are involved in projects similar to ours, and the waysin which we can learn from what they have written, how it can con-tribute to our search for philosophical enlightenment. At the sametime, we cannot ignore the ways in which they differ from us, the wayin which their programs differ from ours, the way in which they ask dif-ferent questions and make different assumptions. Both are importantto a genuine historical understanding of the philosophical past, but justas important, we as philosophers can learn from both.


PART II

METHOD, ORDER, AND CERTAINTY

2

DESCARTES AND METHOD IN 1637

The Discourse on the Method and the three essays that were published withit, the Dioptrics, the Meteors, and the Geometry, make up a very curiousbook. The very title page emphasizes the preliminary discourse, andthat discourse, the Discourse on the Method, emphasizes method, theimportance that method had for Descartes in making the discoverieshe made, the importance that the method Descartes claims to havefound will have for the progress of the sciences and for the benefit ofhumankind as a whole. Descartes is not, of course, telling us that weare obligated to follow his method; the Discourse is, after all, proposed“as a story, or, if you prefer, as a fable” (AT VI 4). But Descartes expectsthat we will all see the light, the light of reason, of course, and followhis example. It is curious, then, that Descartes gives the reader onlybrief hints of what that method is, four brief, vague, and unimpressiverules that, taken by themselves, would hardly seem to justify Descartes’enthusiasm, not to mention a whole discourse in their honor. Further-more, explicit methodological concerns are hardly in evidence in theDioptrics, the Meteors, and the Geometry, which are, Descartes claims,“essays in this method,” as he identifies them on his title page. Indeed,one is hard pressed to find much evidence of the method at all after1637, either explicit discussions of the method or explicit applicationsof the method in any of Descartes’ writings, published or unpublished.Very curious.

These observations raise quite a number of questions about thedevelopment of Descartes’ thought and the state of his program as of1637. In this essay I shall address two of these questions: (1) What pre-cisely was the method Descartes had in mind in 1637, when he sang its

33

praises so enthusiastically? and (2) Why does that method appear solittle in the publications of 1637 and appear to drop out altogetherafter that? Briefly, I shall argue that the method of 1637 was just themethod Descartes had put forward more clearly in the earlier Rules forthe Direction of the Mind, or, at least, the dominant method that showsthrough the latest stages in its composition. But, I shall argue, perhapsby 1637 and certainly after, that method began to show its limitations,and the method that was one of Descartes’ first discoveries, one of his first inspirations proved itself inadequate to the mature program that it led Descartes to undertake. Obviously there is not the space topresent the detailed discussions these questions require. But I shall try to present in broad strokes one way of understanding the develop-ment of Descartes’ methodological thought as he passed from youth tomaturity.

I

I have claimed that the method of 1637 is essentially the method of theRules for the Direction of the Mind, and to make good on that claim, wemust first turn to that work. The Rules, started as early as 1619 and aban-doned in 1628, is a very difficult work; despite its superficial organiza-tion, it is often strikingly unmethodical and disorderly for a work thatis supposed to be Descartes’ most systematic exposition of his method.It is blatantly a work in progress that never progressed to anything likea finished draft, and the text we have shows obvious signs of havingbeen picked up and put down at different times throughout the periodof composition.1

To begin unraveling Descartes’ complex thought on method in theRules we must look to the earliest strata of the work, where Descartessets out the goal of the method in passages likely to have been writtenin November 1619, shortly after the dreams of November 10.2 Descarteswrote:

34 method, order, and certainty

1 For questions of dating, see J.-P. Weber, La constitution du texte des Regulae (Paris: Sociétéd’édition d’enseignement supérieur, 1964). Weber believes that Descartes wrote the textof the Rules in ten discrete “phases.” Though the stages of composition are difficult to dis-tinguish with such exactitude, Weber’s arguments are often useful for dating particular pas-sages of the Rules. I have also used datings suggested by John Schuster in his “Descartes’Mathesis Universalis, 1619–28,” in S. Gaukroger (ed.), Descartes: Philosophy, Mathematics andPhysics (Sussex: the Harvester Press, 1980).

2 See Weber, La constitution, §§ 13, 55.

The goal [finis] of studies ought to be the direction of one’s mind [ingenium]toward having solid and true judgments about everything which comesbefore it. (AT X 359)

But, Descartes thinks, such “solid and true judgments,” such “certainand indubitable cognition” (AT X 362) as he calls it in the followingrule, can come to us in only two ways, through intuition, or throughdeduction, “for in no other way is knowledge [scientia] acquired” (ATX 366). And so, what we should seek is an intuition, “the undoubtedconception of a pure and attentive mind” (AT X 368), or a deduction,a chain of such intuitions, grounded in intuition, arrived at through “acertain movement of our mind [ingenium],” inferring one thing fromanother (AT X 407). To find such knowledge, though, Descartes thinksthat we need a method (Rule 4). But what is this method and how is itsupposed to work?

From the start Descartes had in mind a two-stage process. Writing inRule 5, again from late 1619,3 Descartes summarized his rule of methodas follows:

This rule is observed exactly if we reduce involved and obscure propositionsstep by step to simpler ones, and thus from an intuition of the simplest wetry to ascend by those same steps to a knowledge of all the rest. (AT X 379)

The rule of method thus has two steps. First there is a reductive step inwhich “involved and obscure propositions” are reduced to simpler ones.This is followed by a constructive step, in which we proceed from anintuition of the simplest back to the more complex.4

But what in concrete terms does the method come to? How is it tobe used in specific cases? It is quite possible that Descartes’ vision in

descartes and method in 1637 35

3 See ibid. §§ 19, 55.4 To avoid confusion, I am breaking with most commentators, who refer to these as the ana-

lytic and synthetic steps, following the distinction Descartes draws in the Second Replies. See,for example, Ch. Serrus, La méthode de Descartes et son application à la métaphysique (Paris:Librarie Félix Alcan, 1933), chapter I; Geneviève Rodis-Lewis, L’oeuvre de Descartes (Paris: J.Vrin, 1971), pp. 173ff; L. J. Beck, The Method of Descartes (Oxford: Oxford University Press,1952), chapter XI, etc. This is a distinction that has little direct relevance to the stages of the method of the Rules. In the Rules we are dealing with a distinction between two parts of a single method; though they are distinct, both are necessary for a true applica-tion of the method. But the distinction between analysis and synthesis in the Second Repliesis completely different. There we are dealing with different ways of setting out a single lineof argumentation, and we must choose one or the other. See AT VII 155–56 or AT IX–A212. On analysis and synthesis, see Garber and Cohen, “A Point of Order,” essay 3 in thisvolume.

1619 was cloaked in poetic enthusiasm, and that Descartes himself maynot have had a clear and distinct conception of precisely how themethod was to work in actual practice. Matters are clarified consider-ably in an example Descartes gave late in the composition of the Rules,where the programmatic bravado of the earlier years is translated intopractice. The example I have in mind is the anaclastic line, whichDescartes discusses in Rule 8. The example is closely connected withoptical investigations Descartes undertook probably between 1626 and1628, and probably dates from that period.5 But whenever it dates from,it displays what I take to be the method as Descartes understood it atthe time he abandoned the project of the Rules, and represents whathe means by method in 1637, I shall argue.

The argument is set out in Table 1. The problem Descartes poses isthat of finding the shape of a line (lens) that focuses parallel rays oflight to the same point (AT X 394). Now, Descartes notices – and thisseems to be the first step in the reduction – “the determination of this[anaclastic] line depends on the relation between the angle of inci-dence and the angle of refraction.” But this question is still “compositeand relative,” and we must proceed further in the reduction, to thequestion of how this refraction is caused by light passing from onemedium into another, which in turn raises the question as to “how theray penetrates into the whole transparent thing, and the knowledge ofthis penetration presupposes that the nature of illumination is alsoknown” (AT X 394–95). But in order to understand what light is,Descartes claims, we must know what a “natural power [potentia natu-ralis]” is. This is where the reduction ends. At this point Descartes seemsto think that we can “clearly see through an intuition of the mind” (ATX 395) what a natural power is, something that we understand in termsof local motion.6 Once we have this intuition, we can then begin the


5 See Schuster, “Mathesis,” pp. 55, 88 n.68. Weber dated this text to the year 1621, basing hisargument on the dating of the discovery of the law of refraction that G. Milhaud proposed;see Weber, La constitution, § 23bis. I follow Schuster here. Setting aside the dating of thelaw of refraction, it appears clear that the text concerning the “noblest example,” an appli-cation of the method to epistemological questions (AT X 395, l. 17), a text that Weber cor-rectly links to the anaclastic example, does not date from 1621. This text on the “noblestexample” is intimately connected with the following text, AT X 396 l. 26ff, which Weberdates to the years 1625–27; see La constitution, §§ 23–24, 60. We discuss the epistemologi-cal project of Rule 8 in part III of this essay.

6 In Rule 9 Descartes says that “if one wants to examine” this natural power, one must turnto “local motions of bodies” [AT X 402]. According to Schuster, this passage probably datesfrom the same period as the anaclastic line example; see Schuster, “Mathesis,” p. 87 n.60.

constructive step, and follow back in order through the questions raiseduntil we have answered the original question, Q6 allowing us to deducean answer to Q5, Q5 allowing us to deduce an answer to Q4, and so onuntil we reach an answer to Q1, deductively.7

This example suggests the following conception of method. Method-ical investigation begins with a question. This question is reduced tosimpler questions, questions whose solution is presupposed for the solu-tion of the question originally posed. That is, Q1 is reduced to Q2 ifwe must answer Q2 before we can answer Q1. Descartes thinks that thisprocess leads us from more specific questions to more general, morebasic, more fundamental questions, from the shape of a specific lens,to the law of refraction, to the nature of light and the nature of a naturalpower. Descartes thinks that when we follow out this reductive series,we will ultimately reach an intuition. Here the reduction ends and con-struction begins. At this point we can turn the procedure on its head,and begin deducing answers to the questions that we have successivelyraised, in an order the reverse of the order in which we raised them.When we are finished it is evident that we shall have certain knowledge;the answer arrived at in this way will constitute a conclusion deducedultimately from an initial intuition.


Table 1. Anaclastic Line Example (Rules for the Direction of the Mind, Rule 8)

Q1. What is the shape of a line (lens) that focuses parallel rays of light to thesame point?

Q2. What is the relation between angle of incidence and angle of refraction (i.e.,the law of refraction)?

Q3. How is refraction caused by light passing from one medium into another?

Q4. How does a ray of light penetrate a transparent body?

Q5. What is light?

Q6. What is a natural power?

Intuition: A natural power is. . . .

Construction: The construction consists in traversing the series of questions fromQ5 to Q1, deducing the answer to each question from that of the precedingquestion.

7 For a lucid discussion of the anaclastic line example, see Pierre Costabel, Démarches origi-nales de Descartes savant (Paris: J. Vrin, 1982), pp. 53–58.

Descartes’ strategy here is extremely ingenious. The stated goal ofthe method is certain knowledge, a science deduced from intuitivelyknown premises. What the method circa 1628 gives us is a workable pro-cedure for finding an intuition and a deductive chain from which suchknowledge can be attained. This workable procedure is the reductionof a question to more and more basic questions, questions we can iden-tify as questions whose answers are presupposed for answering the ques-tion originally posed. The efficacy of the reductive step of the methoddepends upon a substantive assumption about knowledge, the assump-tion that knowledge, scientia, is structured in a very specific way, a doc-trine that Descartes seems to have held in one form or another sincethe crucial night of 10 November 1619 (cf. AT X 204, 215, 255, 361).It is not at all clear how in detail Descartes may have seen this structurein 1619. But Rule 12 of the Rules suggests that by 1628 Descartes sawall knowledge grounded in intuitions about the very most general fea-tures of the world, thought, extension, shape, motion, existence, dura-tion, etc. On these intuitions are grounded layers of successively lessgeneral propositions. If knowledge is structured in this way, thenDescartes thinks we should be able to solve any problem in an orderlyand methodical way, tracing step by step through the layers, back towardthe intuition, and deducing down from there to the question that interests us.

My account of method in the Rules ignores numerous complexities.I have said relatively little about the stages of composition of the Rules,and nothing about simple natures or the use of experiment in themethod (though I will touch on that a bit later). Also, I have saidnothing about the mathesis universalis of Rule 4 and Rule 14, which someargue is identical to the method (they are wrong, I think, but it wouldtake me too far from my main theme to argue the case).8 And finally,


8 The question of the mathesis universalis and its connection to the method is very importantfor my interpretation; to the extent that we are interested in the usage (and the lack ofusage) of the method in Descartes’ thought, we have to understand what Descartes under-stood by “method.” Now, some have supposed that Descartes identified the method withthe mathesis universalis; see, e.g., G. Milhaud, Descartes savant (Paris: Félix Alcan, 1921), p.69; Paul Mouy, Le développment de la physique cartésienne: 1646–1712 (Paris: J. Vrin, 1934)pp. 4–5; Jean-Luc Marion, Sur l’ontologie grise de Descartes (Paris: J. Vrin, 1981), §§ 9–11;Desmond Clarke, Descartes’ Philosophy of Science (University Park, PA: The Pennsylvania StateUniversity Press, 1982), pp. 166ff. The question is complicated, but I think that this inter-pretation is mistaken. My aim here is not to establish my conclusion with certainty, but itappears to me that the method of the anaclastic line example is the definitive method ofthe Rules, that which reappears in the Discourse, as we have seen. It is true that, from timeto time in the Rules, Descartes seems to identify his method with the mathesis universalis,

I have neglected to mention the numerous other assumptions, largelyunwarranted, I think, that Descartes needs to make his method work.9

But what I have given is an account of the method Descartes held in 1628 or so when he stopped work on the Rules to turn to the con-struction of his system.

II

It is this method, I claim, that Descartes had in mind in 1637 when he published the Discourse on the Method. The method I attributed toDescartes in the Rules agrees well enough with the brief exposition ofthe method, the four rules that he gives in part IV, particularly thesecond and third of those rules. The second rule requires us “to divideevery difficulty . . . into as many parts as one can,” and the third sug-gests that I conduct “my thoughts in order, beginning with the simplestobjects, those easiest to understand, to rise little by little, as by degrees,up to the most composite knowledge” (AT VI 18). Although I think thatcommentators have not, in general, grasped the method Descartes recommends in the Rules, the obvious correspondence between thetwo-stage method Descartes recommends in the Rules, the reductionfollowed by the construction, and these two rules he recommends inthe Discourse have often been noted.10


above all in Rule 14. My hypothesis is this. In the last stage of the composition of the Rules,Descartes had a brilliant idea. The most important thing about the method, as presentedin the anaclastic line example, is order. But Descartes had been interested perhaps tenyears earlier in a science of pure order, that is, in what he called the mathesis universalis.Descartes might have thought that this science of order was applicable to the method of Rule 8, and in one way or another, that one could use the mathesis universalis in the methodical solution of problems. That seems to me to be the idea of Rule 14. Andperhaps, in this same moment of enthusiasm, Descartes attached an older exposition ofthe mathesis universalis to what was then extant of Rule 4, where he introduced the method,intending to return to Rule 4 and integrate the old with the new, a conjecture which mightexplain certain strange aspects of the text in this Rule; cf. Weber, La constitution, chapterI. But it seems to me that this marriage between mathesis and method did not work, andDescartes abandoned the idea very quickly. There is not a single example of the methodthat Descartes suggests in Rule 14, and there is no reason to think that Descartes had anymore than a vague and impressionistic idea of the mathesis as method.

9 For example, Descartes supposed that the simpler questions are also those that are meta-physically more fundamental. Furthermore, in the anaclastic line example, the intuition,the step presented as the last in the reduction, is very obscure. Finally, it is not certainthat one can deduce, properly speaking, the answer to a question originally posed fromthe intuition to which the method leads one.

10 See, for example, Étienne Gilson, René Descartes: Discours de la méthode, texte et commentaire,4me éd. (Paris: J. Vrin, 1967), p. 205; L. J. Beck, Method, pp. 149ff; etc.

But our account of Descartes’ views on method in 1637 cannot stopwith the Discourse. The Discourse, Descartes tells his correspondents,does not contain a genuine exposition of the method. Descartes wroteto Mersenne on 20 April 1637, discussing the title he chose for his Discourse:

I did not call it Treatise on the Method but Discourse on the Method, which is thesame as Advice on the Method, to show that I did not intend to teach it, butonly to discuss it. Since, as one can see from what I said about it, it consistsmore in practice than in theory. (AT I 349)

The method, then, “consists more in practice than in theory.” But what“practice” should we examine? In writing to P. Vatier about the methodon 22 February 1638, Descartes makes a suggestion: “I have given aglimpse [of the method] in describing the rainbow” (AT I 559). Thereference here is to the eighth discourse of the Meteors, where Descartesgives his celebrated account of the rainbow. Descartes there tells thereader that

I could not choose material more appropriate to show how, by the method Iuse, one can arrive at knowledge which those whose writings we have didn’tpossess. (AT VI 325)

A study of the account Descartes gives of the rainbow is, then, sup-posed to teach us the method by showing us how it works “in practice.”But, as Descartes also told Vatier, “the matter is very difficult” (AT I559), and it is not at all easy to discern the clear outlines of Descartes’method in the mists that surround the rainbow.

Very briefly,11 Descartes uses a combination of reasoning and exper-iments with spherical flasks for water and with prisms to lead him to anexplanation of the two principal features of the rainbow, the colors wesee, and the fact that the rainbow is always composed of two separateregions of color that are separated by a dark space. From the experi-ments with prisms, Descartes concludes that colors arise when light isbent in refraction; he argues that the color is caused by the tendencyto rotate that the balls receive during refraction. From observations on the flask of water, and calculations made with the help of his law of refraction, Descartes concludes that most sunlight passing into a


11 This paragraph is a summary of the argument given in Discourse 8 of the Meteors, AT VI325ff. For a fuller development of the claim that this example exemplifies the method,see “Descartes and Experiment in the Discourse and Essays,” essay 5 in this collection.

droplet of water and following certain paths will leave at one of twoangles, about 42 degrees and about 52 degrees. Putting these togetherwe have the rainbow, roughly speaking, two regions of color that arisethrough refraction, separated by about 10 degrees.

It is not easy to extract a method from this morass of detail, but onecan see in Descartes’ account the outlines of the two-step method, thereductive step followed by the constructive step that constitutes the coreof Descartes’ method in the Rules. In Table 2 have rearranged the argu-ment a bit to show its structure. In the diagram, Q1–Q5 constitute thereduction of the initial question, the ordered succession of questionsDescartes would answer to answer the question originally posed. Thereduction ends with an intuition about the nature of light and how it passes through bodies. (In the Meteors Descartes actually appeals tothe Dioptrics, where the nature of light is presented as a hypothesis. SeeAT VI 331 and 84.) D1–D4 constitute the constructive stage of themethod, where Descartes goes from the intuition to the solution to the problem originally posed. (Again, Descartes actually appeals hereto results that are derived in the Dioptrics, the law of refraction. See ATVI 337 and 93ff.) Viewed in the way I suggest, the account of therainbow nicely displays the method of the Rules that we saw in the ana-clastic line example.

While it does not pertain to my main theme in this essay, I shouldpoint out how nicely the rainbow example shows us the role of experi-ment in the method.12 It is worth noticing that experiment enters onlyin the reductive stage of the method; it helps us to find a path fromour complex question to the intuition from which that question will beanswered. But the answer itself is purely deductive, and makes no useof experiment. The chain of causes that the Cartesian scientist seeks inreason is exemplified in the causal connections one finds in natureitself. Insofar as these later connections are open to experimental deter-mination, we can use experiment to sketch out the chain of causes andfind what causally depends on what, and thus use the connections wefind in nature as a guide to the connections we seek in reason. It maynot be obvious how we can go deductively from the nature of light tothe rainbow, but poking about with water droplets, flasks, and prismsmay suggest a path for the deduction to follow. But this does not makethe deduction superfluous; while it may be through effects that we are


12 For a fuller development, see, again, “Descartes and Experiment,” essay 5 in this collection.

Table 2. Descartes’ Account of the Rainbow (Meteors, Eighth Discourse)

Q1. What causes the rainbow (two regions of color)?

[Rainbows appear only in the presence of water droplets; size is irrelevant to the phenomenon.]

Q2. What causes the two regions of color in any spherical ball?Q2a. What causes the two regions? Q2b. What causes the color?

[The two regions result from two [Color is produced without a curvedcombinations of reflection and surface and without reflection; it refraction.] requires a restricted stream of light,

and a refraction.]

Q3a. Why do the two combinations Q3b. How does refraction cause colorof reflection and refraction result in under appropriate circumstances?two discrete regions?

Q4. How does light pass through media?Q5. What is light?Intuition: The nature of light, and how it passes through media [Cf. Q5, Q4].

D1a. Law of refraction D1b. The only change in a restricted stream of light passing from one medium to another (refraction aside)is a differential tendency to rotation.

D2a. All parallel rays of light D2b. Color can only be the converge into two discrete streams differential tendency to rotation after two refractions and one or two produced in passing from one reflections, emerging from the drop medium to another in refraction(flask) in two discrete regions [Cf. Q3b].[Cf. Q3a].

D3. Parallel rays of light produce two discrete regions of color on a spherical ballof water [Cf. Q2].

D4. Sunlight (parallel rays of light) on a region of water droplets will producetwo regions of color [Cf. Q1].

led to causes, it is from knowledge of those causes and the deductionswe can make from them that our knowledge actually derives.

But let us return to the main theme. An examination of the rainbowexample, Descartes’ own announced example of the method in 1637,strongly suggests that the method Descartes had in mind in the contextof the Discourse and Essays was just the method of the Rules, the two-stage method we saw in the anaclastic line example, the reduction of aquestion to an intuition, and the construction of an answer to that ques-tion from intuition. But it is interesting to note that the account of therainbow we have been discussing is probably not contemporaneous withthe Discourse; while it is impossible to be certain, it is likely that thatportion of the Meteors dates from late 1629, not long after the Ruleswere set aside.13 When the account of the rainbow appears eight yearslater in the Meteors, it appears as a kind of ghost from an earlier period.This is significant, for the account of the rainbow is the only place inthe Essays where Descartes explicitly calls attention to the method of hispreliminary discourse and it is the only example of the method to whichhe calls attention in his letters. Though the method “consists more inpractice than in theory” (AT I 349), the practice in question is notexemplified elsewhere in the Essays. The Essays are, of course, notunconnected in Descartes’ mind with the method. Descartes wrote toMersenne in April 1637:

I call the treatises that follow essays in this method because I claim that thethings they contain couldn’t have been found without it, and that through[what I have discovered] one can know the value [of the method]. (AT I 349)

But though they show the value of the method, the Essays do not them-selves use the method. Writing to Vatier on 22 February 1638, Descartesexplains this as follows:

I couldn’t show the usage of the method in the three treatises which I gavebecause [the method] requires an order for investigating things that is very different from that which I thought necessary to use to explain them.(AT I 559)


13 In a letter of 8 October 1629 Descartes wrote to Mersenne that he is working on “a smalltreatise which will contain the explanation of the colors of the rainbow (to which I havegiven more care than all the rest) and generally the explanation of all sublunar phe-nomena” (AT I 23). This small treatise will doubtless become the Meteors, and Descartes’words to Mersenne suggest that Descartes probably solved the problem of the rainbowbefore October 1629.

The mode of exposition Descartes chose for the Dioptrics and the Meteorswas, of course, hypothetical. Both works begin with appropriate hypothe-ses which ground the results which follow, hypotheses that allowDescartes to show some of his results, but in a way that does not forcehim to divulge the first principles of his system, something for which, hebelieved, the public was not ready (AT I 370, 563–64; AT III 39).

But it is interesting to note that even in other contexts, whereDescartes is not too shy to divulge the foundations of his system, the method is hardly in evidence. In the earlier World, for example,Descartes divulges more of the foundations of his physics than he willdo later in the Essays; though certain metaphysical issues that Descarteswas concerned with at the time are hidden, he is forthcoming aboutthe nature of matter, the nature of light, the role God plays in main-taining the world, and so on. But it is difficult to discern the formalmethod of the Rules in the World. And when a few years later Descartessets aside his scruples and presents his system in its full and proper formin the Meditations and the Principles of Philosophy, there is as little of themethod as there was in the World and the Essays. Descartes does con-tinue to build on first principles, to start with intuition (ultimately thecogito), and deduce down from there, from the more general and moremetaphysical to the more specific. This, of course, is a feature of theorder of reasons that M. Gueroult emphasized, and it looks a great deallike the constructive stage of the method. And his continued interestin experiment and observation show that he is still keenly aware of theproblem of finding an appropriate path from intuition to the solutionof particular problems in physics. For example, Descartes’ keen inter-est in embryology and sexual reproduction in the 1640s was, I think,part of an attempt to bridge physics and biology;14 perhaps an under-standing of how purely mechanical processes result in the genesis of anew organism will show how in nature organisms arose from lifelessmatter, Descartes thought. But there is little evidence of the earliermethod in his later writings, in particular, little evidence of a formalreduction that precedes the constructive deduction of conclusions fromintuition, the reduction that earlier had constituted the principal secretof the method. This is so even in the Meditations, a work whose origin,Descartes tells the Doctors of the Sorbonne, was in part a response toa request for him to apply his celebrated method to God and the soul(AT VII 3), a work written in the analytic mode, Descartes tells the


14 See, for example what Descartes says in the Discourse, AT VII 45–46, and the commentaryin Gilson, Commentaire, pp. 393ff.

second objectors, a work that is intended to follow “the true waythrough which a thing was . . . discovered” (AT IXA 121). In the Medi-tations, the intuition that constitutes the starting place of the deduction,the cogito, is carefully prepared in the First Meditation. But the prepa-ration does not seem to be a reduction in the precise sense of the term.The First Meditation does many things; it clears away prejudice, estab-lishes a standard for certainty, introduces the problem of knowing ourcreator as the essential preliminary for any further knowledge. But itdoes not sketch out the sequence of steps to be followed in resolving aquestion, the way a proper reduction is supposed to do.

One cannot deny that the Meditations are carefully organized andordered. But even though there is an order, this order is not evident tothe meditator at the start of the Meditations. From the cogito of Medita-tion I to the end of Meditation VI there are numerous places where themeditator tries to lead the argument into a dead end, where the med-itator begins an argument that simply does not pan out. For example,at the beginning of Meditation III, the meditator tries to demonstratethe existence of the external world, before giving the proof for the exis-tence of God. However, at this point in the argument of the Meditations,the meditator doesn’t have the means to make his proof work, and hemust set the question aside, and turn to another question, to God,leaving aside the question of the external world until Meditations V andVI, where it can finally be settled. These digressions are very importantto the structure of the Meditations.15 The Meditations are addressed, inpart, to a very specific audience that Descartes knows quite well, to theunconverted, readers full of prejudice for their senses and for the mate-rial world, and these digressions are very important to convince themthat the arguments that they are inclined to accept, arguments that takefor granted a faith in the senses, arguments that take for granted a pri-ority in belief in the external world – these arguments Descartes wantsto show are mistaken. And the way he does this is by letting the meditator try to show that they work, only to show that they don’t. This is the function of the failed argument for the existence of body inMeditation III, for the wax example of Meditation II, and for otherarguments in the Meditations.16


15 Other important digressions include the celebrated piece of wax in Meditation II and theargument for the existence of the external world drawn from the faculty of imaginationat the beginning of Meditation VI.

16 See Garber, “Semel in vita: The Scientific Background to Descartes’ Meditations,” essay 11in this volume for an elaboration of some of these themes.

There is method in this procedure, to be sure, but the method is notthe strict method of the Rules or the Discourse. In the method of hisyouth, the reductive step brings it about that the entire constructivestep is sketched out, before the first deduction, and the constructionfollows directly the order as set out in the reductive step; this, indeed,is the main point of having a reduction, so that one will know how to perform the deduction, and this reductive step is the principal secret of the method, what makes it work. In this method there is no place for the sorts of digressions so important to the purpose of the Meditations. Furthermore, it is not clear to me that one can isolateone well-defined question to which Descartes addresses himself in the Meditations – a minimal condition required for the method of the Rules to apply. In this sense one can say that the meditator doesn’t followthe method, nor can the reader learn the method by reading the Meditations.

In claiming that Descartes’ later works do not display his earliermethod I am making a controversial claim, one that would be chal-lenged by other scholars, who have claimed to find the method of theRules and Discourse in the Meditations, at very least.17 But even if they areright (and I don’t think they are), it is beyond dispute that Descarteshimself hardly mentions his method after the Discourse and the letters that immediately follow its publication. If method is the key toknowledge and the key to the later Cartesian system (as it seemed tobe in 1637), Descartes himself does not call attention to that fact.Indeed, when the earlier method comes up in his later writings, it hasa decidedly subordinate role in his thought. In the Letter to Picot thatserves as a preface to the French edition of the Principles of Philosophyof 1647, Descartes recommends that the student of philosophy “oughtto study logic, not that of the schools, but that which teaches one how to conduct his reason to discover truths that one doesn’t know”(AT IXB 13–14). It would be good, Descartes says, for him to “prac-tice the rules concerning easy and simple questions for a long time”until “one acquires a certain habitude for finding truth in these ques-tions” (AT IXB 14). But in this respect, the method has roughly thestatus of the provisional morality (which immediately precedes it in theLetter), one of those preliminaries that should be undertaken by thestudent of nature before undertaking the serious business of philoso-


17 See Serrus, La méthode, chapter III; Beck, Method, chapter XVIII; Peter Schouls, The Imposition of Method (Oxford: Oxford University Press, 1980), chapters IV–V; etc.

phy; it is an exercise useful primarily in sharpening the mind andhelping us to recognize truth, an exercise that has in 1647 roughly thesame role that Descartes earlier gave the scholastic logic he otherwiserejected in the Rules (cf. AT X 363–64). Whatever it is, it is clearly notnearly so important to Descartes in the 1640s as it appeared to be in 1637.

How can we account for these curious facts? How can we explain thefact that method gets such little play in Descartes’ actual scientific writ-ings? How can we explain the fact that the method, the central focusof his theory of knowledge and inquiry in 1637, is barely mentioned inlater writings? My claim is this. The method was Descartes’ first inspi-ration, and was crucial for the first results of his system, as he reportsin the Discourse. But, I shall argue, two basic changes in Descartes’thought made the method largely obsolete.

III

Descartes’ method first dates from mid- and late November 1619, it is generally agreed, the days and weeks following the crucial threedreams. It had been a year since Descartes had run into the young IsaacBeeckman in Breda and had his first sympathetic introduction to themechanical approach to nature that was later to dominate his thought.Beeckman was not a systematic thinker, it is fair to say, in the sense thathe had no large, overarching system. He was interested in the solutionof individual problems, and it is with the discussion of individual prob-lems, taken one by one, that his notebooks are filled.18 It was this wayof doing physics and mathematics that he transmitted to the youngRenatus Picto, or René du Perron as Descartes styled himself at thetime. Beeckman’s notebooks show that Descartes worked on a numberof such problems, set for him by his older friend, including the be-havior of water contained in a vessel, the behavior of a body in free-fall,


18 See the summary of the questions on which Beeckman worked between 10 November1618 and January 1619, when he was in contact with the young Descartes, as given in hisjournal, AT X 41–45. The very variety of the questions is very impressive. But it is alsointeresting to note the form of the articles in his journal. Most often the questions arequite specific and deal with specific phenomena: “candelarum scintillatio unde oriatur,”“cometarum caudae quid sint,” “aves cur in aere volare possint,” etc. There are, to be sure, ques-tions about motion and the laws of motion, but there are no questions about the natureof a natural power, as one finds in Rule 8. And, above all, there is little interest in anycomprehensive system encompassing all the sciences.

and numerous problems in music and geometry (AT X 46–78).19 It isnot surprising, then, that the method that Descartes first attempted toformulate in November 1619 and developed in the 10 years that fol-lowed, was a method for the solution of individual problems. To makeuse of the method, we must first set a specific question for ourselves,what is the shape of a lens with such-and-such properties, or, what causesthe rainbow, or whatever. Once we have a specific question, we can thenapply the method, reduce the question to simpler questions until wereach an intuition, and deduce back up to an answer to the questionoriginally posed. The method is a method for doing science as, say,Beeckman conceived of it, as a series of discrete questions about thenatural world.

But as I noted earlier in discussing the method of the Rules, themethod presupposes a certain conception of the structure of knowl-edge. All knowledge, for Descartes, is interconnected, grounded ulti-mately in a small number of intuitively knowable propositions fromwhich all else follows deductively. This, as I noted, was one of the thingsthat Descartes probably learned in that night of enthusiasm in Novem-ber 1619, and this is the key to the method he developed in the yearsfollowing. It is precisely because all knowledge is interconnected in thisway that the method is possible, that it is possible to take a question andreduce it to an intuition from which an answer could be deduced. Butthis very doctrine that makes the method possible leads to its demise.For if all knowledge is interconnected, then what we should be doingis not solving individual problems, but constructing the completesystem of knowledge, the interconnected body of knowledge that startsfrom intuition and comes to encompass everything capable of beingknown. Though he may have recognized this implication from the start,in 1619, it will be ten years before he begins such a system, in 1629with the first metaphysics, unfortunately lost, followed immediately bythe composition of the World. This project is what is striking and dis-tinctive about Descartes’ mature system, the system we find sketched inparts IV and V of the Discourse and developed in the Meditations andPrinciples of Philosophy. Unlike those of others, Galileo, for example (cf.AT II 380), Descartes’ strategy is to start not with individual questions,


19 According to Gouhier, it is also probable that a part of the Cogitationes Privatae containingthe Parnassus presents problems that Descartes discussed with Beeckman. See AT X219–48, and Henri Gouhier, Les premières pensées de Descartes (Paris: J. Vrin, 1958), pp. 15,24.

but to start at the beginning, with the intuitively graspable first princi-ples that ground the rest, and progress step by step from there down-ward to more particular matters. No longer a mere problem solver,Descartes has become a system builder.

But as a system builder, what role can he find for a method whosegoal is the solution of individual problems? With this crucial change inDescartes’ conception of scientific activity, a change motivated by thesame doctrine of the interconnection of knowledge that motivated hismethod, the method becomes obsolete; or if not obsolete, at very leastit is less central than it once had been.

This is one way in which the evolution of the Cartesian program ledto the demise of method. But there is another consideration as well.The method is a procedure for answering a question by deducing ananswer from intuition; it tells us how to find the appropriate intuition,and how to find a path from intuition to the answer we seek. But thisnaturally leads us to the question as to why we should trust intuitionand deduction at all, and why we should consider them to be the only source of knowledge. The history of Descartes’ struggle with thisproblem is very complex and I can only sketch briefly some of the mostimportant stages. The issue first arises in Rule 8, in what is probably thevery last stage of the composition of the Rules, just before Descartes setit aside in 1628 or so.20 There it appears as the “noblest example” ofthe method, something useful for preventing ourselves from attempt-ing to solve problems beyond our ability, or preliminary to the actualuse of the method in the same way that it is useful for the blacksmithto build sturdy tools before attempting to make horseshoes [AT X395–98].21 It is not altogether clear what status this investigation hadin 1628, whether it was a mere preliminary to investigation, or part ofthe system of knowledge itself, whether it is essential in order for us tohave any knowledge, or whether it is simply a practical suggestion aboutwhere we might begin. The status of this investigation of the grounds


20 See Schuster, “Mathesis,” pp. 58–59.21 In this very interesting but very complicated text, Descartes puts forward three distinct

and different versions of the project: (1) AT X 395, l. 17 to 396, l. 25; (2) 396, l. 26 to397, l. 3 or to 397, l. 26; and (3) 397, l. 4 or 397, l. 27 to 400 l. 11. In the first version,the project is described as the “most noble” example of the method (AT X 395). Theblacksmith analogy is used in a paragraph that could belong to either the second or thethird version. The differences between these three successive versions shows Descartes atthe very moment when he is launching his epistemological program, and where he isreflecting with great care about the exact formulation of the problem of knowledge. Butthis isn’t our task here.

of intuition and deduction, now clear and distinct perception, is alsodifficult to determine in the Discourse, where it appears to be somethingof a digression, part of the answer Descartes gives the reader whoobjects to the metaphysics presented earlier in part IV of the Discourse.If you continue to think, contrary to what I have written (Descartes says)that the material world of suns, stars, and tables is better known thanGod and the soul, reflect on the fact that if we did not know that Godexists and is not a deceiver, then we could know nothing at all (AT VI37ff.). But by the 1640s, the epistemological project, the investigationof the trustworthiness of intuition and deduction, clear and distinct perception has become the essential foundation of all knowledge; thetree of knowledge from the 1620s, grounded in the intuition of themost general notions concerning extension and thought, has grownroots, and it is essential for us to understand the foundation of ourbeliefs in God’s veracity for us to have any genuine knowledge at all(AT IXB 14).

But with this change, method by itself can no longer lead us togenuine knowledge. The reductive stage of the method starts with aquestion, and then takes us back to questions presupposed, until wefinally reach an intuition. But when the reduction has reached an intuition it goes no further. Thus the method of the Rules can at bestgive us imperfect knowledge, the moral certainty we get when we takeintuitions for granted, rather than the metaphysical certainty thatcomes from knowing that our clear and distinct perceptions are the creation of a God who does not deceive (AT VI 37–38; AT VII 141).

I have argued that two changes in Descartes’ thought conspired tomake the method of the Rules largely inapplicable to the system ofknowledge he hoped to build: (1) the change from a problem-solvingconception of scientific activity to a system-building conception; and(2) the adoption of the idea that intuition cannot be taken for grantedand must be validated, and that this is the essential preliminary to anysystem of knowledge. Given these features of Descartes’ mature systemof the 1640s, it is no wonder that Descartes came to have relatively littleuse for the method of the Rules, oriented to the solution of individualproblems, and incapable of leading us to metaphysical certainty.

But these considerations may also help to explain why the methoddoes not appear very much in the Essays either. Descartes suggests thathe does not use the method in the Essays because he did not want toreveal the foundations of his physics. But this cannot be the whole story.On the one hand, he was quite capable of using his method without


revealing any more of his foundations than he wanted to, as he did inthe rainbow example. And, on the other hand, even when he was notespecially worried about exposing the foundations of his physics, as inthe earlier World, method seems to play no substantive role. My own sus-picion is that many of the changes in Descartes’ thought that make theearlier method obsolete in the 1640s may also be present as early asthe first sketch of metaphysics Descartes attempted in 1629–30. Notthat Descartes was aware of what was going on. I suspect rather thatstarting perhaps as early as the winter of 1629–30 method is no longerrelevant to his scientific practices, and is simply not used in the project.And so, I suspect, when in the mid 1630s he sat down to gather togethersome of this material and present it in his Essays, the method had aslittle role to play as it did in its sources. But Descartes was perhaps notaware of the change his thought had undergone. And so, when he satdown to compose the preliminary discourse, out came the Discourse on Method, a work that expressed a conception of scientific inquiry that belonged to the earlier and somewhat more naive M. du Perron.Though cognizant of the fact that his Essays did not make much use ofthe method, he may not have realized why. This is a conjecture, ofcourse, and a very risky one. But it is indisputable that as his systemgrew, perhaps from the first metaphysics of 1629–30 onward, methodbecame, first in practice, and then after 1637 in theory, less and lessimportant to Descartes.

If I am right, then, the volume Descartes published 350 years ago in 1637 is a curious work, a beginning and, at the same time, an end.It is, of course, the beginning of Descartes’ public career, and it con-tains a preliminary sketch of the full system he will develop in suc-ceeding years, the interconnected body of knowledge grounded in firstphilosophy. But it also marks the end of a period, the last work in whichDescartes was to emphasize method as the key to knowledge. Descartesin 1637 is, in a sense, like the butterfly, emerging from his cocoon,spreading his new wings to dry in the sun, not yet fully aware that heis no longer a caterpillar.


3

A POINT OF ORDER

Analysis, Synthesis, and Descartes’ Principles

The serious student of Descartes’ philosophy must deal with the factthat Descartes’ metaphysics is presented in a number of different ways in a number of different works. While the Meditations ought to be regarded as the authoritative text, it is important to account for the sometimes significantly different versions of the philosophy thatDescartes presents in the Discourse, the Principles of Philosophy, the SearchAfter Truth, and in numerous remarks scattered throughout the corre-spondence. In this essay we shall examine one attempt to explain theprincipal differences between two of these works: the Meditations andthe Principles. It is often claimed that these differences can be explainedby the fact that the Meditations are written in accordance with the ana-lytic method, whereas the Principles are written in accordance with thesynthetic method. We shall argue against two somewhat different ver-sions of this thesis. Although we have no counter-thesis of comparablepower or simplicity to offer, we shall suggest some ways of understand-ing the relations between these two central works that better reflect thetexts and what appear to be Descartes’ intentions.

The main source for our understanding of Descartes’ distinctionbetween analysis and synthesis is the difficult thought often citedpassage at the end of the Second Replies (AT VII 155–156).1 In the Second

52

This essay was written jointly with Lesley Cohen.1 References to Descartes’ works will generally be given in the text. All translations are our

own.The technical terms “analysis” and “synthesis” come up very infrequently in Descartes’

writings. “Analysis” is mentioned in connection with the procedure of the Meditations inonly one other place, in the Fourth Replies (AT VII 249). All other appearances of the tech-

Objections, Descartes is requested to present his argument in more geo-metrico, with the full apparatus of definitions, postulates, and axioms(AT VII, 128). Descartes complies with this request in the GeometricalAppendix which follows his Second Replies where he provides a geometri-cal exposition of some of his arguments. But first Descartes gives ageneral discussion of the geometrical method of presentation. This dis-cussion begins with a distinction between two aspects (res) of the geo-metrical mode of writing (modus scribendi): ordo and ratio demonstrandi.Ordo, Descartes says, is simply the arrangement of material in such away that that which is presented earlier can be known without havingto appeal to that which follows. The terms “analysis” and “synthesis” areintroduced when Descartes attempts to distinguish between two differ-ent kinds of rationes demonstrandi that one could follow, presumablywithout violating ordo. Analysis is presented as the ratio which shows “thetrue way by which a thing was methodically and, as it were, a priori discovered [methodice & tanquam a priori inventa est]” (AT VII 155).Descartes’ account of synthesis is somewhat more complicated. Heexplains:

Synthesis on the contrary, clearly demonstrates its conclusions in an oppo-site way, proceding as it were a posteriori [tanquam a posteriori quaesitam](although the proof is here more often a priori than in the preceding case),and makes use of a long series of definitions, postulates, axioms, theorems,and problems. (AT VII 156)

In the Second Replies Descartes explicitly relates this distinctionbetween analysis and synthesis to his procedure in the Meditations.There he states:

In my Meditations I followed only analysis, which is the true and best way forteaching [via . . . ad docendum]. (AT VII, 156)

a point of order 53

nical terms are in mathematical contexts. See, e.g., AT II 22, 30, 82, 337, 394, 400, 438,637; AT III 99; AT VI 17–18, 20; AT X 373. For informal and non-technical uses of theterm “analysis” see, e.g., AT I 236–237; AT VII 444, 446. The only place in the corpus whereDescartes attempts explicitly to characterize the notions of analysis and synthesis and dis-tinguish between the two is in the passage from the Second Replies that we discuss. In thisessay, we shall be concerned with the notions of analysis and synthesis only insofar as theyhave been used by commentators to explain the differences between the Meditations andthe Principles. For more general historical accounts of analysis, synthesis, and the closelyrelated notions of resolution, composition, and method in general, see, e.g., J. Hintikkaand U. Remes, The Method of Analysis (Dordrecht: 1974); J. Hintikka, “A Discourse onDescartes’s Method,” in M. Hooker, ed., Descartes (Baltimore: 1978), pp. 75–88; and J. H.Randall, The School of Padua and the Emergence of Modern Science (Padua: 1961).

However, the Second Replies itself provides no direct evidence as to howthe Principles fit into the distinction drawn there. Although Descartesdoes present an example of synthetic argumentation in the GeometricalAppendix to the Second Replies, he does not mention the as yet uncom-pleted Principles in that connection. The only passage in the Cartesiancorpus in which there is a direct statement that the Principles are syn-thetic occurs in the Conversation with Burman. Burman raises a questionrelating to the two kinds of proofs for the existence of God offered inthe Meditations. In the course of his answer, Descartes points out thatin the Principles, unlike in the Meditations, the a priori argument pre-cedes the a posteriori arguments. The explanation Burman reports is this:

The way and order of discovery [via et ordo inveniendi] is one thing, that ofteaching [docendi] another; in the Principles he teaches, and proceeds synthetically. (AT V 153)2

There is some doubt about the reliability of this passage, as with all ofthe Conversation with Burman, particularly insofar as teaching is associ-ated with synthetic method here rather than with analytic method as itis in the unquestionably genuine Second Replies.3 But it does provide atleast prima facie evidence that Descartes thought that the Principlesare synthetic, and that he saw this as explaining at least some of the differences between that work and the analytic Meditations.

These observations, however, are of little use in understanding thedifferences between the two works in question until some furthercontent is given to the rather obscure distinction between analysis andsynthesis that Descartes offers in the Second Replies. One account of thisdistinction is offered by Martial Gueroult in his numerous influentialwritings on Descartes.4 According to Gueroult, the distinction between


2 It is interesting to note that this explanation for the divergence between the Meditationsand the Principles on this point is found in the literature on Descartes even before the firstpublication of the Conversation in 1896. See, e.g., Joseph Millet, Descartes, sa vie, ses travaux,ses découvertes, avant 1637 (Paris: 1867), pp. 216–217. Millet gives his account as if it werecommon knowledge, and offers no documentation.

3 For resolutions of this seeming inconsistency, see John Cottingham trans. and ed., Descartes’Conversation with Burman (Oxford: 1976), pp. 70–71 and Martial Gueroult, Descartes selonl’ordre des raisons (Paris: 1953 and 1968), vol. I, pp. 357–358, note 58.

4 See Descartes selon l’ordre des raisons (Paris: 1953 and 1968), vol. I, pp. 22–28, 357–360; Nou-velles réflexions sur la preuve ontologique de Descartes (Paris: 1955), pp. 17–20; and “La véritéde la science et la vérité de la chose dans les preuves de l’existence de Dieu,” in Descartes(Cahiers de Royaumont) (Paris: 1957), pp. 108–120, esp. pp. 112–117. This last paper is

analysis and synthesis is properly understood as a distinction betweentwo orders of presentation, namely the order of knowledge (ratiocognoscendi, la vérité de la science) and the order of being (ratio essendi, lavérité de la chose). The order of knowledge, or the analytic order, followsthe order of things as they are known. Consequently, an analytic pre-sentation of Cartesian metaphysics must, according to Gueroult, beginwith one’s own existence established by means of the Cogito, the firstthing which is known to us, and proceed from there to the existenceof other things, e.g., God and the material world, whose knowledgedepends on the knowledge of oneself. The order of being, or the syn-thetic order, on the other hand, proceeds in quite a different way asGueroult understands it, presenting things in an order that reflects the real dependencies that things have with respect to one another,independent of our knowledge of them. Consequently, on this understanding of the distinction, a synthetic presentation of Cartesianmetaphysics must begin not with the self and the Cogito, but with God,the real cause on which all else, including one’s own existence,depends.

Although Descartes himself never presents an account of the dis-tinction between analysis and synthesis in quite these terms, a plausiblecase can be made that this is what he had in mind. Descartes distin-guishes between the order of knowledge and the order of being in apassage from the Rules for the Direction of Mind which Gueroult oftencites as support for this position: “Individual things ought to be vieweddifferently in relation to the order they have with respect to our knowl-edge, than if we speak of them as they really exist” (AT X 418). WhileDescartes does not explicitly use the terms “analysis” and “synthesis” inthis connection, it is natural to associate this distinction between theorder of knowledge and the order of things with the distinctionDescartes draws between the two rationes demonstrandi in the SecondReplies, as Gueroult does. The order of things “with respect to ourknowledge” in the Rules seems exactly what Descartes is referring tosome years later when he characterizes the analytic ratio demonstrandi asshowing the “true way by which a thing is discovered.” While synthesisis not characterized in terms that directly suggest the order of being,

a point of order 55

followed by an interesting discussion (pp. 121–140) to which we shall later refer. The inter-pretation presented below is taken from the writings here cited. It is fair to say that the dis-tinction between analysis and synthesis as Gueroult draws it plays a central role in hiselaborate interpretation of Cartesian metaphysics.

there is nothing in the characterization Descartes gives in the SecondReplies which prevents identifying synthesis with order of being, thuscompleting the parallelism between the two passages.5 Such a conjec-ture would make reasonable sense of Descartes’ remarks as reported byBurman regarding the relative positions of the a posteriori and a prioriarguments for the existence of God in the Meditations and the Princi-ples. If a synthetic exposition is one that follows the order of being, thenone should expect a synthetic treatment of Cartesian metaphysics toput the a priori argument, which proceeds from the essence of God tohis existence, before the a posteriori argument, which proceeds froma particular idea we have to the existence of God as a necessary causeof that idea.

As elegant as Gueroult’s interpretation is, it unfortunately will notstand up to the actual texts. Gueroult’s thesis offers a plausible and intu-itively satisfying account of the different positions of the a posterioriand a priori arguments for the existence of God in the Meditations andthe Principles. However, his reading runs up against a basic similaritybetween the two works. Although the two presentations of the meta-physics differ with respect to many important details, the two works


5 Well, almost nothing. The somewhat peculiar language of the Second Replies does raise some-thing of a problem for relating those two passages and identifying analysis with the orderof knowledge and synthesis with the order of being, a problem that Gueroult does not dealwith. In the Second Replies, analysis is characterized as proceeding “tanquam a priori” and syn-thesis as proceeding “tanquam a posteriori.” But Descartes, like his contemporaries, identi-fied a priori arguments with arguments that proceed from cause to effect, and a posterioriarguments with arguments that proceed from effect to cause. See AT I 250–251, 563; ATII 433; AT IV 689; AT XI 47. And since causes are clearly prior to their effects in the orderof things, the Second Replies would thus seem to identify analysis with the ratio essendi andsynthesis with the ratio cognoscendi, exactly the opposite of what Gueroult claims! These pas-sages also raise a more general problem of interpretation. While Gueroult’s interpretationsof the terms in question are in apparent contradiction with the Second Replies, they are inaccord with the traditional understanding of those terms, in accordance with which analy-sis was almost invariably associated with a posteriori arguments from effect to cause, and syn-thesis with a priori arguments from cause to effect. See, e.g., Lisa Jardine’s discussion of theRenaissance uses of this terminology in Francis Bacon: Discovery and the Art of Discourse (Cam-bridge, England: 1974), pp. 249–250, and Louis Couturat’s discussion in La Logique deLeibniz (Paris: 1901), pp. 176–179. Thus, the obvious reading of the Second Replies makesDescartes’ usage of the terms “analysis” and “synthesis” radically at variance with the wayin which his contemporaries used them. For different resolutions of these problems, allfavorable to the Gueroult thesis, see F. Alquié, ed., Œuvres philosophiques de Descartes (Paris:1963–1973), vol. II, p. 582, note 1; J. Brunschwig, “La preuve ontologique interprétée parM. Gueroult,” Revue Philosophique 150 (1960), pp. 251–265, esp. pp. 257–259; and J.-M.Beyssade, “L’Ordre dans les Principia,” Les Études Philosophiques 1976, pp. 387–403, esp. pp. 394–395.

seem constructed on largely the same plan. Both works begin withdoubt, both proceed from there to the Cogito, from the Cogito to God,and from God to the external world. Given the similarities between thestructures of the two works, it is hard to understand how one could hold that one work follows the order of knowledge and the other workfollows the order of being. Something, it seems, must be wrong withGueroult’s reading; either analysis and synthesis are not connected withthe distinction between order of knowledge and order of being, or thePrinciples are not synthetic after all.6

However, it may be possible to retain the thesis that the Meditationsare analytic and the Principles synthetic if a different interpretation ofthese terms can be offered, one that is more consistent with the texts.Edwin Curley presents and argues for such an account in a recentpaper, “Spinoza as an expositor of Descartes.”7 Curley’s intuition issimple. We know that the Geometrical Appendix to the Second Replies is syn-thetic, and have good reason to believe that the Principles are as well. Ifwe are to discover what synthesis is and how it differs from analysis, thenthe question we must ask is clear: What do the Principles and the appen-dix to the Second Replies have in common that differentiates both ofthem from the analytic Meditations?

Approaching the problem in this way, Curley presents two featureswhich, he claims, differentiate synthetic works from analytic presenta-tions of the same material: the framing of “formal definitions of impor-tant concepts,” and the “prompt and explicit recognition of eternaltruths.”8 In the Meditations key concepts, like that of clarity and dis-tinctness, are introduced by examples, rather than by definition, as inthe Principles. And it is the Principles, not the Meditations, in whichDescartes seems to admit that the Cogito depends on the principle that

a point of order 57

6 A similar point is made by J. Brunschwig in “La preuve ontologique interprétée par M.Gueroult,” loc. cit. esp. pp. 255–257. Brunschwig’s arguments are attacked in B. Rochot,“La preuve ontologique interprétée par M. Gueroult (Response aux “Objections” de M.Jacques Brunschwig),” Revue Philosophique 151 (1961), pp. 125–130, and defended in J.Brunschwig, “Reponse aux objections de M. Rochot,” Revue Philosophique 152 (1962), pp.365–370. The question also arises in the discussion following Gueroult’s “La vérité de lascience et la vérité de la chose,” loc. cit., in remarks made by J. Hyppolite (pp. 125–126)and F. Alquié (pp. 134–135). Gueroult’s initial response is to say that the Principles “sontquelque chose d’un peu bâtard” insofar as they are really a mixture of analysis and synthesis(see pp. 126 and 137). This position is also endorsed by Henri Gouhier. See his La penseemétaphysique de Descartes (Paris: 1968), p. 109. J.-M. Beyssade, Op. Cit., works this positionout in some detail.

7 In Siegfried Hessing (ed.), Speculum Spinozanum (London: 1977), pp. 133–142.8 Ibid., pp. 136–137.

what thinks must exist. Curley’s basic strategy might be used to uncovereven further differences between the purportedly synthetic Principlesand Geometrical Appendix, and the analytic Meditations, yielding eventu-ally a rich and interesting account of the distinction between analysisand synthesis, an account that does not suffer from the problem wefound in Gueroult. Following Curley’s line of thought, one might pointout that the Meditations are written in the first person, while the Prin-ciples and the Geometrical Appendix are both written impersonally, or,perhaps more substantively, the Meditations can be differentiated fromthe purportedly synthetic works by virtue of the fact that in the Medi-tations, unlike the other two works, we find whole chains of reasoning,including false starts, heuristic arguments meant to motivate particularpremises, and strict arguments essential to establish conclusions. Thefirst causal proof for the existence of God as presented in Meditation IIIillustrates this well. The argument proper is preceded by an investi-gation based on the distinction between innate, adventitious, and factitious ideas, an argument that leads, unfortunately, to no certainknowledge (AT VII 37–40). The causal argument itself, when finallypresented, contains a number of lengthy subarguments. For example,Descartes gives a long heuristic argument to motivate the premise thatthere must be at least as much formal reality in the cause as there isobjective reality in the effect (AT VII 40–42). Also, the final conclu-sion, that God must exist as the cause of our idea of Him, is given onlyafter a lengthy enumeration of our ideas and their possible causes (ATVII 42–45). This contrasts radically with the presentation of the sameargument in the Principles and in the Geometrical Appendix. In both ofthese works, there are no false starts or dead ends, and little heuristicargument. The proof and its premises are presented unadorned andbare (see AT VII 167; AT VIIIA 11–12).

But despite the attractiveness of Curley’s account, one large difficultyremains. While Curley’s strategy is capable of yielding a plausibleaccount of the distinction that fits the texts, in the end it rests on an unstable foundation. While Curley shows us how the concepts of analysis and synthesis can be made to fit the Meditations and the Principles, neither he nor Gueroult has shown us why we ought to see the texts in that way. Neither has established with sufficient evidencethe basic premise in this exercise in interpretation, the claim thatDescartes really saw the distinction between analysis and synthesis as being relevant to the differences between the Meditations and thePrinciples.


No one, of course, can question the claim that Descartes wrote theMeditations according to the analytic ratio demonstrandi. He explicitlytells us he did this in the Second Replies. But the direct evidence thatDescartes wrote the metaphysical part of the Principles synthetically isvery weak. The only textual evidence for this claim comes from the Con-versation with Burman. But, it must be remembered, these words are not from Descartes’ own hand. They are filtered through Burman andalmost certainly through Clauberg, and clearly contain a number ofmistakes.9 Thus it is difficult to be sure that the particular wording ofany given passage represents Descartes’ intentions, particularly whenthe remarks relate to such an obscure point as the distinction betweenanalysis and synthesis. It is defensible to use that document to supportan interpretation drawn from more reliable texts. But it seems ques-tionable to use passages from the Conversation as the basis of an inter-pretation, which one must do if one is to maintain that the Principlesare synthetic.

In addition to the general concerns about the reliability of the Con-versation, there are some rather more specific reasons for questioningwhether Burman’s report is trustworthy on this point. Descartes, ofcourse, never directly says that the Principles are not synthetic, any morethan he says that they are, outside of the Conversation. But it does seemsignificant that in a number of contexts in which Descartes could quitenaturally have connected the Principles with the synthetic mode ofwriting, he does not.

Descartes’ correspondence allows us to trace out the history of thePrinciples and the Objections and Replies with some confidence. Descartesseems to have finished his manuscript of the Meditations by April of1640, for by 5 May 1640 he began to send it out for comment (AT III61). During the time he was putting the final touches on the Medita-tions, soliciting objections, and writing the replies that were to be pub-lished with them, he began to work on his Principles. The earliestreference to the Principles is in a letter written to Mersenne on 11November 1640, where he talks about his intention

a point of order 59

9 For recent discussions of the reliability of the Conversation, see F. Alquié (ed.), Œuvresphilosophiques de Descartes (Paris: 1963–1973), vol. III, pp. 765–767, Roger Ariew’s review of J. Cottingham (trans. and ed.), Descartes’ Conversation with Burman (Oxford: 1976), Studia Cartesiana 1 (1979), pp. 183–187, and Cottingham’s reply to Ariew, Ibid., pp.187–189. Ariew also shared with us his “Descartes Really Said That?,” given at the PacificDivision Meetings of the APA, March 1980. Curley discusses this question in Op. Cit., 140,note 9.

to write a completely ordered course of my philosophy in the form of theseswhere, without any excess of words, I will present only my conclusions alongwith the true reasons from which I derive them. (AT III 233; cf. AT III259–260)

By the end of December, it is clear that Descartes has actually begun towork on the first part, that which contains his metaphysics. He writesMersenne in a letter of 31 December 1640:

I have resolved to spend [this year] writing my philosophy in such an orderthat it can easily be taught. And the first part, that which I am working onnow, contains almost the same things as the Meditations which you have,except that they are in an entirely different style. (AT III 276)

It is only after the Principles were in progress that Descartes received theSecond Objections, the reply to which contains the discussion of analysisand synthesis. Mersenne promised to send them in December 1640 (ATIII 265), but Descartes does not seem to have received them untilJanuary 1641 (AT III 282). Descartes worked on the response throughJanuary and February (AT III 286, 293), and sent it to Mersenne byearly March 1641 (AT III 328). This raises a serious problem for thethesis that the Principles were intended to be synthetic: if Descartes wasalready well into the metaphysical sections of the Principles by the timethat he wrote the Second Replies, why does he not mention them? Afterdistinguishing between analysis and synthesis there, Descartes presents“a certain few things [from the Meditations] in synthetic style . . . fromwhich, I hope, [my readers] will get some help” (AT VII 159). IfDescartes really thought of his Principles as synthetic, it would have been very natural for him to have informed his readers that they couldexpect the whole of his metaphysics in synthetic style in a work then inprogress. That he does not mention the Principles in this connection issignificant.

It could be objected here that Descartes may not have wanted to publicize the Principles until they were further along. There is some-thing to this objection, to be sure. When Descartes first tells Mersenneof his new project in November and December of 1640, he does ask him to keep the project secret (AT III 233, 259). But Descartesseems to have changed his mind fairly soon. In the Fourth Replies, in apassage that was written by the end of March 1641, within a month ofthe completion of the Second Replies, Descartes refers to the work in


progress.10 If he was willing to refer to the Principles in answeringArnauld, it seems strange that he would neglect to mention them inthe discussion of analysis and synthesis in the Second Replies, if in fact hethought of the new work as being synthetic. Still more difficult toexplain is why, if he considered the Principles to be synthetic, Descarteswould have neglected to refer to them in the French translation of theSecond Replies, which appeared in 1647, three years after the Principleswere published. In the translation there is significant alteration of the sections of the Second Replies dealing with analysis and synthesis,doubtless with Descartes’s approval and probably from his own hand.After distinguishing between analysis and synthesis and before givingthe example of synthetic argumentation in his Geometrical Appendix,Descartes eliminates a large section of the Latin text and replaces itwith the following short paragraph:

But, nevertheless, to show how I defer to your advice, I shall try here to imitatethe synthesis of the Geometers, and make an abridgement of the principalarguments which I have used to demonstrate the existence of God and thedistinction between the human mind and body. This might perhaps serve tolessen the attention required of the reader a bit. (AT IXA 123; cf. AT VII157–159)

Surely, if Descartes really did think that the metaphysics was presentedsynthetically in the Principles, this would have been a perfect opportu-nity to tell his readers so, and refer them to that work. That he did not is at least some evidence that the Principles were not meant to be synthetic.

It is thus significant, we think, that Descartes does not mention thePrinciples when he talks about analysis and synthesis. But it is perhaps

a point of order 61

10 The reference to the Principles is given in AT VII 254. This reference, which is part of along discussion of transubstantiation, was not published in the Paris edition of 1641, andfirst appeared in the Amsterdam edition of 1642. There is strong evidence, though, thatit was written in March 1641. In a letter of 18 March 1641 Descartes refers to the lastsheet of his reply to Arnauld, “where I explicate transubstantiation in accordance with myprinciples,” as being in progress (AT III 340). It seems to have been finished and sent toMersenne by 31 March 1641 (AT III 349). Mersenne, though, suggested that he elimi-nate this passage in order more easily to obtain the approbation of the authorities, a sug-gestion that Descartes took (AT III 416). When the Paris edition appeared, the longsection on transubstantiation was reduced to a single sentence (given in the textual noteto line 21 in AT VII 252) which also contains a reference to his yet to be completed Prin-ciples. The full discussion was restored for the Amsterdam edition at Descartes’s request(AT III 449).

even more significant that he does not talk at all about analysis and syn-thesis when he discusses the relations between the metaphysics of theMeditations and the Principles, as he does on a number of occasionsoutside of the Conversation. Sometimes Descartes describes the meta-physics of the Principles as an “abrégé” of his philosophy (AT III 259;AT V 291; cf. AT IXB 16). Sometimes Descartes focuses on the fact thatthe Principles, unlike his previous writings, are written in short articles(AT VII 577), or that the work is a simplified version of his Meditations,containing only “my conclusions, with the true arguments from whichI derive them” (AT III 233). Sometimes he informs his correspondentsthat the principal difference between the two works is that “that whichis given at length in the one is considerably shortened in the other, andvice versa” (AT III 276). But nowhere in his correspondence or his pub-lished writings does Descartes ever mention the distinction betweenanalysis and synthesis in connection with the Principles. This would bevery strange indeed if Descartes really thought that the Principles weresynthetic.

Thus, it seems reasonable to deny that Descartes intended the Prin-ciples to be an example of the synthetic ratio demonstrandi. But in doingso, we do not want to assert that they are analytic either. The discussionof the Principles and their relation to the Meditations lacks any referenceat all to the distinction between analysis and synthesis. This strongly sug-gests that the distinction between analysis and synthesis may be entirelyirrelevant to understanding the true relations between the metaphysicalarguments of the Meditations and the Principles.

This position leaves us with a problem: If we cannot appeal to thedistinction between analysis and synthesis how, then, are we to under-stand the important differences between the two works? It seems to usthat there is no clear and simple answer to this question; Descartes’sown words and our common sense are all we have to rely on. The brevityof the metaphysical sections of the Principles may be attributed to thefact that Descartes conceived of Part I of the Principles as a preface to ascientific treatise, and not as a metaphysical treatise to stand on its own(cf. AT III 523; AT IXB 16).11 Similarly, certain other features of its


11 Given this, it might be interesting to compare the metaphysics of the Principles with theversion of the metaphysics presented in part IV of the Discourse, another work intendedas the preface to a scientific work. While the two presentations differ in many importantrespects, there are some striking similarities between the two. For example, both lack thehypothesis of the evil genius, and in both, the real distinction between mind and bodyseems to be proved before Descartes proves that God exists.

intended use may explain the use of explicit definitions and quasi-syllogistic argument in the Principles. Descartes’ hope that his Principlesmight be used as a textbook in the schools might have influenced himto set his arguments out in a more explicit way, more like a typicalscholastic textbook, than he did in the Meditations (see AT III 276; ATVII 577). Also, he seems originally to have conceived of the Principlesas part of a larger publication, which was to include an annotatedscholastic treatise on metaphysics, and an explicit comparison betweenhis philosophy and the philosophy of the schools.12 This may haveinduced Descartes to give explicit definitions and careful arguments, sothat the similarities and differences between his philosophy and that of the Scholastics would be more apparent to the reader (cf. AT III259–260).

These considerations do not explain all of the important differencesbetween the Meditations and the Principles by any means. For example,they cannot explain why Descartes orders the arguments for the exis-tence of God differently in the two works.13 Giving up the claim thatthe Principles are synthetic does make the commentator’s job somewhatmore difficult. But, it seems to us, nothing is gained by trying to explainthe differences between the Meditations and the Principles in termsforeign to Descartes’ own conception of their relations.14

a point of order 63

12 See AT III 233, 259–260. The text he mentions in this connection is Eustachius a SanctoPaulo’s Summa Philosophica, published first in Paris in 1609, but reprinted often through-out the 17th century. Descartes refers to this as “the best book that has ever been writtenon this material” (AT III 232; cf. AT III 251). Descartes abandoned this project in favorof a straight presentation of his own ideas in part because Eustachius’ death on 26 Decem-ber 1640 prevented Descartes from getting his permission to use his book in that way (ATIII 260, 286), and in part because he came to think that an explicit attack on the Scholas-tics was not needed (AT III 470).

13 It should be noted, in this connection, that even if one accepts the claim that the Princi-ples are synthetic, this difference between the Meditations and Principles is not easilyexplained. Curley’s account of analysis and synthesis, for example, seems to leave thisdivergence between the two texts unexplained.

14 We would like to thank Roger Ariew, Edwin Curley, Alan Donagan, Harry Frankfurt, andStephan Voss for helpful discussions and correspondence concerning the matters dis-cussed in this essay.

4

J.-B. MORIN AND THE SECOND OBJECTIONS

Of the seven sets of objections to the Meditations, two stand out as beinga bit different, the Second and the Sixth. In every other case we canidentify one person, a philosopher or a theologian, who is the authorof those objections. In the case of the Second and the Sixth, though,we are dealing with objections that have been collected by one person,the ever-present Father Mersenne, but that purport to represent thework of a number of other scholars, who remain unidentified. For mostpurely philosophical purposes, this does not matter a great deal; afterall, an idea is an idea, whoever happens to have it, and if what is impor-tant is just the confrontation of ideas with one another, then the par-ticular identity of the authors in question, those who contributed tothese two sets of objections, is relatively unimportant.

But for those of us with a more historical approach to the texts, thisis an unfortunate gap. First of all, it is intrinsically interesting from anhistorical point of view to know who may have contributed to the draft-ing of these objections. But more important, in order to understandthe objections, their meaning and import, it is very important to knowsomething about their authors. In particular, I shall argue that, behindthe scenes in the Second Objections and Replies, there is not merely an author but a text that is important for understanding what Descartesis doing, a text that is implicitly referred to in the Second Objections

64

Work for this essay was supported by the National Endowment for the Humanities, an inde-pendent Federal Agency, under grant RH-20947 and by the National Science Foundationunder grant DIR-9011998. I would like to express my sincere thanks to those agencies fortheir kind support.

and is the direct object of Descartes’ reply in the geometrical presen-tation of the arguments that follows the Second Replies.

The author in question, I claim, is Jean-Baptiste Morin, astrologer,physician, and professor of mathematics at the Collège Royal, and thetext in question, his Quod Deus sit, a small tract, published in Paris in1635, in which he presents an argument for the existence of God ingeometrical form, with definitions, axioms, and a string of theorems. Iwill begin with a brief biographical sketch of Morin, one of the morecurious savants of his time. Then I will discuss his relations withMersenne and Descartes and make the case that he was behind at leastcertain portions of the Second Objections to the Meditations. Finally, I will discuss Descartes’ reaction to Morin’s pamphlet, and Morin’s laterreaction to Descartes, concentrating in both cases on the question ofthe geometrical presentation of metaphysics, Morin’s advocacy, andDescartes’ critique.

Jean-Baptiste Morin

Jean-Baptiste Morin was born on February 23, 1583, in Villefranche-en-Beaujolais, which made him just a bit more than thirteen yearsDescartes’ senior.1 Morin’s early years were not easy; illness and thenecessity of earning his keep prevented him from pursuing the studiesin natural philosophy that interested him from his earliest years. (Oneinteresting event is a grave injury he suffered in 1605, at the age oftwenty-two, mulieris causa, causing him to flee Villefranche. Though he never married, Morin seemed always to have had a weakness for the ladies.) Finally, in 1608, at the advanced age of twenty-five, he ob-tained the protection of Guillaume du Vair, Premier-Président of theParlement d’Aix, who enabled him to have lessons in mathematics, thenhelped him resume his studies, first in philosophy, then in medicine.

morin and the SECOND OBJECTIONS 65

1 The biographical sketch that follows is largely taken from an anonymous biography thatappeared not long after Morin’s death, La vie de Maistre Jean Baptiste Morin . . . (Paris, ChezIean Henault, 1660). The editors of the Mersenne correspondence identify the author asGuillaume Tronson; see Marin Mersenne, Correspondence du P. Marin Mersenne, religieuxminime, ed. C. de Waard et al., 17 vols. [Paris: Beau-Chesne (vol. 1), Presses Universitairesde France (vols. 2–4), CNRS (vols. 5–17), 1932–1988], vol. III, pp. 127–28. I also madeuse of the excellent biblio-biographical sketch by Monette Martinet, “Jean-Baptiste Morin(1583–1656),” in Pierre Costabel and Monette Martinet, Quelques savants et amateurs descience au XVIIe siècle: Sept notices biobibliographiques caractéristiques (Cahiers d’histoire et de philoso-phie des sciences, NS no. 14). (Paris: Société Française d’Histoire des Sciences et des Tech-niques et Éditions Belin, 1986), pp. 69–87.

Morin graduated from Avignon in May 1613. Shortly thereafter he wentto Paris, where he entered the service of the bishop of Boulogne,Claude Dormy, as physician. Dormy encouraged Morin’s studies ofastrology and alchemy, and he sent Morin on a journey of discovery tothe mines of Hungary and Transylvania, a trip from which resultedMorin’s first book, Nova mundi sublunaris anatomia, an account of theinterior of the earth, published in 1619 and dedicated to his formerpatron du Vair. While in the mines, Morin had noticed the unusualheat, and he wrote the book to offer an astrological explanation of it,referring to the influence of the stars.

Morin even then had astrological inclinations, to be sure. But uponhis return to Paris, they were strengthened. Morin made the astrologi-cal prediction that Dormy was in danger of death or imprisonment.Sure enough, Dormy was carted off in 1617, though given Dormy’sinvolvement in court politics, one would not have to have been a master astrologer to have made that prediction. But Morin was furtherconfirmed in his vocation, and he went on to make numerous cele-brated predictions, some of which were actually borne out.2 AfterDormy, Morin passed first to the patronship of the abbé de la Bretonnière, with whom he spent four relatively quiet years. In 1621he passed on to the service of the duc de Luxembourg, brother of the favorite of the king, the duc de Luynes, who soon fell from favor.During this period Morin composed a number of works, including twoastrological tracts and, in 1624, an interesting pamphlet attacking agroup of young scholars who had announced a public disputation in which they proposed to refute the foundations of Aristotle’s naturalphilosophy and replace it with a form of atomism.3 In the latter tractMorin came out in favor of form and matter and against atoms –indeed, against innovation in natural philosophy in general. In theastrological tracts he came out against Copernicus and in favor ofTycho, though in general he showed his interest in bringing astrology


2 See La vie de . . . Morin, pp. 62–91 for an account of Morin’s predictions. See also the entertaining, though not altogether reliable account in Anne Soprani, Les rois et leurs astrologues (Paris: MA Editions, 1987), pp. 175–82. Morin’s most disastrous prediction was the incorrect prediction of the imminent death of Gassendi in the course of a pamphlet war.

3 The two astrological works are: Astrologicarum domorum cabala detecta (Paris, 1623) and Ad australes et boreales astrologos; pro astrologia restituenda epistolae (Paris, 1628). The polemi-cal work is Réfutation des thèses erronées d’Anthoine Villon . . . & Estienne de Claues . . . (Paris,1624).

up to date by making it consistent with the latest discoveries in obser-vational astronomy, in particular the discovery and mapping of thesouthern skies.4

Though Morin believed in the guidance of the stars, he did not leavehimself to their care alone. From all evidence, he was a firm believerin freedom of the will and the value of self-promotion. And so, when achair of mathematics came open in 1629, he made himself a candidate,and with the influence of the cardinal de Bérulle and the QueenMother, he received the appointment, which he held until his death in1656. (The documents I have read imply that it was not his mathe-matical talent alone that won him the chair.5) It was during this periodthat Morin wrote most of his voluminous writings. Two extended dis-putes stand out. The first concerned a scheme that Morin proposed asearly as 1633 for determining the longitudes of vessels at sea, a dis-covery that Morin hoped would win him a pension from CardinalRichelieu. Though the method worked in theory, it turned out to benot altogether practical, and despite roughly fifteen years of pleas andpamphlets, Morin never got his pension. The second large controversywas with Pierre Gassendi, over Morin’s critique of Gassendi’s atomismand Gassendi’s critique of Morin’s astrology, a dispute that began inearnest in the late 1640s. Also important from these years are twovolumes attacking Copernican astronomy and the Quod Deus sit of 1635,to which we shall return. Through these years, though, starting in theearly 1630s and extending up to the time of his death, Morin wasworking on his magnum opus, the Astrologia gallica, a Summa astrologica,as it were, a work that summarized his career as a natural philosopherand astrologist, published at The Hague in 1661, five years after hisdeath. Though it was directed mainly at outlining his astrologicalsystem, this thick tome begins with a proof for the existence of God (alater version of his Quod Deus sit) followed by a series of books onnatural philosophy, laying the foundations for the more properly astro-logical questions to follow. The natural philosophy that Morin outlines


4 For an account of Morin’s progressive astrology, see Wilhelm Knappich, Histoire de l’astrologie(Paris: Editions du Féslin, 1986), pp. 229–33.

5 See Bérulle’s letter to Richelieu, Corr. de Mersenne, vol. III, pp. 501–2. On the objections toMorin’s elevation to the Collège Royal, see, e.g., M. L. Am. Sédillot, Les professeurs de mathématiques et de physique générale au Collège de France (Rome: Imprimerie de Sciences Mathématiques et Physiques, 1869), p. 101. Abuses of this sort seem to have been quitecommon at the Collège Royal in the early seventeenth century; see Claude-Pierre Goujet,Mémoire historique et littéraire sur le Collège Royal de France (Paris, 1758), vol. I, pp. 206ff.

there is definitely conservative. He explicitly attacks mechanist andatomist views like those of Descartes and Gassendi. While there aresome modern elements – a theory of space that looks as though it isderived from Patrizi, for example – Morin grounds his physics in thetheory of substantial forms.

But yet in a sense Morin considered himself a sort of progressive.When Descartes’ Discourse and Essays came out in 1637, Morin was oneof the savants who received a copy. Descartes had hoped to collect avariety of responses to his first publication and publish them togetherwith his responses, much as he was later to do with the Meditations. Thewhole exchange is quite interesting, and I have discussed it elsewhere.6

For the moment, I would like only to note something that Morin saidin his first reply to Descartes. In his letter of February 22, 1638, Morinwrote: “I do not know, however, what I should expect from you, for Ihave been led to believe that should I discuss matters with you usingthe terms of the Schools ever so little, you would immediately judge memore worthy of scorn than of response. But in reading your discourses,I find that you are not as much of an enemy of the Schools as you aremade out to be.” Morin continues with some remarks about his ownview of the Schools: “The Schools appear to me to have erred onlyinsofar as they are more occupied in speculation directed toward thesearch for the terms that we must use to discuss things, than they arein the search for the truth itself about things through good experi-ments; so they are poor in the latter and rich in the former. That is whyI am like you in this matter; I seek the truth about things in naturealone, and I no longer put my trust in the Schools, which serve me onlyfor terminology.”7

This may strike us as something of a distortion, and it is, in a sense.But one can also see what Morin means. I mentioned Morin’s trip ofdiscovery to Hungary and Transylvania to visit the mines. In the prefaceto his Anatomia of 1619 Morin discusses the motivation for his explo-rations. The great diversity of opinion among the learned forms moreof an obstacle to learning than a help. And so, he argues, we must turnaway from books and to nature itself to discover how things really are.In doing so, Morin thinks that he has found an account of the makeupof the earth that is utterly unknown among the philosophers of the


6 See Daniel Garber, “Descartes, the Aristotelians, and the Revolution that Didn’t Happen in1637,” The Monist, 71 (1988), pp. 471–86.

7 AT I 541.

Schools.8 I am sure that Morin saw this as exactly parallel to Descartes’rejection of learning and his travels in search of experience. LikeDescartes, Morin professes to be following reason, not authority. In theAstrologia gallica he writes: “In what is said below we shall not follow thedoctrines of the Schools, which are often in error, but we shall look tothe nature of things, which alone contains the truth.”9 Even his treat-ment of astrology shows his open-mindedness. Though he agrees withthe tradition that the stars influence what happens here below, he isnot dogmatic about the details and sees the need to revise traditionalastrological doctrines in the face of newly discovered astronomical facts.His is a progressive astrology, so to speak.10

At the same time, Morin’s instincts are undeniably conservative. In doctrine, he follows Aristotle and opposes atomism and Coperni-canism; at root, the traditional philosophy is right, if not in every detail. He is conservative in other respects too. A social climber of sorts, always looking out for a way to advance himself socially and financially, he vigorously opposes challenges to the institutionswhose support and patronage he constantly sought. This, I think, is at least in part behind the vigor of the attack he made in his relative youth against a motley crew of anti-Aristotelians in 1624, insupport of the government’s condemnation and exile of three young scholars who proposed publicly to refute Aristotle, along withParacelsus and the Cabala.11 Though he considers himself open-minded, he has clearly hitched his star to the traditional philosophy of the Schools.


8 See Morin, Nova mundi sublunaris anatomia (Paris: 1619), dedication (unpaginated), letterto the reader, and, esp. chapter 5. Morin opens the letter to the reader with a frank dec-laration of the novelty of his view: “Hic habes . . . Novam Mundi sublunaris divisionem,novas divisionis causas, novaque de rebus physicis disserendi fundamenta.”

9 “Neque in infra dicendis sequemur doctrinas scholarum quae frequentius fallunt, sednaturam rerum spectabimus, quae sola veritatem continet.” Astrologia gallica (The Hague,1661), p. 39.

10 This is the main project of Ad australes et boreales astrologos of 1628. Of particular concernto him there are the recent observations of the southern sky, and how they affect astrology.

11 This is quite evident in the 1624 pamphlet, Réfutation des thèses erronées . . . , where he com-plains more than once of the arrogance of the attack on Aristotle in the great city of Paris:“Ils [i.e., Villon and de Clave] affichent . . . un defi publique à toutes les Escoles, sects &grands Esprits . . . Et cecy non dans un village, mais dans une ville de Paris, à la face de laSorbonne, de toute l’université, & du plus fameux Senat qui soit au monde” [p. 6]. Moringoes on to say that one of the reasons why he is attacking Villon and de Clave publicly is“pour l’honneur de ceste Cité tres celebre de Paris” [pp. 19–20].

Morin, Mersenne, and Descartes

In general I find Morin to be a fascinating character. While his instinctsare conservative, he is not a dogmatic Schoolman (I wonder whetheranyone really was), and while he is a sort of progressive, he is noDescartes or Gassendi. But interesting as it would be to continue thisdiscussion of Morin’s life and works, we must turn now more specifi-cally to his relations with Descartes and with Mersenne, Descartes’friend and the collector of the Second and Sixth Objections.

By the time Descartes had finished the Meditations and begun to cir-culate it for comments, Descartes, Morin, and Mersenne had knownone another for quite some time. Mersenne, too, had opposed the anti-Aristotelians of 1624 in print, as Morin had, and no doubt they becameacquainted then, if they did not know one another before.12 Descartes’acquaintance with Morin is usually dated from 1626 or 1628, andMorin is known to have helped Descartes get a piece of optical equip-ment made in the late 1620s.13 Though Mersenne always opposedastrology and came to support Copernicanism, he always seemed toconsider Morin a friend. When Morin was about to publish his argu-ments against Copernicus, Mersenne, along with Gassendi, counseledhim against publishing the book, but after it was published, Mersennedid not disown him.14 He even sent the book to Descartes, whose reac-tion was, predictably, caustic.15 Still, Descartes and Mersenne soughtMorin’s opinion of Descartes’ Discourse and Essays when the book cameout in 1637. In this interesting exchange, Mersenne acted as a sort ofintermediary. He was with Morin when he received Descartes’ firstletter, in reply to Morin’s, and he continued to encourage Morin tothink of Descartes as someone knowledgeable in the School philosophyand not unsympathetic to it, and to encourage Descartes to continueto give Morin this (false) impression.16

Morin was clearly a friend of Mersenne’s. He may not have been a


12 See Marin Mersenne, La vérité des sciences (Paris, 1625), pp. 76–84 and 96–113.13 See Charles Adam, Vie et oeuvres de Descartes (Paris: L. Cerf, 1910) p. 90, and Corr. de

Mersenne, vol. II, p. 420. There are references to Morin in letters of 1629 and 1630; seeAT I 33, 124, 129–31, etc. In a letter from 22 February 1638, Morin begins by recallinghis earlier acquaintance with Descartes in Paris, presumably before Descartes left forHolland in 1628; see AT I 537.

14 See Gassendi to Joseph Gaultier, 9 July 1631, Corr. de Mersenne, vol. III, p. 173.15 See Descartes to Mersenne, Summer 1632 (?), AT I 258.16 On this, see Garber, “Descartes, the Aristotelians. . . .”

member of the circle who met regularly in Father Mersenne’s room inthe Minim Couvent near the Place Royale.17 He is, though, certainly areasonable candidate for membership in the anonymous group ofpeople who contributed to the Second and Sixth Objections, whichMersenne collected.18 But one can go further than that. It can be estab-lished with reasonable certainty that Morin was a part of that group ofobjectors, and something plausible can be said about which specificobjections he might have contributed to the enterprise.

Most relevant here is Morin’s short treatise Quod Deus sit of 1635.Briefly (we shall look into it more carefully below), Morin’s book is anargument in geometrical style for the existence of God.19 Starting witha series of formal definitions and axioms, the book comprises thirty theorems purporting to establish the existence of God and his relationto the world. Mersenne knew this book and seems to have thought wellenough of it to call it to Descartes’ attention. He sent Descartes a note,now lost, apparently summarizing one of the arguments of Morin’sQuod Deus sit; Descartes responded on November 11, 1640.20 Descartes’response to Mersenne contains a critique of an argument that does notcorrespond exactly to anything that I can find in Morin’s book itself,and it is impossible to evaluate Descartes’ criticism without seeingexactly how Mersenne represented the argument. But the book itselffollowed shortly thereafter; it came from Mersenne via Huygens andarrived on January 21, 1641.21

Later we shall look more carefully at Descartes’ response. But for themoment I would like to note only that Morin’s name, and the name of his pamphlet, almost certainly appeared in the first version of the


17 So argues Bernard Rochot; see his comment in Corr. de Mersenne, vol. X, p. 410n. I’m notsure that Rochot is right.

18 There may be some precedent for Morin and Mersenne collaborating on a critique ofDescartes. Baillet reports that some of the objections to the Discourse and Essays Morinsent Descartes may actually be due to Mersenne: “Le Père Mersenne sembloit avoir jointquelques-unes de ses difficultez avec les objections de M. Morin” [Vie de M. Descartes (Paris,1691), vol. I, p. 356].

19 For a detailed discussion of this text, see Joseph Iwanicki, Morin et les démonstrations math-ématiques de l’existence de Dieu (Paris: Vrin, 1936). Iwanicki offers a good discussion of thetexts, the arguments, Descartes’ critique of the arguments, and the history of geometri-cal arguments for the existence of God; he also notes the connection between Morin andthe Second Objections. However, Iwnaicki sees himself as rather an advocate for the histor-ical and philosophical importance of Morin, and winds up greatly exaggerating Morin’splace. Indeed, he concludes rather implausibly that it is Morin, not Descartes, that hasthe best of their exchange on proofs for the existence of God.

20 AT III 233–34. 21 AT III 283.

Second Objections that Descartes received. Writing to Huygens onJanuary 16, 1641, Descartes notes: “I have been very eager to see thebook, Quod sit Deus [sic], because it is cited in the objections that FatherMersenne wrote you that he would send me.”22 The point at which itmay have been cited is relatively easy to determine. At the end of theSecond Objections we find the following passage: “After giving your solutions to these difficulties it would be worthwhile if you set out theentire argument in geometrical fashion, starting from a number of defi-nitions, postulates, and axioms. You are highly experienced in employ-ing this method, and it would enable you to fill the mind of each readerso that he could see everything as it were at a single glance, and be per-meated with awareness of the divine power.”23 While this passage doesnot make direct reference to Morin and his book (no passage in thefinal published text does), it seems quite plausible that this is thepassage to which Descartes refers in his letter to Huygens. It is not alto-gether clear why Morin’s name was dropped from the final publishedtext. It could be that Descartes made it a general policy not to mentionliving authors by name. Another factor may have been the fact thatDescartes was not impressed with the book. Rather than saying some-thing uncomplimentary about Morin, thus offending him, somethingthat he explicitly told Mersenne he did not want to do, Descartes mayhave decided to drop the reference altogether.24 Perhaps, too, he didnot want to start the sort of pamphlet war in which the somewhat iras-cible Morin had been known to engage with relish. Be that as it may, itseems reasonably certain that Morin and his Quod Deus sit stand behindthis passage of the Second Objections.

With a little imagination, we may also be able to see Morin’s handin other passages of the Objections that Mersenne is known to haveassembled. Morin’s Astrologia gallica, a vast, encyclopedic work, beginswith a full account of metaphysics and physics, in order to ground theastrology in the later sections. Scattered throughout these introductorysections are passages criticizing Descartes, both his metaphysics and hisphysics. While Descartes’ name occurs often in these pages, especiallyinteresting is a critique that appears at the very beginning of the book,in a section entitled “Liber primus: De Vera cognitione Dei ex lumineNaturae; per Theoremata adversus Ethnicos & Atheos Mathematicomore demonstrata.” This, not unsurprisingly, is an expanded version of


22 AT III 765–66. 23 AT VII 128.24 See Descartes to Mersenne, 28 January 1641, AT III 294.

the pamphlet of 1635, the Quod Deus sit. (The expanded version hadbeen published in 1655 as a separate pamphlet.) As we shall see, anumber of interesting changes were made to the 1635 version, perhapsas a response to Descartes’ criticism of that earlier work.

The most significant change is in the preface. While much of thepreface of the 1635 version is retained, Morin added much new mate-rial, in fact a long and rather critical discussion of Descartes’ meta-physics and his proofs for the existence of God. What is interesting hereis a certain correspondence between the criticisms in this new prefaceand the contents of the Second Objections. Morin refers to the SecondObjections and Descartes’ replies a number of times, and to other sets of objections only rarely. In Astrologia gallica the first serious objectionto the Meditations is the objection that the certainty of the cogito pre-supposes the proof for the existence of God. This, in essence a versionof the circle objection that Arnauld brings up in the Fourth Objectionsin a somewhat more direct way, is the third point in the Second Objec-tions.25 Morin objects to the innateness of the idea of God andDescartes’ causal principle, which corresponds to the second point inthe Second Objections.26 Morin also challenges Descartes’ version of the ontological argument, a challenge that appears as the sixth pointin the Second Objections.27 Now, Morin does not bring up other pointsfrom the Second Objections, and he does bring up a number of other crit-icisms that are not found anywhere in the Objections, either in the Second Objections or in any others. But the correspondence betweenMorin’s later criticisms in the Astrologia gallica and specific sections ofthe Second Objections suggests to me that Morin may well have con-tributed those objections to the pool, and that even if Mersenne wrotethe final text, Morin may have formulated the criticisms. (It is inter-esting that there is no correspondence at all between Morin’s critiquein the Astrologia gallica and the Sixth Objections. This suggests to me that Morin probably had no hand in the later set.) There is a certainamount of conjecture in my claim that Morin may have been responsi-ble for these parts of the Second Objections, and I do not want to insiston it. Morin was self-important enough that I suspect he might havetold his readers that he was behind those parts of the Second Objec-tions if indeed he was. In the Astrologia gallica, for example, he certainly


25 Morin, Astrologia gallica, p. 2; AT VII 124–25.26 Morin, Astrologia gallica, p. 4; AT VII 123–24.27 Morin, Astrologia gallica, p. 5; AT VII 127.

claims credit for having elicited the geometrical appendix to the Second Replies that is the answer to the request at the end of the Second Objections for a geometrical development of the main argu-ments.28 So while I think that it is quite possible that Morin did con-tribute these objections, it is by no means certain.

Descartes’ Critique of Morin

I have set out the facts, so far as I can establish them. Morin standsbehind the Second Objections, in a sense; his Quod Deus sit was known to Descartes, and Morin is probably responsible for the request thatDescartes present his arguments in geometrical form. Furthermore, itis possible (though not certain) that Morin made other contributionsto the Second Objections, which Mersenne assembled. These are the facts. But what about the philosophy? While a great deal could be saidhere, I would like to concentrate on the issues raised by the geometri-cal proof for the existence of God that Morin offers in the Quod Deussit, by Descartes’ response to this work, both in his letters and in theSecond Replies, and by Morin’s response to Descartes’ response to hisQuod Deus sit in his Astrologia gallica.

From the introduction to Morin’s booklet, his letter of dedication tothe Sacra Comitia Gallica, the assembled clergy of France, who gath-ered in Paris in 1635, one can see why Mersenne would have foundMorin’s project sympathetic.29 In that introduction, Morin explicitlyopposes himself to “the unchecked sect of the atheists, who now havebecome so haughty.”30 He does so by appealing to the certainty of mathematical method to prove to atheists that they are constrained toaccept the existence of God. He writes:

I have never doubted that one could show, not what God is, but that he is byway of the most evident light of nature alone. I also grasped that the great-est good deriving from this lies in the fact that the natural light still remains


28 See Astrologia gallica, p. 5. See also Morin’s Defensio dissertationis . . . (Paris, 1651), p. 90,where he indirectly implies that he was behind Mersenne’s question to Descartes: “Carte-sius fuerit provocatus a . . . R. P. Mersenno, ut simili methodo conaretur demonstrare existentiam Dei, . . . nec tamen viris doctis satisfaceret.”

29 There is also a dedication to the same group in the second edition of 1655, on the occa-sion of the next meeting of the group in Paris. Bayle suggests Morin had hoped to gaina pension from that group; see Pierre Bayle, Dictionaire historique et critique (Amsterdam:1720), vol. III, p. 2015, art. Morin, note H.

30 Morin, Quod Deus sit (Paris: 1635), p. 4.

to the atheists, though they resist it, and with the help of that alone, theyremain capable of grasping the first principles of nature, which they cannotfail to perceive even while they are denying them, because they are the per seobjects of that [natural] light. Consequently, at the very least this path fordiscussing the existence of God with atheists is open to us, so that they mightknow their greatest error. Therefore, having undertaken this task for theglory of God, for the confirmation of faith, and to return the atheists to theirsenses, using a mathematical method, I carried it out to such an extent thatonce they concede those things I laid down as principles, perceptible by thelight of nature alone, atheists cannot deny that God exists, that he createdthis world in time, and that he governs it by his providence, unless they them-selves also deny that they exist.31

(Morin, like Descartes, appeals frequently to the light of nature.)Morin’s introduction resembles Mersenne’s project in the commentaryon Genesis from 1623, which begins with a ferocious attack on the athe-ists of his day, followed by a multitude of arguments for the existenceof God drawn from every conceivable premise.32 It also resembles theletter of dedication to the doctors of the Sorbonne, which precedesDescartes’ Meditations, emphasizing the necessity of refuting atheism byproving God’s existence.33 Like Descartes, Morin argues that knowledgeof the existence of God is foundational for all other knowledge. Heclaims that many other important theorems can be derived from theones he gives. Indeed, Morin claims: “It is not difficult to extend theprinciples I posit, and the theorems I set out to many other wonderfultheorems about God and his creatures; indeed, . . . using the samemethod, one can prove the immortality of the soul and all the naturalsciences.”34 In this respect Morin seems to resemble Descartes and theproject of the Meditations. One can see why Mersenne would havethought that Descartes would find this sympathetic. Mersenne, insending him the booklet, no doubt thought that he would draw Morinand Descartes closer together; it is interesting (and perhaps revealingof Mersenne’s character) that he miscalculated so badly. Despite theirsuperficial similarity, Descartes found Morin to be quite a differentkettle of fish.


31 Ibid., pp. 5–6.32 Mersenne, Quaestiones celeberrimae in Genesim (Paris, 1623). The proofs for the existence

of God begin on col. 25, and the subject isn’t set aside until cols. 669–74, where Mersennepresents a long diatribe against atheism. In all, Mersenne offers 36 arguments for the existence of God.

33 AT VII 1–2. 34 Morin, Quod Deus sit, p. 7.

Morin’s Quod Deus sit contains a total of thirty theorems, but the argu-ment for the existence of God is really quite simple and can be foundin Theorems 14–16. Theorem 14 reads: “Omne ens finitum habet esseab Ente infinito,” “Every finite thing has its being from an infinitething.”35 Morin offers two proofs for this theorem. The first is a directregress argument. “Whatever there is must derive either from itself [sitseipso] or have its being from something else.” This is one of Morin’saxioms. But as a finite thing, something cannot derive from itself(“nullum ens finitum est seipso”); this is true by Morin’s Theorem 12.Since there cannot be an infinite regress of causes or a circle (Prop.13), there must be, somewhere in the series, an infinite cause. Thusevery finite thing must have its being from an infinite cause, eitherdirectly or mediated by other finite causes. The second proof is some-what different. Morin begins with a curious proof that would seem toestablish that there can only be a finite number of finite things. Con-sider the number of people. Suppose that it is infinite. Then, Morinargues, it will contain all people who were, are, and will be, by his Axiom10 (“There can be nothing greater or larger than the infinite, nor canany such thing be conceived”36). But, Morin notes, experience showsthat new people are born every day. This, he infers, could not happenif the number of people were infinite, since, presumably, one cannotadd anything to a number that is already infinite. Thus, the number ofpeople must be finite.37 But since each finite thing needs a cause, itfollows that there must be something that is not finite that is the causeof everything else. And so, again, every finite thing has its being froman infinite cause.

Theorem 15 then establishes that “mundus hic finitus est,” “thisworld is finite.” The principal argument for this conclusion is groundedon Morin’s refutation of the Copernican claim and his view that theearth is at rest in the center of the universe. Now assume that the uni-verse is infinite. If so, the universe would occupy an infinite space. Sincethe universe turns around the earth once every twenty-four hours, itwould then follow that matter would traverse an infinite space in a finiteamount of time, which is absurd. And so, Morin concludes, the universemust be finite.

From these two theorems, it follows directly that there must be aninfinite being. For if finite things have their being from something


35 Ibid., p. 15. 36 Ibid., p. 10.37 Ibid., pp. 16–17. A similar argument is found later, in Theorem 17.

infinite (Theorem 14) and if this universe is finite (Theorem 15), thenthere must be an infinite being. This is Morin’s Theorem 16.

The earlier theorems deal with more general questions about infi-nite and finite beings. Before establishing that God exists, Morin estab-lishes, for example, that the infinite being is purus actus (Theorem 2),that there are not two infinite beings (Theorem 5 – shades of Spinoza),that the infinite being is indivisible and simple (Theorems 6 and 7),and so on. After establishing that the infinite being exists, Morin estab-lishes that the infinite being produced everything by a simple act of will(Theorem 21), that the infinite being continually produces and con-serves the finite beings he produces (Theorem 22), that the world wascreated in time (Theorem 27), and, finally, that the infinite being is theultimate end ( finis) of all finite beings (Theorem 30).

Needless to say, Descartes was not altogether impressed with this. Hedidn’t expect much to start with. His dealings with Morin on the subjectof his Dioptrics mostly left him unimpressed. Writing to Mersenne onDecember 31, 1640, Descartes noted: “I would not be unhappy to seewhat M. Morin has written about God because you say that he proceedsas a mathematician. But just between you and me, I don’t expect verymuch of it, since I have never before heard of him involving himself witha writing of this kind.”38 Descartes’ expectations were not disappointed.When he finally received the book, shortly after writing this note toMersenne, he must have read it immediately. In his letter of January 28,1641, Descartes transmitted his comments on the book to Mersenne:

I perused M. Morin’s little book. Its main shortcoming is that throughout hetreats infinity as if his mind were above it, and could comprehend its prop-erties. This is a shortcoming common to almost everyone, which I have care-fully tried to avoid, since I have never treated infinity except to submit myselfto it, and not to determine what it is or what it is not. Then, before givingany explanation of controversial matters, in his sixteenth theorem, where hebegins to try to prove that God exists, he bases his reasoning on his purportedrefutation of the motion of the earth, and on the claim that the entireheavens move around it, something that he hasn’t proved at all. And he alsoassumes that one cannot have an infinite number there, etc., something thathe doesn’t know how to prove either. Thus, everything he sets out right upuntil the end is quite far from being evident and quite far from the geomet-rical certainty that he seems to promise at the beginning. I say this justbetween ourselves, if you please, because I don’t want to displease him at all.39


38 AT III 275. 39 AT III 293–94.

Descartes’ reaction here seems quite fair. Morin’s proofs, like those ofSpinoza, who would later offer a very different geometrical proof forthe existence of God, are strongly based on the notion of infinity;Morin’s God is from the first and primarily an infinite being, and it ison this divine attribute that Morin’s arguments are grounded. AndDescartes is certainly correct to note that Morin makes some very oddstatements about infinity. Furthermore, Descartes correctly observesthat Morin’s geometrical proof in the Quod Deus sit depends cruciallyon the nongeometrical premise that the earth is at rest in the center ofthe universe.

This was the last time Descartes mentioned Morin in his correspon-dence, at least in that which survives, and it is the only passage in whichDescartes explicitly addressed Morin’s book. But it seems to me thatmuch of what Descartes has to say about the geometrical mode of expo-sition in the Second Replies is also directed specifically against Morin.Descartes begins by distinguishing between two things, the order ofexposition, the ordo scribendi, and the way of demonstrating, the ratiodemonstrandi. To write in order is simply to write in such a way that “theitems which are put forward first must be known entirely without theaid of what comes later; and the remaining items must be arranged insuch a way that their demonstration depends solely on what has gonebefore.” The ratio demonstrandi, on the other hand, is twofold and rep-resents two different ways of realizing order. The ratio of analysis “showsthe true way by means of which the thing in question was discovered.”This, Descartes tells us, is what he used in the Meditations. “Synthesis,by contrast . . . demonstrates the conclusion clearly and employs a longseries of definitions, postulates, axioms, theorems, and problems, sothat if anyone denies one of the conclusions it can be shown at oncethat it is contained in what has gone before.”40 By synthesis hereDescartes clearly means quite specifically the sort of method that Morinused in Quod Deus sit, a quasi-geometrical demonstration using defini-tions, axioms, and so on.

Descartes makes no bones about it: analysis is vastly to be preferredto synthesis, at least in metaphysics. He writes:

In metaphysics . . . there is nothing which causes so much effort as makingour perception of the primary notions clear and distinct. Admittedly, theyare by their nature as evident as, or even more evident than, the primary


40 AT VII 155–56.

notions which the geometers study; but they conflict with many preconceivedopinions derived from the senses. . . . And so only those who really concen-trate and meditate and withdraw their minds from corporeal things, so far asis possible, will achieve perfect knowledge of them. Indeed, if they were putforward in isolation they could easily be denied by those who like to contra-dict just for the sake of it. This is why I wrote “Meditations” rather than “Dis-putations,” as the philosophers have done, or “Theorems and Problems,” asthe geometers have done.41

This last phrase seems to be a clear reference to Morin’s pamphlet. Anda few lines later there is another, I think: “I am therefore right to requireparticularly careful attention from my readers; and the style of writingthat I selected was one which I thought would be most capable of gen-erating such attention. I am convinced that my readers will derive morebenefit from this than they will themselves realize; for when the syn-thetic method of writing is used, people generally think that they havelearned more than is in fact the case.”42 Descartes thus has very littleregard for the use of the geometrical or synthetic style of writing inmetaphysics. At best, he argues, it is a style appropriate for geometry,where “the primary notions which are presupposed for the demon-strations of geometrical truths are readily accepted by anyone, sincethey accord with the use of our senses.” But, it should be noted,Descartes is not even particularly happy with the use of the geometri-cal style of writing in geometry. He writes: “It was synthesis alone thatthe ancient geometers usually employed in their writings. But in myview this was not because they were utterly ignorant of analysis, butbecause they had such a high regard for it that they kept it to them-selves like a sacred mystery.”43

Synthesis thus seems to be good for very little. (This is yet anotherreason to be suspicious of the often-made claim that Descartes volun-tarily decided that he was going to write his Principles in the syntheticstyle.) But yet, Descartes goes ahead and responds to the request of theauthors of the Second Objections and presents his arguments in the style of the geometers. Given what he said about synthesis, this is not alittle puzzling. Granting the difficulty of his Meditations, Descartes tellshis readers that he is giving them this morsel of the argument not as asubstitute for the analytic Meditations, but in order to give them help incomprehending some specific arguments that are particularly difficultand particularly important.


41 AT VII 157. 42 AT VII 158–59. 43 AT VII 156.

But even this example of synthesis is an implicit critique of Morin’sprocedure in Quod Deus sit. Like Morin’s book, Descartes’ geometricalarguments have definitions, axioms, and theorems. In general it is notilluminating to compare in detail Descartes’ text with Morin’s. UnlikeMorin, Descartes seems to do his best to avoid the notion of infinity.44

The propositions simply formalize arguments found already in the Meditations; there seems little there that can be regarded as a specificreply to Morin’s pamphlet. But Descartes’ geometrical exposition hassomething that Morin does not: postulates. In a standard Euclideangeometry there is little to distinguish postulates from axioms; in bothcases we are dealing with propositions that must be assumed to doproofs. But in Descartes’ geometrical arguments, the postulates aresomething else, not propositions at all:

The first request I make of my readers is that they should realize how feebleare the reasons that have led them to trust their senses up till now. . . . I askthem to reflect long and often on this point. . . . Second I ask them to reflecton their own mind and all its attributes. . . . Fifth I ask my readers to spenda great deal of time and effort on contemplating the nature of the supremelyperfect being. Above all they should reflect on the fact that the ideas of allother natures contain possible existence, whereas the idea of God containsnot only possible but wholly necessary existence. This alone, without a formalargument, will make them realize that God exists.45

These are hardly postulates of the usual sort. They are in fact demands,as the Latin postulare would suggest, things we are asked to do, not merelyto accept. In including such postulates in his geometrical presentation,Descartes is answering the criticisms of the geometrical mode of writinghe made in the Second Replies; it is only because he includes such pos-tulates, Descartes thinks, that the geometrical mode of presentation iscapable of leading us to knowledge of things metaphysical. In this way,the differences between Descartes’ and Morin’s geometrical argumentsfor the existence of God simply underscore Descartes’ rejection ofMorin’s chosen form of presentation. Thus the geometrical presentationthat follows the Second Replies can be read not only as a clarification ofthe arguments, terminology, and assumptions used in the Meditations,not only as a civil answer to a civil question from the authors of the Second Objections, but also as a philosophical exercise directed against the Quod Deus sit of Jean-Baptiste Morin.


44 However, note Descartes’ Axiom 6, in AT VII 165–66. 45 AT VII 162–63.

Morin’s Response

The Second Replies is the last text in which Descartes has anything to say about Morin; as far as Descartes was concerned, the less said, thebetter. But though Descartes may not have had anything more to sayabout Morin, Morin had quite a lot to say about his more famous colleague.

The response is found in Morin’s posthumously published Astrologiagallica. While there is no direct evidence that Mersenne actually showedMorin the direct criticisms Descartes made of his work, the paragraphin the letter quoted above, the alterations Morin made in the newedition of the Quod Deus sit – included in the Astrologia – suggest thatMersenne may well have transmitted the essence of those criticisms.Though in the end he does not give up his strong dependence on infin-ity, nor does he actually alter many of the details of his proofs, therearrangements and the additional axioms and definitions show somesensitivity to Descartes’ concerns.46 Also, later in the Astrologia there isconsiderably more discussion of Descartes, particularly his physics. Altogether, this amounts to an additional set of objections against theMeditations, and against the Principles, too, objections especially worthstudy given Morin’s rather interesting position in the intellectual worldof mid-seventeenth-century France. But rather than trying to survey thewhole of Morin’s attack against Descartes, let me just touch on a fewissues with respect to the questions of analysis versus synthesis andDescartes’ geometrical arguments.

Morin begins his discussion of Descartes’ geometrical exposition bynoting that it was he, Morin, and his Quod Deus sit that elicited the discussion:

Although my little book against the atheists pleased everyone, after the publication of his Meditations, those who were not satisfied with his demon-strations for the existence of God through our idea of him requestedDescartes to prove the same a posteriori through his creatures, as I had done.To that same end, that same little book was requested of me, which the


46 For example, in the new version of theorem 15, now theorem 22 (“this world is finite”),Morin eliminates the argument he had used earlier, and which had offended Descartesso much, which depends on his refutation of Copernicus. See Astrologia gallica, p. 11. Similarly, in the new version of theorem 27 (now theorem 35) Morin eliminated theassumption about infinity that Descartes found so problematic, that one cannot add any-thing to an infinite number. See Astrologia gallica, p. 13. There are other, smaller changesas well that are suggestive.

Reverend Father Mersenne, known to all of the learned, sent him in Holland, so that he might see my method for proceeding in the geometricalfashion.47

Morin goes on to examine the three proofs that Descartes gives in hisgeometrical appendix, finding them, one by one, unsatisfactory. Mostinteresting, though, are his comments on Descartes’ remarks on theanalytic and synthetic modes of reasoning. He criticizes Descartes’ useof both ways of proceeding.

Morin notes that Descartes does try to give a geometrical account,like Morin’s own, using definitions, axioms, and theorems. But he alsotakes note of the fact that Descartes makes use of postulates: “Thenthere are also seven postulates, by which the mind binds itself. However,I have demanded [postulaverim] nothing. Rather, I have left the mindwith its freedom of judgment.”48 It is interesting here that Morin doesnot seem to understand exactly why Descartes adds the postulates inthe way he does, nor does Morin understand the rather radical differ-ence between Descartes’ postulates and those more commonly foundin the tradition. All he says is that they seem to bind the intellect in away that he does not want to. He does continue, however, with a ratheruncharacteristically penetrating critique of Descartes’ Postulate 5: “I askmy readers to spend a great deal of time and effort on contemplatingthe nature of the supremely perfect being. Above all they should reflecton the fact that the ideas of all other natures contain possible existence,whereas the idea of God contains not only possible but wholly neces-sary existence. This alone, without a formal argument, will make themrealize that God exists.”49 Morin comments: “Once we have concededthis postulate, then no definitions, no axioms, nor any demonstrationsare needed, either through analysis or through synthesis.”50 Morin’spoint is a good one: Take this particular postulate seriously, and thereis no need for argument at all.

Morin does not discuss Descartes’ general remarks on the prefer-ability of analysis over synthesis for metaphysics; the general theoreti-cal position seems to escape him. But he does say why he thinks thatanalysis is not an appropriate way of proving the existence of God.Morin writes:


47 Morin, Astrologia gallica, p. 5. 48 Ibid. 49 AT VII 163.50 Morin, Astrologia gallica, p. 6. Morin goes on to say that if we don’t concede the postulate,

then we cannot pass from the idea of God in the mind to his existence in reality, but thatis a longer story.

And indeed it seems remarkable to me that M. Descartes chose the analyticmethod for proving the existence of God, which is utterly inappropriate forthis purpose. Analysis is defined by Viète as the assumption of that which is soughtas if it were conceded, then through consequences [passing] to that which is generallyconceded as true. If it is generally conceded as true that he [Descartes] existsfrom the fact that he thinks and, indeed, that he has an idea of an infinitelyperfect being, which he calls God, then that which we seek will be whetherGod exists. Now, from the definition of analysis we should assume that Godexists, as if it were conceded, and from that concession, we should seek [toshow] as a consequence that M. Descartes, or he who has the idea of an infi-nitely perfect being, that is, God, thinks and therefore exists. But in his analy-sis, he demonstrates nothing of the sort; indeed, nothing of the sort can bedemonstrated. For God exists from eternity, but M. Descartes has not thoughtfrom eternity, and therefore did not exist, nor did he have the idea of God[from eternity]. Therefore it is obvious that analysis can do nothing towardproving the existence of God from the idea of God which he says that he has,considering that idea as the concept of a being of greatest perfection or ofinfinite nature, as he often does.51

The criticism is just, if we assume that Descartes had in mind Viète’sconception of analysis here. While it would take us too far afield todemonstrate this, I think that Morin’s criticism shows that he simplymisunderstood what Descartes was up to in calling the Meditations ana-lytic, just as he missed the deeper points behind Descartes’ critique ofthe geometrical mode of writing in metaphysics. It is quite clear thatDescartes’ Meditations are not intended to be analytic in the sense inwhich Viète’s mathematics is.52

But Morin kept insisting, stubbornly, on the fact that he was right onthis issue, as on others, and Descartes was wrong. His final proof was,in his eyes, definitive: the doctors of the Sorbonne gave his Quod Deussit the approbation that they denied Descartes’ Meditations. Here is anargument from authority if ever there was one.53


51 Ibid., p. 7.52 For one interpretation of what Descartes means when he calls the Meditations analytic, see

the discussion of the Meditations in Daniel Garber, “Semel in vita: the Scientific Backgroundto Descartes’ Meditations,” in Amélie Rorty (ed.), Essays in Descartes’ Meditations (Universityof California Press, 1986), pp. 81–116, essay 11 in this volume.

53 Jean-Robert Armogathe has recently argued that contrary to what Morin thought,Descartes actually did receive the approbation of the Sorbonne. See J.-R. Armogathe,“L’approbabion des Meditationes par la Faculté de Théologie de Paris (1641),” BulletinCartésien 21 (1994) [in Archives de Philosophie 57 (1994)], pp. 1–3.

In this essay I have argued that Jean-Baptiste Morin and his Quod Deussit stand behind at least parts of the Second Objections, and that it wasspecifically to Morin and his little book that Descartes was respondingat the end of the Second Replies and in the geometrical appendix. Buthow does this change our understanding of those passages? Perhaps notat all; interesting as that bit of historical information may be to thoseof us with an antiquarian bent, it may not have any real philosophicalbearing. But then maybe it does.

I would like to end with a kind of conjecture, a stab at an argumentthat one might make on the basis of my historical argument. I thinkthat what I have presented here strengthens the case for saying thathowever important it might be for earlier thinkers, however much itmay be emphasized by later commentators, the doctrine of analysis andsynthesis may not be a central tenet in Descartes’ own thought, not abasic category in terms of which Descartes liked to think of his workand that of others. Rather, I suspect that it is a very specific responseto a very specific proposal for how to do metaphysics, a proposal embod-ied in the example of Morin’s Quod Deus sit. And, I think, it is a clearrejection of that way of doing metaphysics. Even though Descartes doesdevelop his ideas in synthetic form in the geometrical appendix to theSecond Replies, it must be emphasized that this is largely (only?) to show the inadequacy of that form and the problems inherent in anenterprise of the sort that Morin was attempting to undertake. Thisdoes not establish for certain that Descartes did not then generalize thenotion of synthesis, or take it seriously in his own later works. But, Ithink, the argument should somewhat undermine whatever temptationwe might have to see synthesis as a more general category and to try toinclude the Principles as synthetic, as many readers from Martial Gueroult to Edwin Curley and J. M. Beyssade have done.54 In late 1640and early 1641, when Descartes confronted the geometrical argumentof Morin and penned his response, both his private response toMersenne and his more public response in the Second Replies, and when he began drafting what was to become the Principles of Philosophy,he saw nothing to recommend a geometrical metaphysics of the sortthat Morin was attempting to establish.


54 For a more systematic attack on the idea that the Principles should be understood as syn-thetic, see Daniel Garber and Lesley Cohen, “A Point of Order: Analysis, Synthesis, andDescartes’s Principles,” in Archiv für Geschichte der Philosophie 64 (1982), pp. 136–47, essay3 in this volume.

5

DESCARTES AND EXPERIMENT IN THEDISCOURSE AND ESSAYS

It is generally recognized that knowledge for Descartes is the clear anddistinct perception of propositions by the intellect; knowledge in thestrictest sense is certain, indeed indubitable, and grounded in thepurely rational apprehension of truth. But it is also generally recog-nized that Descartes was a serious experimenter, at least in his biologyand his optics, and that in these areas, at least, he seemed to hold thatknowledge requires an appeal to experience and experiment. Writing,for example, in Part VI of the Discourse on Method, Descartes laments thefact that he has neither the time nor the resources to perform all theexperiments (expériences) necessary to complete his system, and callsupon his readers to “communicate to me those that they have alreadymade, and to help me in performing those which remain to be done”(AT VI 65). (One can see in the Discourse a clear anticipation of animportant later literary form, the grant application.)

To the twentieth-century philosopher this looks a bit puzzling: Howcan Descartes be both a rationalist, who sees knowledge as deriving fromthe intellect, and an experimentalist, who sees experiment and obser-vation as essential to the enterprise of knowledge? This is the puzzle Iwould like to address in this essay. I shall argue that not only is thereno contradiction here, but that the appeal to experience is an essentialpart of the method for constructing a deductive science. We shall begin

85

Other than the abbreviations used throughout this book (AT, CSM, CSMK), when quotingthe Meteors or the Dioptrics I use the following abbreviation:

Ols Descartes, Discourse on Method, Optics, Geometry, and Meteorology, trans. Paul J. Olscamp.Indianapolis: Bobbs-Merrill, 1965.

with a brief account of Descartes’ procedure for constructing hisscience, his method. While Descartes’ method is discussed at greatlength in any number of books and papers, there is hardly a clearaccount in any of the literature of what it is in practice. Then, once wehave a clear picture of what Descartes’ method is, and the precisedeductive structure of the body of knowledge that he is building, wecan turn directly to the question of experiment, and see how it fits intothe program.

Method

I hold the view that Descartes, in an important sense, gave up hisfamous method sometime in the late 1630s or early 1640s, and so I donot want to identify the question of Descartes’ scientific procedure withthat of his method.1 But to understand Descartes’ procedure in scienceit will be helpful to begin with a brief account of the method as it is initself and as it is in application, and work from there. In discussing themethod, I shall concentrate on the account Descartes gives in the earlyRules for the Direction of the Mind, which Descartes worked on inter-mittently from 1618 or so until 1628 or thereabouts; though never finished and never published, it is by far the most thorough account of method in the Cartesian corpus, far more intelligible than the briefand enigmatic account of the method Descartes gives in Part IV of the Discourse.

In order to understand the method, we must understand the goal ofinquiry in the Rules, for the method of the Rules is precisely a methodof attaining that goal. The goal of inquiry is the subject of the first two rules:

The goal [finis] of studies ought to be the direction of one’s native abilities[ingenium] toward having solid and true judgments about everything whichcomes before it. . . . We should concern ourselves only with those objects ofwhich our native abilities seem capable of certain and indubitable cognition.(AT X 359; AT X 362)

By “certain and evident cognition” here, Descartes seems to meanknowledge grounded in what he calls intuition and deduction. In RuleIII Descartes defines intuition:


1 For a full defense of this view, see D. Garber, “Descartes and Method in 1637,” essay 2 inthis volume, and Garber, Descartes’ Metaphysical Physics (Chicago: University of Chicago Press,1992), chapter 2. The account of method in this essay is drawn from these sources.

By intuition I understand not the fluctuating faith in the senses, nor thedeceitful judgment of a poorly composed imagination, but a conception ofa pure and attentive mind, so easy and distinct that concerning that whichwe understand no further doubt remains; or, what is the same, the undoubtedconception of a pure and attentive mind, which arises from the light of reasonalone. (AT X 368)

Deduction is defined in terms of intuition; it is a chain of intuitions,the intuitive grasping of a connection between one proposition andanother (AT X 369–370, 407). This, Descartes argues in the Rules, isthe only way to knowledge (AT X 366).

Method is what, in the Rules, is supposed to lead us to such knowl-edge. But what is this method? Descartes writes in Rule IV:

By method I understand certain and easy rules which are such that whoeverfollows them exactly will never take that which is false to be true, and withoutconsuming any mental effort uselessly . . . will arrive at the true knowledge[vera cognitio] of everything of which he is capable. (AT X 371–372)

Descartes summarizes these “certain and easy rules” in Rule V:

The whole of method consists in the order and disposition of those thingstoward which the mental insight [mentis acies] is to be directed so that we dis-cover some truth. And this rule is observed exactly if we reduce involved andobscure propositions step by step to simpler ones, and thus from an intuitionof the simplest we try to ascend by those same steps to a knowlege of all therest. (AT X 379)

Descartes’ rule of method has two steps, a reductive step, in which“involved and obscure propositions” are reduced to simpler ones, anda constructive step, in which we proceed from simpler propositions backto the more complex.2 But the rule makes little sense, nor does itconnect very clearly with the account of knowledge and certainty interms of intuition and deduction, unless we know what Descartes meanshere by the reduction to simples, and the construction of the complexfrom the simples.

The precise method Descartes has in mind is nicely illustrated by anexample he gives of methodical investigation in Rule VIII (see Table1). The problem Descartes poses for himself is that of finding the ana-clastic line, that is, the shape of a surface “in which parallel rays are

descartes and experiment 87

2 I should point out that I am breaking with most commentators, who refer to these as theanalytic and synthetic steps. See my remarks on this in “Descartes and Method in 1637,”note 4, essay 3 in this volume.

refracted in such a way that they all intersect in a single point afterrefraction” (AT X 394). Now, Descartes notices – and this seems to bethe first step in the reduction – that “the determination of this [ana-clastic] line depends on the relation between the angle of incidenceand the angle of refraction” (AT X 394). But, Descartes notes, this ques-tion is still “composite and relative,” that is, not sufficiently simple, andwe must proceed further in the reduction. Rejecting an empirical inves-tigation of the relation in question, Descartes suggests that we must nextask how the relation between the angles of incidence and refraction iscaused by the difference between the two media, for example, air andglass, which in turn raises the question as to “how the ray penetratesthe whole transparent thing, and the knowledge of this penetration presupposes that the nature of the illumination is also known” (AT X394–395). But, Descartes claims, in order to understand what illumi-nation is we must know what a natural power (potentia naturalis) is. Thisis where the reductive step ends. At this point, Descartes seems to thinkthat we can “clearly see through an intuition of the mind” what a naturalpower is (AT X 395). Other passages suggest that this intuition is inti-mately connected with motion.3 Once we have such an intuition, we can


Table 1. Anaclastic Line Example (Rules for the Direction of the Mind, Rule VIII)

Q1. What is the shape of a line (lens) that focuses parallel rays of light to thesame point?

Q2. What is the relation between angle of incidence and angle of refraction (i.e.,the law of refraction)?

Q3. How is refraction caused by light passing from one medium into another?

Q4. How does a ray of light penetrate a transparent body?

Q5. What is light?

Q6. What is a natural power?

Intuition: A natural power is. . . .

Construction: The construction consists in traversing the series of questions fromQ5 to Q1, deducing the answer to each question from that of the precedingquestion.

3 Rule IX tells us that in order to understand the notion of a natural power, “I will reflecton the local motions of bodies” (AT X 402). What this suggests is that the understandingof illumination is, somehow, an intuitive judgment about the simple nature, motion, thoughit is not clear how exactly he thought this would work.

begin the constructive step, and follow, in order, through the questionsraised until we have answered the original question, that of the shapeof the anaclastic line. This would involve understanding the nature ofillumination from the nature of a natural power,4 understanding theways rays penetrate transparent bodies from the nature of illumination,and the relation between angle of incidence and angle of refractionfrom all that precedes. And finally, once we know how angle of inci-dence and angle of refraction are related, we can solve the problem ofthe anaclastic line.5

This example develops the programmatic statement of the methodas given in Rule V in a fairly concrete way. If we take the anaclastic lineexample as our guide, then methodical investigation begins with a ques-tion, a question which, in turn, is reduced to questions whose answersare presupposed for the resolution of the original question posed (i.e., Q1 is reduced to Q2 if and only if we must answer Q2 before wecan answer Q1). The reductive step of the method thus involves, asDescartes suggests in Rule VI, ordering things “insofar as some can beknown from others, so that whenever some difficulty arises, we willimmediately be able to perceive whether it will be helpful to examinesome other [question], and what, and in what order” (AT X 381). Andso, in a sense, the reduction leads us to more basic and fundamentalquestions, from the anaclastic line, to the law of refraction, and backeventually to the nature of a natural power and to the motion of bodies.Ultimately, Descartes thinks, when we follow out this series of questions,from the one that first interests us, to the “simpler” and more basicquestions on which it depends, we will eventually reach an intuition.When the reductive stage is taken to this point, then we can begin theconstructive stage. Having intuited the answer to the last question inthe reductive series, we can turn the procedure on its head, and begin


4 Descartes writes, “If, at the second step, he is unable to discern at once what the nature of light’s action is . . . he will make an enumeration of all the other natural powers, in the hope that a knowledge of some other natural power will help him understand this one,if only by analogy” (AT X 395). In personal correspondence John Nicholas has emphasizedto me the importance (and complexity) of this step in the construction. He suggests, plausibly, I think, that “human limitations are such that in practice we commonly cannotcarry out the downward deduction, and have to fall back on the surrogate step of analogizing and comparing with other natural agencies that the targeted one.” Insofar asthis analogizing may depend on our experience with the phenomenon in question, as wellas with other phenomena, this suggests to him that there may be another use of experi-ence in Descartes than the one that I emphasize in the following sections. He might wellbe right.

5 See Pierre Costabel, Démarches originales de Descartes savant (Paris: Vrin, 1982), 53–58, foran account of the historical background to this example.

answering the questions that we have successively raised, in an orderthe reverse of the order in which we have raised them. What this shouldinvolve is starting with the intuition that we have attained through thereductive step, and deducing down from there, until we have answeredthe question originally raised. Should everything work out as Descarteshopes it will, when we are finished it is evident that we will have certainknowledge as Descartes understands it in the earliest portions of theRules; an answer arrived at in this way will constitute a conclusiondeduced ultimately from an initial intuition.

Descartes’ strategy here is extremely ingenious. The stated goal ofthe method is certain knowledge, a science deduced from intuitivelyknown premises. What the method gives us is a workable procedure fordiscovering an appropriate intuition, one from which the answer to thequestion posed can be deduced, and it shows us the path that deduc-tion must follow. This workable procedure is the reduction of a ques-tion to more and more basic questions, questions that we can identifyas questions whose answers are presupposed for answering the questionoriginally posed; this reduction both leads us to an intuition, Descartesthinks, and shows how we can go from that intuition back to the ques-tion originally posed.

This is the story as of 1628 or so, when Descartes abandoned thecomposition of the Rules. As noted earlier in this section, I think thatDescartes’ thinking about method changes in his later years. Put briefly,while Descartes always maintains the view that knowledge is to begrounded in intuition, in the immediate apprehension of truths, hechanges his mind about which truths lie at the bottom, and about howit is that we are to find them. In the Rules he seems to take the view thatour knowledge of the physical world is grounded in certain truths,immediately grasped, about the nature of bodies, natural powers, andso forth.6 But in the later writings, the grounding is ultimately in meta-physics, our knowledge of ourselves and God, and in God’s role as theguarantor of our clear and distinct perceptions; in the later writings,the intuitions he takes for granted in the Rules must be grounded inGod our creator and in us, God’s creation. And furthermore, in the


6 See especially the development in Rule XII (AT X 419) where Descartes discusses the so-called simple natures on which all our knowledge is supposed to be grounded. The simplenatures divide into three classes: intellectual, material, and common. The intellectualsimple natures include knowledge, doubt, ignorance, volition. The material simple naturesinclude shape, extension, and motion. The common simple natures include existence,unity, and duration.

latter writings, the reductive step of the method, a step that can lead usonly as far as the unjustified intuitions, is abandoned in favor of a directattack on the foundations.7 Despite these changes, though, it will behelpful to begin attacking the question of experiment in Descartes byexamining the role it plays in his method.

Method and Experiment

In the previous section of this essay I emphasized what might be calledthe deductive structure of Descartes’ project, the view of a completedscience as a deduction from initial intuitions. In calling the structuredeductive I do not mean to say that it is deductive in precisely themodern sense, or that it is deductive in any precise sense at all. It mustbe remembered that when Descartes introduces the notion of deduc-tion in the Rules it is in explicit contrast to the formal logic of theSchools, indeed, in explicit contrast to any formal procedures at all. ForDescartes, intuition and deduction are the immediate grasping of thetruth of propositions and the inferential connections between propo-sitions, and so there is no in principle reason why a deduction cannotbe an ampliative inference in the modern sense of the term, as, forexample, the cogito seems to be.8 But despite Descartes’ refusal to pindown the notion of a deduction in any formal way, a completed scienceis supposed to be deductive for him in a rather strict sense; derivativeand more complex propositions are supposed to be deduced in hissense from propositions simpler and more basic, and grounded ulti-mately in intuition.

However, Descartes is clear, his natural philosophy is definitely notsupposed to be a priori in the modern sense of the term, knowledgeobtained without the help of experience. Although Descartes seems towant to proceed deductively, experience and experiment have a signif-icant role to play in this business. It is, of course, well known by nowthat Descartes was a dedicated experimenter, observer, and dissector,and that the empirical investigation of nature is given significant atten-


7 For a fuller account of the changes, see the references given in note 1 in this essay.8 See Garber, “Science and Certainty in Descartes,” in Descartes: Critical and Interpretive Essays,

ed. Michael Hooker (Baltimore: The Johns Hopkins University Press, 1978), 114–151, esp.116–123. Desmond Clarke argues that the term “deduction” is so broad for Descartes that even hypothetical arguments count as deductions for him. See D. Clarke, Descartes’ Philosophy of Science (University Park, PA: The Pennsylvania State University Press, 1982),63–70, 201–202, 207–210.

tion in the Rules, the Discourse, and other writings where he discusseshis natural philosophy. Of course, this raises an important problem:How is the appeal to experience consistent with the apparently deduc-tive structure of Descartes’ project? There is a considerable literatureon this basic question, and answers range from denying (or better,ignoring) the interest in experiment, to denying that Descartes’ sciencewas ever intended to be deductive, to claiming that Descartes was simplyinconsistent – deductive in theory, and empirical in practice.9 This isthe problem I would like to address in this section. I shall try to showsomething that may sound a bit paradoxical, that for Descartes experi-ment functions as an important and, in fact, indispensable tool for discovery in his deductive science, and it is to experience that we must turn to help us sort our the details of the deductive hierarchy ofknowledge.

A reasonable place to begin is with a passage from Part VI of the Discourse, where Descartes attempts to explain to the reader the use ofexperiment in his thought. The passage begins with a lengthy accountof where experiment is not really necessary. Descartes reports that hebegan his investigations with “the first principles or first causes” ofeverything, which can be discovered from “certain seeds of truth whichare naturally in our souls.” From this Descartes derived “the first andmost ordinary effects that one can deduce from these causes,” theheavens, stars, the earth, water, air, fire, and so on. The passage thencontinues as follows:

Then when I wanted to descend to those which were more particular, I waspresented with so many different kinds of things that I did not think that itwas possible for the human mind to distinguish the forms or kinds of bodieswhich are on the earth from an infinity of others that could have been there,if God had wanted to put them there, nor, consequently, to make them usefulto us, unless one proceeded to the causes through their effects, and attendedto many particular experiments. Afterward, reviewing in my mind all of theobjects which have ever been presented to my senses, I venture to say that Ihave never noticed any thing that I could not easily enough explain by theprinciples that I have found. But I must also admit that the power of natureis so ample and so vast, and that these principles are so simple and so general,that I have found hardly any particular effect which from the first I did not know could be deduced in many ways, and [I admit] that my greatest difficulty is ordinarily to find in which of these ways it depends on these principles. (AT VI 64–65)


9 For a survey of the various views taken in the literature, see Clarke, Descartes’ Philosophy ofScience, 9–10.

Experiment seems not to be at issue in the early stages of investiga-tion. Where experiment becomes important, Descartes indicates, iswhen we move from the very most general features of the world, and,as he puts it, descend to particulars. There, he says, the direct deduc-tion from first principles must stop, and we must “proceed to the causesthrough their effects, and attend to many particular experiments.” Thishas suggested to many, and not implausibly, that at this stage sciencemust become a posteriori, arguing from effect to cause by a kind ofhypothetico-deductive method of the kind practiced in the Essays anddefended in the correspondence of 1637 and 1638.10 While this maydescribe Descartes’ views later, in certain pessimistic sections of thePrinciples, this is not, I think, what Descartes had in mind in the Dis-course.11 In the passage in question, Descartes seems clear that he is still interested in deduction, even after he has descended to particulars.The problem is that in any given case, there are many possible ways inwhich one can deduce from the general principles, “so simple and sogeneral,” to the particular effects we observe. Experiment is somehowsupposed to help us find the right deductions, the ones that pertain toour world and to the phenomena that concern us. In this way, experi-ments seem not to replace deductions, but to aid us in making the properdeductions.12

The view is initially quite paradoxical. How can some deductions beright and others wrong? How can it be that experiment is essential for


10 Charles Larmore suggests such a view in “Descartes’ Empirical Epistemology,” in Descartes:Philosophy, Mathematics and Physics, ed. Stephen Gaukroger (Sussex: The Harvester Press,1980), 6–22, esp. 9, 12. I presented a similar view in “Science and Certainty in Descartes,”though I no longer think that it is correct.

11 One might point here to the obvious use of hypotheses in the Dioptrics and Meteors, wellbefore the Principles of Philosophy; see AT VI 83ff., 233ff.; Ols 66ff., 264ff. But the Essaysconstitute an attempt to give the results of inquiry without revealing the full system, andthey are not intended to replace proper argument in natural philosophy, which proceedsfrom cause to effect. By arguing from hypotheses he thought that he could show some ofhis results without having to divulge the first principles of his physics, for which, hebelieved, the public was not ready. But, while pleased with his Essays, he was clear thatthey represent not the definitive treatment of his thought, in accordance with his methodof inquiry, but, rather, interesting experiments in exposition. There is an extended dis-cussion of this in Part VI of the Discourse : AT VI 76–77. This theme also runs throughDescartes’ correspondence in the period; see AT I 562–564; AT II 141–144, 199–200;CSMK 87–88, 103–104, 107. See also Garber, “Science and Certainty in Descartes” andDescartes’ Metaphysical Physics, chapter 2.

12 See also Descartes’ remarks in Principles III 4. There he talks about having to turn to thephenomena at that stage in his exposition, “not to deduce an account of causes from theireffects,” but “to direct our mind to a consideration of some effects rather than others fromamong the countless effects which we take to be producible from the selfsame causes.”

a deductive explanation of a phenomenon? And how could Descartespossibly have maintained a deductive structure in his science, if headmits that there are circumstances in which we must “proceed tocauses through their effects”? To see how this might work, let us turnto some examples.

As discussed previously, the anaclastic line problem from Rule VIIIinvolves finding the shape of a surface that refracts all parallel rays intoa single point. Descartes’ solution to the problem requires us to followa certain series of steps, first a reduction of the problem to a series ofsimpler ones, then a constructive step, where the reductive series is tra-versed backwards, resulting in a deductive solution to the problem, if allworks well. Descartes never tells us here where we can or must appeal toexperience; experience comes up only in a negative way, where Descartesasserts that we should not try to discover the relation between the angleof incidence and the angle of refraction through experiment, for thatwould violate Rule III, which tells us that only intuition and deductionare sources of real knowledge (AT X 368). But there is at least one placein the reduction where an appeal to experience would seem to behelpful, if not altogether obligatory. In the very next step of the reduc-tion, Descartes says that the investigator must notice that the relationbetween the angles of incidence and refraction itself depends on thechanges in these angles due to the differences in the media throughwhich the ray is passing (e.g., from air into glass, or water into air), andthat these changes, in turn, depend on the way in which the ray pene-trates the transparent body (AT X 394). Descartes does not mentionexperiment or experience in this context. But it is difficult to imaginethat this is a step that we can make on the basis of the “seeds of truth”alone. While it may not require sophisticated optical experiments, itseems that we at least require some minimal experience with light raysand lenses, or other actual instances of refraction, in order to see thatlight is typically bent by passing from one medium into another, and tocome to the realization that in order to discover the law refraction obeyswe must first understand how light passes through media of differentsorts. In this way experience would seem to help us to see how we mightproceed in our investigation by suggesting what further questions itmight be useful for us to look into.

Experiment comes up at best only implicitly in the anaclastic lineexample. But it is quite a visible feature of another example Descartesgives of his method. The example I have in mind is the account Descartesgives of the rainbow in the Eighth Discourse of his Meteors. This


passage contains the only explicit mention of the method in all of theEssays, and it is singled out in a letter from 1638 as an exemplary use ofthe method in practice (see AT VI, 325; Ols. 332 and Descartes to Vatier,22 February 1638; AT I 559; CSMK, 85). The example is a very compli-cated one, one of Descartes’ best but most complex scientific arguments.I shall begin by summarizing the argument, and then try to show howthe mass of experimental detail and complex argument sorts itself outinto a methodical framework (see Fig. 1).13

The problem is to explain how it is that rainbows come about. Theaccount begins with the observation that rainbows appear when and onlywhen there are water droplets in the air. Descartes then turns to thestudy of large spherical flasks of water which, he claims, duplicate the effects seen in individual droplets of water that appear to cause therainbow. Observations on the flask allow Descartes to measure theangles at which colors are observed, and allow him to determine thatthere are two regions of color whose red portions are about 42 and 52degrees from the angle at which they are hit by the rays of the sun (seeFig. 1). These experiments also allow Descartes to determine that thesetwo regions of color derive from two different combinations of reflec-tion and refraction within the water flask; the brighter color region(which corresponds to what is now called the primary bow) at 42degrees results from two refractions and one internal reflection, whilethe dimmer color region (the secondary bow) at 52 degrees resultsfrom two refractions and two internal reflections. (The two paths canbe discerned within the flask represented in Fig. 1.) These investiga-tions led Descartes to two further questions, why there is color at all inthese cases, and why it is that the colors appear at two specific angles.The first question, that of color, is explored experimentally, through aprism, in which, like the flask, colors are produced through the reflec-tion and refraction of light (see Fig. 2). Observations made with theprism show that a curved surface, like that of the raindrop or the flask,


13 My own interest in the rainbow case here is largely as an illustration of the method of theRules. For discussions of Descartes’ account of the rainbow that emphasize its place in thehistory of such discussions and in the history of optics more generally, see Carl B. Boyer,The Rainbow (Princeton, NJ: Princeton University Press, 1987), Chapter 8, and Jean-RobertArmogathe, “L’arc-en-ciel dans les Météores, in J.-L. Marion and N. Grimaldi, eds. Le Discourset sa méthode, 145–162. Considering Descartes’ account in its historical perspective makesit quite clear that despite the impression he gives in the Meteorology of having discoveredeverything himself, he owes a great deal to previous investigators. Interesting and impor-tant as these historical considerations are, I will focus instead on Descartes’ presentation ofhis theory in an attempt to untangle the methodological underpinnings of his argument.

is not needed to produce color; nor is a reflection necessary, Descartesdiscovers through experiment. What seems to be necessary, Descartesfinds, is at least one refraction, and a restricted stream of light. But inorder to understand how the refraction of a restricted beam of lightcan produce color, we must press deeper into the nature of light andthe way it passes through a transparent body, the very questions that wewere pressed back into in the anaclastic line case. The nature of lightwe know from the Dioptrics: “[The nature of light is] the action of move-ment of a certain very fine material whose particles must be picturedas small balls rolling in the pores of earthly bodies” (AT VI, 331; Ols.336).14 And, Descartes argues, what happens when a restricted beam of


Figure 1

14 This is the paraphrase Descartes gives in the Meteors; the passage he is referring to in theDioptrics can be found at AT VI 89–93.

light passes from one medium into another in refraction is that the ballsare given differential tendencies to rotate, depending on where theyare in the stream (see Figs. 2 and 3). Since, refraction aside, that is theonly mechanical effect that passing from one medium into another hason the light, Descartes argues that color just must be caused by the dif-ferential tendency to rotation. Those balls with a greater tendency torotate produce the color red in us, Descartes claims, while those thathave a lesser tendency to rotate produce the color blue/violet in us.(Remember, of course, Descartes held that in the strictest sense, coloris only in the mind, and not in bodies.) And so, from the nature of lightand the way it passes through media, we have shown how colors are pro-duced, Descartes thinks. But it still remains to show why the colors areproduced in two discrete regions, at characteristic angles from that ofsunlight. To solve that problem, Descartes turns back to the flask.Appealing to the law of refraction, which Descartes alludes to in theanaclastic line example, and derives (after a fashion) in the Second Dis-course of the Dioptrics, he demonstrates that after two refractions andone reflection, the vast majority of a bundle of parallel rays hitting theflask, wherever they may hit, will emerge from the flask between 41 and42 degrees with respect to the angle of the incident light, and after two


Figure 2

refractions and two reflections, the majority will emerge at between 51and 52 degrees (AT VI 336ff.; Ols. 339ff ).15 From this it follows that at those two regions on the surface of the sphere, there will be two discrete streams of light that emerge from the flask, moving from onemedium into another. And from the previous argument, this will resultin two regions of color at the two angles earlier observed. And so, from


Figure 3

15 Descartes does the calculation by considering a spherical droplet of water hit by parallelrays over one hemisphere, and calculating where various of the rays would emerge afteran appropriate number of reflections and refractions. His conclusion, carefully stated,reads:

I found that after one reflection and two refractions, very many more of [the rays] canbe seen under the angle of 41 to 42 degrees than under any lesser one; and that noneof them can be seen under a larger angle. Then I also found that after two reflectionsand two refractions, very many more of them come toward the eye under a 51 to 52 degree angle, than under any larger one; and no such rays come under a lesser.(AT VI 336; Ols. 339)

While the conclusion is arrived at by calculation, that calculation must make explicitappeal to the index of refraction for water. When the question comes up in the SecondDiscourse of the Dioptrics, he notes that we must appeal to experience in order to deter-mine the value of this constant for various sorts of materials (AV VI 101–102). This wouldseem to be another place in which experiment would enter into the method. However,one presumes that Descartes believed that the index of refraction could itself be arrivedat by calculation, were we to know the size, shape, and motion of the corpuscles that makeup water.

the nature of light, the way it passes through media, and the law ofrefraction, it follows that the rays of sunlight hitting the flask will resultin two regions of color at two characteristic angles. When we have amultitude of such drops, we have a rainbow.

It is by no means obvious how this somewhat confused mass of experiment and reasoning can be fitted into the rather rigid mold ofDescartes’ method. The schematic representation of the argumentgiven in Table 2 indicates one plausible way in which the argumentmight fit. In the schematic representation of the argument, Q1 throughQ5 represent the reduction, which leads us from the question originallyposed, “what is the cause of the rainbow,” back to the intuitions whichare the starting point of the Cartesian deduction, intuitions about thenature of light and how it passes through media. But the importantthing is, of course, the specific path that Descartes follows to go fromthe initial question to the intuition, for it is that path that will deter-mine the path followed in the deduction. In this case Descartes pro-ceeds by splitting the question into two questions, one about color andone about the two regions. Included in square brackets are the empir-ical results derived from experiment at the point in the argument inwhich Descartes appeals to them. The path followed after the intuitionis relatively straightforward. Here we are dealing with the same stepsfollowed in the reduction, only in the reverse order, as we pass fromintuition to the final answer to the question originally posed. But unlikethe reduction, experiment and its results seem to play no role in thispart of the argument. The example is certainly much more complexthan the anaclastic line example, but it seems to have much in commonwith it in structure.

Before turning back to my main theme, the use of experiment inthese arguments, I would like to comment on the kind of deductionthat is involved in this case. In the anaclastic line case, we had a defi-nite question, the shape of a lens with such-and-such properties, and atthe conclusion of the procedure we can expect a deductive answer tothe question, a deduction from basic principles (ultimately, the natureof a natural power) that a lens with this-or-that shape will have such-and-such characteristics. But the situation here is a bit different. Whatwe are seeking is the cause of the rainbow. The answer to this questionis, in a sense, not deduced; rather, it is revealed in the deduction itself.The deduction shows us how we can go from the nature of light to thephenomenon of the rainbow; what is deduced, strictly speaking, is justthe phenomenon itself, the patches of color in the sky. But the path


Table 2. Descartes’s Account of the Rainbow (Meteors, Eighth Discourse)

Q1. What causes the rainbow (two regions of color)?

[Rainbows appear only in the presence of water droplets; size isirrelevant to the phenomenon.]

Q2. What causes the two regions of color in any spherical ball?Q2a. What causes the two regions? Q2b. What causes the color?

[The two regions result from two [Color is produced without a curvedcombinations of reflection and surface and without reflection; it refraction.] requires a restricted stream of light,

and a refraction.]

Q3a. Why do the two combinations Q3b. How does refraction cause colorof reflection and refraction result under appropriate circumstances?in two discrete regions?

Q4. How does light pass through media?Q5. What is light?Intuition: The nature of light, and how it passes through media [Cf. Q5, Q4].

D1a. Law of refraction D1b. The only change in a restrictedstream of light passing from one medium to another (refraction aside)is a differential tendency to rotation.

D2a. All parallel rays of light D2b. Color can only be the converge into two discrete streams differential tendency to rotation after two refractions and one or produced in passing from one two reflections, emerging from the medium to another in refractiondrop (flask) in two discrete [Cf. Q3b].regions [Cf. Q3a].

D3. Parallel rays of light produce two discrete regions of color on a spherical ballof water [Cf. Q2].

D4. Sunlight (parallel rays of light) on a region of water droplets will producetwo regions of color [Cf. Q1].

followed in deducing the phenomenon shows us that the cause is thepassing of light from one medium to another, the differential tendencyto rotate this passage gives the particles of light, and the way that the law of refraction causes light rays to converge into two discretestreams at two characteristic angles. This a deduction, but a deductionof a very different sort from the one in the anaclastic line example. One can quite plausibly ask if Descartes can really be sure that he has given the true sequence of causes that produce the rainbow, asopposed to a possible sequence that produced the same appearances.Descartes himself will later come to see that as a problem.16 But in theMeteors it is not; he seems confident that the methodical procedure ofinvestigation he is following assures him that he has captured the real causes.

To return to my main thread, a number of interesting things emergefrom these two examples. First of all, it would appear that experimentfunctions strictly at the reductive stage of method, the stage in which weare trying to go from a question posed to the intuition from which theanswer is to be derived; experiment seems not to be involved in theactual deduction. And in that initial stage of inquiry, it seems to func-tion in two not altogether separable roles. First of all, it helps betterdefine the phenomenon to be deduced or the problem to be solved.This is not at issue in the anaclastic line example, where the problemis set with sufficient precision. But it is an important function of exper-iment in the rainbow example, where Descartes appeals to experimentto fix what the rainbow is, that it consists of two separate bows, and thatthe two bows are always at such-and-such an angle with respect to therays of the sun; in this way, experiment clarifies the question that is tobe answered.17 But just as important, experiment aids the reduction bysuggesting how things depend on one another, and, in that way, sug-gesting at a given juncture what question we might turn to next. It isbecause we know from experiment that refraction depends on a lightray passing from one medium to another that we know that we mustinvestigate light rays, media, and how light passes through a mediumin order to determine the law of refraction. Similarly, it is because ofexperiments with the prism that we know that reflection is irrelevant to


16 See, for example, Descartes’ remarks in Principles IV 204–206; see the discussion of thesepassages in Garber, “Science and Certainty in Descartes.”

17 See the discussion in Rule XIII, AT X 430–431, where Descartes discusses the importanceof specifying in exact terms what is being sought in an investigation.

color, but refraction is not, and it is because we know that colors canarise from the refraction of light that we know that the nature of coloris to be sought in an examination of what light is, and how it is alteredby refraction. Once we understand Descartes’ method and the rolesthat experiment does (and does not) play in it, it should come as nosurprise that Descartes might suggest that “it would be very useful ifsome . . . person were to write the history of celestial phenomena inaccordance with the Baconian method . . . without any arguments orhypotheses” (Descartes to Mersenne, 10 May 1632: AT I 251; CSMK38). The sorts of tales that Bacon recommends to the investigator inBook II of his Novum Organum can tell us, for instance, that factor A(color, say) is always accompanied by factor B (refraction, say), but thatfactor C (say reflection) is present in some cases but absent in others.In an investigation of A, this could lead us to questions about B, andprevent us from raising irrelevant questions about C, as when in therainbow example we learn that refraction is relevant to color, but reflec-tion is not. Such tables of phenomena and their correlations with oneanother, independent of any theory, are precisely what Descartes needsto define problems and to determine the relations of dependence ofone phenomenon on another necessary to perform the reductive stepof the method.

In this way, it seems that experiment is not a replacement for deduction,but part of the step preliminary to making a deduction. Science remainsdeductive for Descartes; in the end our knowledge of the cause of therainbow depends on our performing a deduction of the phenomenafrom an initial intuition. But experiment seems to play its role inpreparing the deduction. Insofar as it helps perform the reductive partof the method, the sequence of steps that leads from a question to anintuition, it helps determine the deduction, the same steps followed inreverse order that leads from intuition to the answer to the questionposed. The deductive chain that the Cartesian scientist seeks in reason,the chain that goes from more basic to less, is exemplified in the con-nections one finds in nature itself. Insofar as these latter connectionsare open to experimental determination, we can use experiment tosketch out the chain of connections in nature and find out whatdepends on what, and thus we can use the connections we find in nature asa guide to the connections we seek in reason. It may not be obvious to us atfirst just how we can go deductively from the nature of light to therainbow, but poking about with water droplets, flasks, and prisms maysuggest a path our deduction might follow.


This understanding of how experiment and observation may beuseful in a deductive science of the sort that Descartes was attemptingto construct allows us to make some sense of some of the more puz-zling aspects of Descartes’ remarks. On this understanding, we do findcauses through their effects, in a sense; experiment is quite necessaryin solving problems and helping us to discover the real causes of phenomena in our world. But in no sense are we replacing deductivewith a posteriori reasoning. Though we must appeal to experiment,experiment only prepares the deduction that will establish the cause.Furthermore, we can now see how experiment can point the way to the“correct” deduction, and eliminate the “incorrect” deductions. Therecan be alternative derivations of a given phenomenon in the sense thatthe same bare effect may be produced by different chains of causes. Forexample, a distribution of colors in a pair of bows in the sky (a bareeffect) may be produced by the reflection and refraction of lightthrough raindrops (as it actually is in our world), or by a distributionof tiny colored balls suspended in the air, or by colors projected by aslide projector on a cloud of dust, or by any number of other perversemeans. But experiment helps us find the correct deduction, that is, thecorrect chain of causes, by making the phenomenon more precise, andsuggesting how it is that the phenomenon is actually produced in thisworld. In this way experiment can lead us to the correct derivation,correct in the sense that it represents the way the phenomena arecaused in our part of the universe. Alternative deductions are notwrong, strictly speaking; one might be able to produce something thatlooks to us very much like a rainbow in any number of ways. But it’sjust that it is not the way things are done here, at least not the way it isdone in nature.

So far I have talked about experiment in the context of Descartes’official method. But, as I pointed out at the very beginning of this essay,I think that Descartes later came to set his method aside. In his laterwritings, those that follow the Discourse, I would argue that Descartesabandoned the reductive stage of his method in favor of a direct attackon the tree of knowledge, starting from intuition (or, rather, first prin-ciples, first philosophy) and deducing on down from there. But I thinkthat much of what I said about experiment in the method also holdsgood for the system-building orientation of later works like the Princi-ples of Philosophy. Though in the later writings an explicit reductive stepis not in evidence, Descartes must find some way of constituting hisdeductive chain, and here experiment will be useful for the same reason


it is in the method. It is, I think, no accident that at the moment thatDescartes was working on extending the system of the Principles fromthe inanimate world, derived by the laws of nature from an initial chaos,to the world of plants and animals, Descartes was also doing experi-ments on the formation of the fetus.18 I am certain that Descartesthought that in sexual reproduction, the development of a living bodyfrom mechanical causes, he might find clues about how living bodiesoriginally arose on this earth through mechanical causes, and that suchclues would help him extend the deduction of terrestrial phenomenabegun in the Principles to living things.

Experiment and the Priority of Reason

In the previous section I tried to show how experiment plays a role inDescartes’ scientific procedure, how experiment is needed in at leastcertain circumstances to aid in the deduction that leads us to genuineknowledge through deduction. But this raises an interesting question.Descartes is usually identified, and rightly so, as the philosopher ofreason, the philosopher who rejected the dependence on the sensesthat characterizes the Aristotelian philosophy that he was eager toreplace, in favor of dependence on clear and distinct perception, theimmediate dictates of the light of reason. I have tried to show howDescartes’ deductive science is not compromised by the way in whichhe appeals to experiment, how the particular conception Descartes has of the deductive structure of knowledge is fully consistent with theuse of reason as an auxiliary to the reductive step of his method. But a deeper question still remains, how any use of experiment at all is consistent with his strictures against the appeal to experience.

Descartes certainly does oppose naive dependence on the senses inpassages too numerous to cite; he warns us that things are not at all asour senses tell us they are, that they are not red and green, sweet orsalty, that our naive belief that all of our knowledge derives ultimatelyfrom our senses is a prejudice of sense- and body-bound youth, a prej-udice that must be rejected before we will be able to penetrate to thetrue nature of things. In his Meditations, he begins with a series of skep-tical arguments that are directed in large part, if not entirely, againstour naive trust in the senses, and in the Fourth Meditation, he appearsto recommend that we must limit ourselves to knowledge derived from


18 See Descartes, La description du corps human, AT XI 252ff.

the light of reason; he appears to argue that only by limiting ourselvesto clear and distinct perception can we guarantee that we do not strayinto intellectual sin, that is, error. And if we are to limit ourselves toclear and distinct perceptions, then there would seem to be no roomfor any appeal to experience at all, even the sort of appeal that I out-lined in the previous section.19

But, I think, the situation is a bit more complex than this textbooksummary of Descartes’ epistemology might suggest. Descartes does certainly favor reason over the senses, but he certainly does not rec-ommend rejecting the senses altogether. The fullest account ofDescartes’ views on the senses and the role that they play in the acqui-sition of knowledge occurs in the Sixth Meditation.

The reconsideration of the senses, rejected earlier in the First Med-itation, begins early in the Sixth Meditation. Earlier and unsuccessfulattempts to prove the existence of bodies led the meditator to considermore carefully the faculty of imagination and the closely related facultyof sensation (AT VII 74). And so the meditator goes back over the con-siderations that led him first to trust the senses, ending with a reviewof the considerations that originally led him to question the senses (AT VII 74–77). At this point, the meditator notes,

But now, after I have begun to know myself and my author a bit better, I donot think that everything that I seem to get from my senses should simply beaccepted, but then I don’t think that everything should be rendered doubt-ful either. (AT VII 77–78)

The senses loom large in the rest of the Meditation. The meditatorfirst distinguishes between the mind and the body. Then the questionturns to the external world, and it is here that the senses make theirfirst positive contribution to the enterprise. The meditator begins:“Now there is in me a certain passive faculty of sensing, that is, of receiv-ing and knowing the ideas of sensible things” (AT VII 79). We have apassive faculty of sensation. But this would be of use only if there were,somewhere, an active faculty for producing these ideas, a cause. This,Descartes argues, could not be in me, for it seems to involve neither myunderstanding nor my will, the two faculties I have. So, the meditatorreasons, the ideas of sensation he has must come from outside of him,


19 For a development of some of these themes in Descartes, see Garber, “Semel in vita: theScientific Background to Descartes’ Meditations,” in Essays on Descartes’ Meditations, ed., A. Rorty (Los Angeles and Berkeley: University of California Press, 1986), 81–116, essay11 in this volume.

either from God or from bodies (i.e., bodies as understood in the FifthMeditation, things extended and extended alone) or from somethingelse. The meditator reasons that it must be from bodies themselves that our ideas derive; God has given me a “great propensity for believ-ing that they come to me from corporeal things,” while he has givenme “no faculty at all” for learning that this propensity might be mis-taken (AT VII 79–80). So, the meditator argues to himself, God wouldbe a deceiver if it turned out that our ideas of bodies come from any-where else but from bodies themselves. And so, he concludes, bodiesexist.20

The argument is a very interesting one. A conclusion is establishednot because we have a clear and distinct perception that bodies exist,exactly, but because the meditator has a “great propensity” for believingsomething, and God has given him no way of correcting that propen-sity.21 Descartes admits here that there are at least some circumstancesin which a belief that we seem to get from sensation, the inclination tobelieve that seems to come to us with the sensation, is worthy of our trust.It may not be as worthy of our trust as a genuine clear and distinct per-ception, as he implies in the Synopsis of the Meditations (AT VII 16), andit may not always be true, as a clear and distinct perception is. But whensensation leads us to a belief, as it does in this case, and when that beliefis not overridden, as it were, by a reason for rejecting it, as is the casewith our beliefs about colors actually being in things, say, then we cantrust the senses.22 This is the strategy that Descartes pursues in theremainder of the Sixth Meditation in his discussion of the senses. Heargues that what he calls the “teachings of nature,” which include thebeliefs that appear to arise spontaneously with sensations, can be trustedas being for the most part true when corroborated by reason, that is whenreason does not give us better grounds for rejecting a judgment from the


20 For a fuller presentation of this argument, see ibid.21 In the version of the argument given in Principles II 1, Descartes does seem to argue

from the fact that “we seem clearly to see” that sensation proceeds to us from the objectof our idea of body to the real existence of body, and does not appeal to the “great propen-sity” that is the nub of the argument in the Meditations. It is not clear why the later textdiffers from the earlier one on this point. It may represent a genuine change in Descartes’epistemology. But then it may simply reflect Descartes’ desire not to enter into his fullaccount of the senses in the Principles. For the relation between the Meditations and thePrinciples, see Garber and Cohen, “A Point of Order,” essay 3 in this volume.

22 That is, we can trust at least some of the judgments that characteristically accompany oursense perceptions. What seems to be at issue here is the third of Descartes’ three gradesof sensation; see AT VII 436–437.

senses, or when reason is in accord with that judgment, or when reasonis silent on the question.

As with clear and distinct perceptions, Descartes is here dealing withsomething that God gave us, beliefs that are, in a certain sense, innate:“I am dealing only with those things that God gave me as a compositeof mind and body” (AT VII 82). As such, Descartes argues, they mustbe in some sense true: “It is doubtless true that everything that natureteaches me has some truth in it” (AT VII 80). When it is truth aboutthe nature of things that we are interested in, it is the light of reason,clear and distinct perceptions, that we must turn to first. Descarteswrites,

And so, my nature teaches me to flee what gives me pain and to seek whatgives me pleasure, and the like. But it does not appear that it teaches us toconclude anything about things outside of us from the perceptions of thesenses without a prior examination of the intellect, since knowing the truth aboutthings seems to pertain to the mind alone, and not to the composite [of mindand body]. (AT VII 82–83; emphasis added)

And so, while some of the teachings of nature will turn out to be true,it is only the intellectual examination of them that will establish this. Inthis way Descartes restores the senses and rejects the hyperbolic rejec-tion of the senses that begins the Meditations; indeed, he goes on toreject even the dream argument that is so prominent in the First Med-itation (AT VII 89–90). But though the teachings of nature, what welearn from our senses, are restored, they are subordinate to reason; theymay be trusted to some extent and in some circumstances, but onlyafter they have been given a clean bill of health by reason.

It is with this in mind that we should return to the use of experimentin the rainbow case discussed earlier. One can say that insofar asDescartes does allow the appeal to the senses in at least a general way,there is no inconsistency in Cartesian epistemology; as long as whatDescartes takes from the experiments to which he appeals falls withinthe bounds of proper caution, there is no special problem here. Butthere is something more interesting to be said in this case about theway in which experience is subordinate to reason.

In the previous section, I showed that while experiment might func-tion as an auxiliary to a deduction, it is the deduction itself and not theexperiment that yields the knowledge. So, for example, in the anaclas-tic line case, while experience might suggest to us that there is somelawlike relation between angle of incidence and angle of refraction, it


is only through deduction that the actual law can be established (seeRule VIII: AT X 394). But the point goes deeper still. In the rainbowcase, Descartes begins by observing that on his flask, the stand-in forthe raindrop, there are two regions of color, at roughly 42 and 52degrees from the ray of sunlight, which angles are then deduced in theend from his theory. After giving his account, Descartes notes that anearlier observer, the sixteenth-century mathematician Franciscus Mau-rolycus, set the angles incorrectly at 45 and 56 degrees, on the basis offaulty observations. Descartes notes that “this shows how little faith oneought to have in observations which are not accompanied by the truereason” (AT VI 340; Ols 342).23 It is only because we can calculate theangles of the primary and secondary bows from the account we haveof the rainbow that we can be sure of what they are, despite the fact thatthe investigation began with an experimental determination of thoseangles.24 Though it is an observation that starts the ball rolling, it is onlythrough a Cartesian deduction that the phenomena and causal depen-dencies observed can actually enter the body of scientific knowledge,strictly speaking. Similarly, it is only because a deduction can, indeed,be made in the reverse order of the causal dependencies that experi-ment has found, that those dependencies ought to be trusted. Descartesis, of course, aware that color can arise not only from refraction of light,but from the reflection of light off of a surface whose texture is appro-priate to cause the changes in the light necessary to produce the colorseen. At one point in his discussion of the rainbow Descartes seems pre-pared to consider such an account of color in the rainbow, because, atfirst glance, the restriction on the beam of light necessary to producecolor through refraction seems to be absent (AT VI 335; Ols 338–339).And so, it seems, the causal dependence of the colors of the rainbowon refraction and reflection suggested by experiment is only provi-sional; while the experimental determination of the path the lightfollows through the droplet may suggest to us a deductive path that we


23 For a discussion of Maurolycus’ theory of the rainbow, see Boyer, The Rainbow, 156–163.The implication of Descartes’ remarks is that Maurolycus’ values for the angles derive fromobservation alone. This is not entirely fair. Maurolycus had his reasons for setting theangles as he did, reasons based on his (incorrect) analysis of the path the light followswithin the raindrop; indeed, he knew that his calculated value differs from what was known through observation, something for which he attempted to offer an explanation(pp. 159–160).

24 We must, of course, remember that the calculation does appeal to an experimentally deter-mined value for the index of refraction; however, as I pointed out earlier, Descartes wouldsurely have thought that a “reason” could be given for that too.

might be able to follow, it is the actual success of the deduction fromintuition to phenomena that actually establishes the causal connectionsthat produce the phenomena. Experiment is important in helping tofind the deduction, but it is the deduction that, in an important sense,fixes both the causal path and the phenomena. Experience is impor-tant, but only under the control of reason, as Descartes took great painsto emphasize in the Sixth Meditation.

This feature of Descartes’ position connects in an interesting waywith an often discussed problem in the philosophy of science, the ques-tion of the theory-ladenness of observation. Whether or not one can have an observation that is not in an important way dependent onsome theory or other is a question too often discussed in the abstract.Descartes’ appeal to experiment in the rainbow case shows an inter-esting complexity in the whole dispute. Descartes does use observationto motivate the theory that he is proposing, or, perhaps, to guide us tothat theory. In this sense, observation would seem to be a-theoreticalfor Descartes. But at the same time it is extremely important to realizethat the observations Descartes presents as motivating his account ofthe rainbow, or at least guiding it, are not to be trusted fully until wehave an account of the matter, until we can derive those observationsfrom more basic principles. There is such a thing as pre-theoreticalobservation for Descartes, and this does seem to have a role to play in his procedure. But, at the same time, there is an important sense inwhich observation does not attain the status of fact until it becomes inte-grated with theory, indeed, until it becomes subordinated to theory.

In this way, for Descartes, experiment by itself can establish no facts;while experiment can lead us to facts, it is only the final deduction ofa phenomenon from intuited first principles that establishes the cre-dentials of a fact, even if first “discovered” through experiment. In hisrecent writings, Ian Hacking argues that experiment must be viewed asin an important sense independent of theorization in science; “experi-ment has a life of its own,” he insists.25 By this he means to point out,among other things, that experiment does not function exclusively inthe service of theoretical argument, furnishing premises for theoreti-cal arguments, testing theories proposed, allowing us to eliminate oneof a pair of competing theories and accept another, and so forth. Thismay be true enough for a wide variety of figures. But it is not true for


25 Ian Hacking, Representing and Intervening (Cambridge: Cambridge University Press, 1983),150.

Descartes. For Descartes, at least in the context of the rainbow, exper-iment plays a carefully regimented role in what is from the start a theoretical project. But, at the same time, neither do experimental phe-nomena have a role assigned to them in standard hypothetico-deduc-tive conceptions of scientific method, as the touchstone of theory, thea-theoretical facts to which we can appeal to adjudicate between alter-native theories. If my account of experiment is correct, then howevermuch experiment might help us to find the correct account, it is ultimately reason, not experiment, that is the touchstone of reality, fortheory as well as for the experimental facts that help us construct theories.26

On the standard view of things, widely shared since the late eigh-teenth century or so, there are two sorts of philosophers: rationalistsand empiricists. Descartes is traditionally viewed as a rationalist, in fact,the founder of the school, in modern times at least. When the extentof Descartes’ dependence on experiment and observation is recog-nized, there is a temptation simply to think that Descartes must havebeen placed in the wrong slot, and conclude that he must really be somesort of empiricist.27 I would resist that temptation. It seems to me thatwhat the case of Descartes shows is how crude the scheme of classifi-cation really is. For Descartes both reason and experience are impor-tant, though in different ways. His genius was in seeing how experienceand experiment might play a role in acquiring knowledge withoutundermining the commitment to a picture of knowledge that had moti-vated him since his youth, a picture of a grand system of certain knowl-edge, grounded in the intuitive apprehension of first principles.


26 Descartes does say some things that would appear to go against my conclusion. Forexample, immediately following the long passage from Part VI of the Discourse I quotedearlier, Descartes writes,

I know of no other means to discover this [i.e., how a particular effect depends on thegeneral principles of nature] than by seeking further experiments [expériences] whoseoutcomes vary according to which of these ways provides the correct explanation. (AT VI 65)

But, I think, this must be understood in the context of the interpretation I have offeredearlier. The experiments in question must be viewed as leading us down one deductive pathrather than down another, and not as a theory-neutral means of choosing between inde-pendently constructed theories; for, as Descartes elsewhere insists, we cannot really be sure of an experimental fact until after we have already determined what the correct deduction is.

27 See, e.g., Clarke, Descartes’ Philosophy of Science, 205.

6

DESCARTES ON KNOWLEDGE AND CERTAINTY

From the Discours to the Principia

Descartes is usually classed among the rationalists, those philosopherswho privilege reason over experience. And indeed he belongs there.On the other hand, though, Descartes was also very interested in exper-iment. The Dioptrique and Météores make a number of references toDescartes’ experiments; the Discours discusses the importance of exper-iments at some length. In the Principia, written starting in early 1641and published in 1644, Descartes refers to a number of experimentalresults to support his views, most visibly in the discussion of the magnet.And at the end of that book, he goes so far as to suggest that his visionof the world is ultimately supported by the fact that it is capable ofexplaining observed phenomena, and nothing more. Where is the real Descartes? Is he mathematician or experimenter? rationalist orempiricist?

This is the larger question that I would like to explore in this essay.But I would like to address it in a rather particular and somewhat specialway. Generally, discussions of Descartes’ views about knowledge andexperience concentrate on texts like the Meditations, and on issues con-cerned with knowledge of the kinds of grand questions that he takesup there, the knowledge of self, body, the distinction between mind andbody, God, and so on. What I want to focus on is something much moremundane. The water we drink every day has a nature, from which followcertain well-known properties; water is wet and liquid at room temper-ature, solid when very cold, quenches thirst, admits light, but causescertain illusions, like the famous bent-stick illusion. All of this issomehow connected with its structure. The question I want to addressis this: how did Descartes think that we could know the internal

111

structures, the natures of particular things like water and wine, goldand wood?

To appreciate Descartes’ problem here we must make a few back-ground remarks. The view about individual natures that Descartes andhis contemporaries learned in school was straightforwardly based onAristotelian principles. According to Aristotle, water (maybe not thewater we encounter in everyday life, but pure water) is an element,defined by a particular substantial form. That form, joined to barematter, gives water the characteristic properties that it has. And so, onthis view, water is just the kind of stuff that by its nature tends to be cold,wet, and liquid, that tends by its nature to fall below the sphere of airand above that of earth, to name two others of the Aristotelian ele-ments; these are just its innate tendencies to behavior. And that is allthe explanation that one can give, period. Mixtures of the elements andtheir properties add considerable complexity to the question. But eventhen, the idea that things have natures, substantial forms that give theminnate tendencies to exhibit the manifest properties they do is basic tothe Aristotelian scheme of things.

But Descartes and his mechanist friends take another view alto-gether. According to Descartes, all body is of the same nature; every-thing in the physical world is extended substance, and its tendencies tobehavior are defined by the laws of motion. Descartes writes in the Principia:

The matter existing in the entire universe is thus one and the same, and it isalways recognized as matter simply in virtue of its being extended. All theproperties which we clearly perceive in it are reducible to its divisibility andconsequent mobility in respect of its parts, and its resulting capacity to beaffected in all the ways which we perceive as being derivable from the move-ment of the parts.1

How then are we to explain the special properties water has? AsDescartes suggests, we can only appeal to “its divisibility and consequentmobility in respect of its parts”. That is, the special properties this waterhas can only be explained in terms of the size, shape, and motion ofthe tiny parts that make it up. Different samples of water presumablyhave a common structure of smaller bodies, corpuscles, whose


1 Principia II, 23. For a fuller account of Descartes’ mechanist program in contrast with Aristotelian hylomorphism, see D. Garber, Descartes’ Metaphysical Physics, Chicago, Univer-sity of Chicago Press, 1992, chapters 3, 4.

characteristic size, shape, and motion give it its characteristicallyobserved properties. Let us call this particular nature its “corpuscularsubstructure.”

This, then, is the problem I would like to explore in Descartes. Forthe scholastic scientist, the characteristic properties of a thing derivefrom a form, often a hidden form, an occult quality. For Descartes,there are no such occult qualities. But there are hidden natures, cor-puscular substructures that are hidden from our view. How can they befound?

There are actually a number of questions here that we might sepa-rate. First of all, how do we discover these hidden mechanisms? Andhaving conjectured a particular candidate for a corpuscular substruc-ture, how do we justify the claim that we have found the correct one?And in this argument, I want to ask, what role does experience and/orexperiment play? And finally, what are the limits of certainty withrespect to our knowledge and belief in the corpuscular substructuresof particular kinds of things?

In order to answer these questions, we must, I shall argue, distinguishthe positions that Descartes takes at different times in his career. Andso, we shall proceed chronologically. First we shall examine the viewsDescartes seems to have had in mid-1630s, when he was completing hisfirst works for publication, the Discours and the accompanying Essais.Then we shall turn to his views a few years later, in the early 1640s, com-posing the Principia. Despite appearances, there is, I shall argue, aradical change between Descartes’ views at the one time and the other.Descartes, I shall claim, moves from the position that we can havegenuine certain knowledge of the corpuscular substructure, to therather different view that our conjectures about corpuscular substruc-tures are at best devices that enable us to predict future experience,and in that way prolong our lives.

1. Knowledge of Particulars in the Discours and Essais

First, then, let us turn to Descartes in the period of the Discours andEssais. The most extensive discussion of the issues connected withknowledge of particulars takes place in Part VI of the Discours, whereDescartes goes to some length to talk about the need for experiments,and argues that his program could progress only if he had sufficientfunds for doing experiments. (Even 350 years ago, scientists had to begfor the money they needed to keep up their laboratories!) The passage

descartes on knowledge and certainty 113

begins with a lengthy account of where experiment is not really neces-sary. Descartes reports that he began his investigations with “the firstprinciples or first causes” of everything, which can be discovered from“certain seeds of truth which are naturally in our souls.” From thisDescartes derived “the first and most ordinary effects that one candeduce from these causes,” the heavens, stars, the earth, water, air, fire,etc. The passage then continues as follows, and addresses more directlyhow it is that we can come to know the corpuscular substructures thatunderlie the greatest part of the particulars we know from everydayexperience:

Then when I wanted to descend to those which were more particular, I waspresented with so many different kinds of things that I did not think that itwas possible for the human mind to distinguish the forms or kinds of bodieswhich are on the earth from an infinity of others that could have been there,if God had wanted to put them there, nor, consequently, to make them usefulto us, unless one proceeded to the causes through their effects, and attendedto many particular experiments.2

It is not easy to interpret this passage. But it is not surprising thatthis is often read as endorsing a certain conception of how we can knowthe natures of particular things by way of what we might call hypothet-ical argument.3

What I call hypothetical argument is suggested later in the Discourswhere Descartes discusses a curious feature of the Météores and Diop-trique. In both of these treatises, Descartes begins by making certain“suppositions,” assumptions or hypotheses about the nature of light, themake-up of water, oils, etc., from which he then derives various featuresof the world. In the Météores, for example, Descartes writes, in a chapterentitled De la nature des cors terrestres:


2 AT VI 64.3 For some recent interpretations along that line, see, e.g., D. Clarke, “Descartes’ Philosophy

of Science and the Scientific Revolution,” in J. Cottingham ed., The Cambridge Companionto Descartes, Cambridge, Cambridge University Press, 1992, pp. 258–285, and Ettore Loja-cono, “L’attitude scientifique de Descartes dans les Principion,” in J.-R. Armogathe andGiulia Belgioioso, eds., Descartes: Principion Philosophiae (1644–1994), Naples, Vivarium,1996, pp. 409–433. What I call hypothetical argument Lojacono calls “le procédé par sup-position.” He emphasizes that this “procédé” should not be called a method, a term thatDescartes reserves for the very different procedure he outlines in the Regulae. I would liketo thank him for correcting my more careless use of language in an earlier draft of this essay.

I assume that the small particles of which water is composed are long, smooth,and slippery like little eels, which are such that however they join and inter-lace, they are never thereby so knotted or hooked together that they cannoteasily be separated; and on the other hand, I assume that nearly all particlesof earth, as well as of air and most other bodies, have very irregular and roughshapes, so that they need be only slightly intertwined in order to becomehooked and bound to each other, as are the various branches of bushes thatgrow together in a hedgerow.4

In a slightly later passage in Discours VI, Descartes writes about thesesuppositions as follows:

Should anyone be shocked at first by some of the statements I make at thebeginning of the Dioptrique and the Météores, because I call them “supposi-tions” and do not seem to care about proving them, let him have the patienceto read the whole book attentively, and I trust that he will be satisfied. For Itake my reasonings to be so closely interconnected that just as the last areproved by the first, which are their causes, so the first are proved by the last,which are their effects. . . . For as experience makes most of these effects [i.e.,observed phenomena] quite certain, the causes from which I deduce themserve not so much to prove them as to explain them; indeed, quite to thecontrary, it is the causes which are proved by the effects.5

To say that the causes are “proved by effects,” as Descartes does, sug-gests very strongly the causes conjectured are established as true by thefact that they are capable of explaining the observed phenomena.There are many other passages from this period, both in the publishedtexts and in the letters that suggest much the same. But the view comesout most clearly in the writings of one of Descartes’ followers, theFrench physicist Jacques Rohault. Writing in his Traité de physique of1671, only a bit more than 20 years after Descartes’ death, he gives hisversion of the proper way of building a natural philosophy:

In order to find out what the Nature of any Thing is, we are to search forsome one Particular in it, that will account for all the Effects which Experi-ence shows us it is capable of producing. Thus, if we would know what theHeat of the Fire is, we must endeavour to find out some particular Thing, bymeans of which, it is capable of producing in us that Sort of Tickling, or pleas-ant agreeable Heat which we feel at a little distance from it. . . . In a word, itmust explain all the Effects that Fire produces. . . . What is now said of Heat,may be applied to all other Things: And by this Rule, every Thing hereafter


4 AT VI 233. 5 AT VI 76; cf. AT II 141–144, 199–200; AT VI 334.

is to be examined, If that which we fix upon, to explain the particular Natureof any Thing, does not account clearly and plainly for every Property of thatThing, or if it be evidently contradicted by any one Experiment; then we areto look upon our Conjecture as false; but if it perfectly agrees with all theProperties of the Thing, then we may esteem it well grounded, and it maypass for very probable.6

This illustrates what I earlier called hypothetical argument. An hypo-thetical argument for some conclusion proceeds as follows. We aretrying to explain some feature of the physical world, say fire and itsheat. We first conjecture a structure of smaller particles in motion; thisis the hypothesis about the nature of the fire that is under considera-tion. We might hypothesize, for example, that fire is made up of small,dagger-shaped corpuscles that move very, very fast. This hypothesis isthen tested against experience; if it is capable of explaining all experi-ments that we can make on fire, and clearly contradicts none, then theconjecture is esteemed “well grounded, and it may pass for very prob-able.” For example, we may imagine that the pain we experience whenputting a finger in the fire is explained by the dagger-like shape andmotion of the particles that make up the fire. But if a conjecture “beevidently contradicted by any one Experiment; then we are to lookupon our Conjecture as false,” says Rohault.

But, I must insist, it is quite wrong to attribute this view to Descartesin the period of the Discours. Immediately after the above quotedpassage, Descartes writes:

And I have called them “suppositions” simply to make it known that I thinkthat I can deduce them from the primary truths I have expounded above;but I have deliberately avoided carrying out these deductions in order toprevent certain ingenious persons from taking the opportunity to construct,on what they believe to be my principles, some extravagant philosophy forwhich I shall be blamed.7

But if it is not the procedure of hypothetical argument that Descartesis espousing here, then what is it? And how does his evident interest in experiment fit in, if it isn’t an hypothetical argument that is at issue here? What kind of deduction does Descartes have in mind here?


6 J. Rohault, A System of Natural Philosophy, Illustrated with Dr. Samuel Clarke’s Notes . . . Done intoEnglish by John Clarke, 2 vols., London, James Knapton, 1723, vol. I, pp. 13–14.

7 AT VI 76; cf. AT I 563; AT II 200; AT III 39.

Let me begin with a brief example from an earlier work, the Regulaead Directionem Ingenii. The assumption behind the method Descartespresents in that book is that real knowledge, knowledge worthy of thename, derives from intuition and deduction. Intuition is a faculty wehave by virtue of which we are capable of grasping truths directly;deduction is a complementary faculty, by virtue of which we can intuitthe connections between one proposition and another.8 Descartes’method in the Regulae consists of a reduction, followed by an intuition,followed by a construction, that is, a deduction of the answer to thequestion originally posed, starting from the intuition that we haveattained.9 This is what we might call the appeal to intuition and deduc-tion, or, more simply, the appeal to intuition, as distinct from the sortof hypothetical argument I noted earlier. If Descartes is right, then allknowledge is derived by deduction from intuition.

In the text of Rule 8 Descartes gives an example of his celebratedmethod.10 The question at issue is the shape of a particular lens, onethat is capable of focusing parallel rays to a single point. The reductionstarts with the question posed, the shape of the lens in question, andleads us back from that by posing a series of presupposed questions. Inorder to determine the shape of the lens in question, we must deter-mine the law of refraction, i.e., the law that governs the bending oflight. But in order to determine that, we must determine the way light is altered when it passes from one medium to another. But todetermine that we must determine how light passes through a medium.Ultimately, we are led back to the question of the nature of a naturalpower. Intuiting the answer to that question, we then pass back theother way, intuiting from the nature of a natural power the answers tosuch questions as the nature of light, the way it passes through amedium, and ultimately, the law of refraction and the shape of the lensin question.

Although he is not terribly explicit about it, I think that this proce-dure is what is behind the view in the Discours and the Essais. This comesout reasonably clearly in Descartes’ treatment of the rainbow in theeighth discourse of the Météores.11 There also it is more evident just how


8 See Regula III, AT X 366 ff. 9 See Regula V and VI, AT X 379 ff.10 See AT X 393 ff.11 The rainbow is discussed in AT VI, 325 ff. For a more detailed discussion of this case, see

D. Garber, “Descartes and Experiment in the ‘Discourse’ and ‘Essays’ ” in S. Voss, ed.,Essays in the Philosophy and Science of René Descartes, New York, Oxford University Press, 1993,pp. 288–301, essay 5 in this volume.

important experiment is to Descartes. The problem posed is the expla-nation in corpuscular terms of the phenomenon of the rainbow.Descartes begins with the experimental and observational fact that therainbow consists of two bows of color that are always at a characteristicangle with respect to sunlight, 42 and 52 degrees, to be exact. Experi-ment is then appealed to, again, to reduce the question, the cause ofthe observed phenomena, to simpler questions. For example, the factthat color can be produced in a prism shows that the cause of color hasnothing to do with a curved surface, and arises when light passes fromone medium into another; in this way, the question of the genesis ofcolor is “reduced” to the question as to how light is changed in passingfrom one medium to another. One proceeds in this way until reachingsomething about which one has direct intuitive knowledge, in this casethe nature of a natural power. The causal explanation is completedwhen one can do a derivation of the observed phenomena from theintuition of the most general principles, using the causal paths sug-gested by the auxiliary experiments and observations that constitute thereduction (“from the way in which light changes when passing fromone medium to another, it follows that color is . . .”); the derivation ofthe phenomena from the most general causes then displays the causalexplanation.12 In this case, Descartes uses this kind of procedure toestablish, with certainty, presumably, the corpuscular substructure thatconstitutes the rainbow: it is an arrangement of water droplets of appro-priate size and arrangement to reflect and refract the incoming sun-light and cause the bands of color that we see in the sky.

This obviously uses experiment, like hypothetical argument, and likehypothetical argument, the point seems to be to fit a hypothesis to thephenomena. But there are important differences. First of all, themicrostructure of the rainbow is not hypothesized to fit the phenom-ena; it is derived making use of the phenomena. At no point in the pro-cedure does one make a hypothesis; when properly used, observationis supposed to lead us directly to the underlying mechanism. But moreimportant, unlike hypothetical argument, the phenomena have novalidity independent of the causal explanation proposed. Descartes’account of the rainbow begins with an observation about the charac-teristic angles of the two bows that make up the rainbow; at the end ofthe argument, these angles are derived from his causal account.Descartes remarks, though, that other observers have observed differ-


12 See D. Garber, “Descartes and Experiment,” cit., pp. 95–101 in this volume.

ent values for these angles. His comment is very significant: “This showshow little faith one should have in observations that are not accompa-nied by the true reason.”13 It is only after we give an explanation, aderivation of the phenomena from first principles that the phenomenaenter the realm of genuine facts, despite the fact that it was theobserved phenomena that started the process in the first place. Obser-vation and experiment may pose problems for us, and may suggestcausal paths for their explanation, but they are not facts until they aresuccessfully deduced from first principles. This is a use of experiment,to be sure, but not an hypothetical argument. Experience is used ratherin the way in which we use diagrams in geometry. We can carefully drawdiagrams on paper, carefully measure sides, angles, and arcs, andhypothesize relationships. But it is only the actual proof of a theoremthat establishes anything as true.

But why didn’t Descartes want to make the direct appeal to intuitionmore visibly than he did in the Discours and Essais? Why did he think itnecessary to use hypothetical argument if it was intuition that he reallypreferred?

Let me remind you of a passage I cited earlier. Referring to the sup-positions he actually used in the Météores and the Dioptrique, Descarteswrites:

And I have called them “suppositions” simply to make it known that I thinkthat I can deduce them from the primary truths I have expounded above;but I have deliberately avoided carrying out these deductions in order toprevent certain ingenious persons from taking the opportunity to construct,on what they believe to be my principles, some extravagant philosophy forwhich I shall be blamed.14

Now it is clear enough how the appeal to intuition involves us in thefoundations of Descartes’ philosophy; the intuitions that he leads us towill reveal the very foundations of his natural philosophy, the nature ofbody, the nature of a natural power, the nature of light, etc. But whywould he want to hide that? Why shouldn’t he, on the contrary, wantto proclaim his new ideas to the world in the most public way?

To understand why not, we must put ourselves back into Descartes’shoes in the mid-1630s. Descartes knew that matters were somewhatdelicate as far as he was concerned. He knew that his philosophy wasin contradiction with the official Aristotelian philosophy, taught both


13 AT VI 340. 14 AT VI 76.

at Catholic universities, like the University of Paris, and at Protestantschools, like the University of Utrecht. That didn’t seem to bother himwhen he wrote Le Monde in the early 1630s, explicitly attacking thesterility of the Aristotelian orthodoxy, indeed, even mocking the Aris-totelian definition of motion! But Descartes was seriously taken abackby the condemnation of Galileo in 1633. He withdrew Le Monde, and,indeed, renounced any ambitions to publish his thought. That did notlast long, though, and within a short time, Descartes was making plansfor a new publication, the Discours and Essais. But there was to be acrucial difference between Le Monde and this later work. Le Monde toldall, and gave the foundations of Descartes’ thought, which made it clearthat he rejected forms and qualities, and placed the sun at the centerof the planetary system, making the earth just another planet. But inthe Discours and Essais, all of this was to be hidden. In writing the Essais,Descartes hoped only to “choose some topics which would not be toocontroversial, which would not force me to divulge more of my princi-ples than I wished to, and which would demonstrate clearly enoughwhat I could or could not do in the science.”15

In the mid- and late 1630s, then, we have a rather clear answer tothe questions posed about the knowledge of particulars. While it maybe hypothetical argument that Descartes chooses for presenting hisconclusions in his published works, it is really the appeal to intuitionthat is close to his heart. Intuition, the immediate apprehension oftruth, and its coordinate faculty, deduction, lead us to a comprehen-sion of the particular nature, the corpuscular substructure that is theground of the manifest properties of things. While experiment comesin, it is just an auxiliary to the intuition and deduction. Reason wouldseem to reign, with experiment in the subordinate position of a trustedadvisor, at best.

Or so it would seem. But all is not well. The intuition that is at thecore of Descartes’ solution to the problem of particular natures andcorpuscular substructures at this time is profoundly mysterious. Can wereally intuit the nature of a natural power? The nature of light? Under-standing deduction in the strict sense Descartes intends, can we reallydeduce from these things that we are supposed to know the way inwhich color arises in the rainbow from the passage of light from onemedium to another? As attractive as the view in the Regulae and Discours


15 AT VI 75. For a fuller account of the story, see D. Garber, Descartes’ Metaphysical Physics,chapter 1.

may seem to be at first glance, I find it ultimately very unsatisfying.Though it promises to reveal the hidden nature of things, to makeoccult qualities and hidden mechanisms manifest, the process by whichsuch hidden natures and mechanisms are supposed to be revealed, theintuition and deduction to which Descartes appeals are themselveshidden and occult in the extreme. This is particularly so when appliedto the knowledge of particular natures. And, I suspect, Descarteshimself came to realize that as well.

2. Knowledge of Particulars in the Principia Philosophiae

Now I would like to turn away from the period of the Discours and Essais,and toward the later period, in the early 1640s, when Descartes wasworking on his Principia Philosophiae. From the very title it is evident thathe had set aside his earlier scruples, and had decided that he wouldpresent his whole system, his true views grounded in their proper meta-physical foundations. The Discours had been published, as had the Med-itations, and the sky had not fallen on Descartes’ head. So, he judged,it was the time to reveal all. And so, in the Principia Descartes beginswith his first philosophy (Part I), proceeding from there to the generalpart of his physics (Part II), including the nature of body, motion, andthe laws of motion, before descending to particulars in the final twoparts, cosmology and the heavens (Part III) and terrestrial physics (PartIV). In the Principia, especially the final two parts, Descartes is verymuch concerned with the nature of particulars, their corpuscular sub-structures, and offers a number of specific analyses. But, we might ask,how does he claim to have knowledge of such particular natures andcorpuscular substructures?

There is reason to believe that Descartes was still, in a way, attractedby the appeal to intuition that we saw displayed in the case of the lensfrom the Regulae or the rainbow example in the Météores, and still some-what suspicious of hypothetical argument, at least as a preferred way ofestablishing his conclusions. For example, in the opening sections ofPart III, Descartes presents a variety of astronomical phenomena to con-sider. Before presenting them, though, he writes:

Our purpose is not to use these phenomena as the basis for proving anything[ad aliquid probandum], for we aim to deduce an account of effects from theircauses, not to deduce an account of causes from their effects. The intentionis simply to direct our mind to a consideration of some effects rather than


others from among the countless effects which we take to be producible fromthe selfsame causes.16

Similarly, in the middle of Part IV, Descartes presents an account ofthe nature and corpuscular substructure of the magnet. Some sectionsafter presenting his account of the magnet, he gives a list of thirty-fourphenomena that he thinks can be explained by his theory. This list ofexperimental facts is preceded by the following remarks:

All these things [i.e., the previously presented account of the nature of themagnet] follow from the principles of Nature expounded above, in such away that even if I were not considering those magnetic properties which Ihave undertaken to explain here [i.e., the experimental phenomena that heis about to present], I nonetheless would not judge these things to be other-wise. However, we shall see that with their help [i.e., the help of the accountof the magnet, iron, etc., that which follows from his principles] a reason forall those properties is furnished so clearly and perfectly that this fact alsowould seem sufficient to convince us of the truth of these things, even if wedid not know that they followed from the principles of Natreu.17

Descartes here does admit that the properties of the magnet discov-ered by experiment certainly do support the account he earlier gave ofthe nature of the magnet on other grounds, and had he not had sucha priori grounds, the experimental fit would have sufficed. As in theDiscours and Essais, it does seem as if hypothetical argument is, at best,a second-best form of argument, and, one might suppose, the appealto intuition is to be preferred.

But I think things are more complex than that. Indeed, there is verygood reason to believe that in the Principia, Descartes ends by com-mitting himself to the very sort of hypothetical argument that he soclearly rejected in his earlier writings.

Descartes turns to the question of the knowledge of particularnatures and corpuscular substructure at the very end of the book, inPart IV. Descartes reminds us that on his view, bodies are made up ofsmall particles, corpuscles too small for us to see. “Who can doubt,” hewrites, “that there are many bodies so minute that we do not detectthem by any of our senses?”18 But, as Descartes realizes, this raises animportant epistemological question: “In view of the fact that I assigndeterminate shapes, sizes, and motions to the imperceptible particlesof bodies just as if I had seen them, but nonetheless maintain that theycannot be perceived, some people may be led to ask how I know what


16 Principia, III 4. 17 Ibid. IV 145. 18 Ibid. IV 201.

these particles are like.”19 His answer is this (at least in the Latin editionof 1644, a bit expanded in the French edition of 1647):

First of all I took the simplest and best-known principles, knowledge of whichis naturally implanted in our minds; and working from these I considered,in general terms, first, what are the principal differences which can existbetween the sizes, shapes, and positions of bodies which are imperceptibleby the senses merely because of their small size, and second, what observableeffects would result from their various interactions. Later on, when I observedjust such effects in objects that can be perceived by the senses, I judged thatthey in fact arose from just such an interaction of bodies that cannot be per-ceived – especially since it seemed impossible to think up any other expla-nation for them.20

The view seems quite clearly to be what I called hypothetical argu-ment earlier. We begin by conjecturing an hypothetical substructurethat, we hope, explains the phenomena. For example, we suppose thatwater is made of eel-shaped particles of a particular size and shape, orthat earth is made up of branch-shaped particles that interconnect withone another. From the conjectured structure we then derive observa-tional consequences, consequences which are compared against expe-rience. To the extent, then, that the observable consequences tallyagainst what we actually observe in the world, the conjectured struc-ture, the conjectured cause, is “proved by the effects.” After a briefexposition of what would appear to be hypothetical argument,Descartes adds a further refinement of this:

But we shall know that we have determined such causes correctly afterwards,when we notice that they serve to explain not only the effects which we wereoriginally looking at, but all these other phenomena, which we were notthinking of beforehand.21

In this way, the proof is stronger to the extent that the effectsexplained were not known at the time that the hypothesis was first putforward.

Descartes also notes the way in which he came to formulate thehypotheses that he puts to the test. He writes:

In this matter I was greatly helped by considering artifacts. For I do not recognize any difference between artifacts and natural bodies except that the operations of artifacts are for the most part performed by mechanisms which are large enough to be easily perceivable by the senses. . . . The effects


19 Ibid. IV 203. 20 Ibid. IV 203. 21 Ibid. III 42.

produced in nature, by contrast, almost always depend on structures whichare so minute that they completely elude our senses.22

And so, Descartes proposes:

No one who uses his reason will, I think, deny the advantage of using whathappens in large bodies, as perceived by our senses, as a model [exemplum]for our ideas about what happens in tiny bodies which elude our sensesmerely because of their small size.23

Putting all these considerations together, the procedure seems to be this, in the end. We begin with our experience of everyday things inthe world, machines of various sorts, perhaps bushes whose branchesbecome entangled, grapes in vats, eels in buckets. On the basis of thisexperience we conjecture possible substructures to explain the behav-ior of things. For example, we might conjecture that the particles ofearth are branched like bushes, to explain why it is that earth coheresin solid clumps, or that water is made up of eel-shaped particles thatcan easily pass over one another, to explain why water is liquid. We thenderive new observable phenomena from our conjectured structure. Ifthe consequences so derived are actually observed, then the conjectureis proved, or, at least, made credible.

So far so good. But while Descartes seems reasonably clear about thenew way of finding knowledge of these substructures, he is not so clearabout the status of such knowledge. How certain can we be of particu-lar natures discovered and “proved” in this way? Here there seems tobe at least some ambivalence in Descartes’ view.

At one extreme, Descartes seems on at least one occasion to suggestthat this procedure gives us genuinely certain knowledge of the innerstructure of things. He writes:

It could scarcely happen that a cause from which all phenomena can clearly be deducedmight be false. . . . We would seem to be doing God an injustice if we suspectedthat the causes of things discovered in this way were false, as if He had givenus such an imperfect nature that we could be deceived by reason, even whenwe were using it properly.24


22 Ibid. IV 203. 23 Ibid. IV 201.24 Ibid. III 43. This is a very strong reading of this passage. The words omitted in the quo-

tation in the text are as follows: “Suppose, then, that we use only principles which we seeto be utterly evident, and that all our subsequent deductions follow by mathematical rea-soning.” . . . This suggests that Descartes may have intended to make the somewhat weakerpoint that if we begin with intuition, and proceed by deduction, then we are entitled tocertainty.

Sometimes, though, Descartes suggests more modestly that while wemay lack the absolute certainty we have in God, mathematics, and thedistinction between mind and body, we still have what he calls moralcertainty about the conjectured corpuscular substructure of particularthings. He writes:

It would be disingenuous, however, not to point out that some things are con-sidered as morally certain, that is, as having sufficient certainty for applica-tion to ordinary life, even though they may be uncertain in relation to theabsolute power of God.25

Descartes then goes on to compare our situation with respect to thehidden natures of particular things with the person trying to decode a letter. If by replacing some letters with others in an orderly way,someone can turn the text into one that makes sense,

he will be in no doubt that the true meaning of the letter is contained inthese words. It is true that his knowledge is based merely on a conjecture,and it is conceivable that the writer [intended] a different message; but thispossibility is so unlikely . . . that it doesn’t seem credible.26

What Descartes is talking about here is not just the corpuscular sub-structures of particular things, but also the general principles on whichthey are based. But his remarks here would seem to hold true for theparticular substructures that Descartes is positing in the Principia.

Elsewhere still in the Principia Descartes takes a step beyond evenmoral certainty when he suggests that the particular natures he positsmay well be false, however useful they may be as a guide to life. Hewrites:

With regard to the things which cannot be perceived by the senses, it is enough to explaintheir possible nature, even though their actual nature may be different. However,although we can understand how all the things in nature could have arisenin this way, it should not therefore be inferred that they were in fact madein this way. Just as the same craftsman could make two clocks which tell thetime equally well and look completely alike from the outside but have com-pletely different assemblies of wheels inside, so the supreme craftsman of thereal world could have produced all that we see in several different ways. . . .I shall think that I have achieved enough provided only that what I havewritten is such as to correspond accurately with all the phenomena of nature.[French version: We shall achieve our aim irrespective of whether these


25 Ibid. IV 205. 26 Ibid.

imagined causes are true or false, since the result is taken to be no different,as far as the observable effects are concerned.] This will indeed be sufficientfor application to ordinary life, since medicine and mechanics, and all theother arts which can be fully developed with the help of physics are directedonly toward items that can be perceived by the senses and are therefore tobe counted among the phenomena of nature.27

Similarly, Descartes writes that he wants the causes he sets out to bemere hypotheses, whose truth is irrelevant. Even if one of his hypothe-ses is taken to be false, he writes,

I think that I shall have achieved something sufficiently worthwhile if every-thing deduced from it agrees with our observations; for if this is so, we shallsee that our hypothesis yields just as much practical benefit for our lives aswe would have derived from knowledge of the actual truth.28

Indeed, he then goes on to make a hypothesis that he knows to befalse, that God created the world not as is set out in Genesis, but in aninitial state of chaos, from which everything we see around us followsout by way of the laws of nature alone.29 This is not the first time thatDescartes has said such things. In the beginning of the Dioptrique, whensetting out the “suppositions” that he will use in the course of that work,Descartes notes:

In this I am imitating the astronomers, whose assumptions are almost all falseor uncertain, but who nevertheless draw many very true and certain conse-quences from them because they are related to various observations they havemade.30

Like the astronomers, he argues, what counts is the ability to predictnew phenomena, and not the truth of the hypothesis. In 1637 this wasonly to justify using hypothetical argument as a provisional mode ofexposition, a way of presenting the results of Descartes’ physics withouthaving to present its foundations. But in 1644 Descartes seems to besaying that this is all we can expect from his physics as far as knowledgeof particulars is concerned, conjectures that we have every reason tobelieve may well be false, but which are useful in the prediction of phenomena. What is extremely interesting here is Descartes’ rather cavalier attitude toward truth when it comes to the corpuscular sub-


27 Ibid. IV 204, Latin version with French variant inserted.28 Ibid. III 44. 29 See ibid., III 45.30 AT VI 83; for other discussions of the use of false or questionable hypotheses among the

astronomers, see AT VII 349; AT X 417; AT II 198–99.

structures and particular natures that he is discussing: It simply does notmatter if the conjectures are false, as long as they agree with the phe-nomena of experiment and observation. What is important for Descartesis now simply that the consequences of his conjectured particular naturesagree with experience. For if they do, then whether true or false, theycan be used to predict future experience, and in that way serve as reli-able guides to life. In this way we can say that for Descartes, experiencedoesn’t confirm the truth of conjectures about the corpuscular sub-structure, but their reliability as predictors of future experience.

The claim that the beliefs about corpuscular substructure arrived atthrough hypothetical argument are only morally certain, or even falseis highly significant. In his very early Regulae, Descartes was quite certainthat anything less than absolutely certain knowledge was simply notworth having. In Rule 1 he declares that “the aim of our studies shouldbe to direct the mind with a view to forming true and sound judgmentsabout whatever comes before it.” Rule 2 is even more demanding: “Weshould attend only to those objects of which our minds seem capableof having certain and indubitable cognition.”31 This strict view of what we as investigators are seeking is, I think, reflected in the officialview of the Discours and Essais. Descartes’ insistence that he is usinghypothetical argument only as a convenient way of presenting his viewwithout divulging its full foundations, and his at least implicit claim thathis conclusions really come from intuition and deduction is, I think,connected with this earlier commitment to certainty; Descartes, I think,was under the illusion that his appeal to intuition yielded certain knowl-edge of the inner nature of things. But Descartes is clear that whateverits virtues, hypothetical argument does not yield certainty, and inendorsing it in the Principia, he is clearly changing his view in a verysignificant way.

Indeed, in treating our knowledge of the inner structure of thingsin the way in which he does, as arguing that it provides us not with cer-tainty but only with a guide to life, Descartes is placing our knowledgeof the inner nature of things at an epistemological level no higher thanthat of sensation itself, and perhaps even lower. In the Meditations,Descartes characterizes sensation as follows:

The proper purpose of the sensory perceptions given me by nature is simplyto inform the mind of what is beneficial or harmful for the composite of


31 AT X 359, 362.

which the mind is a part; and to this extent they are sufficiently clear and dis-tinct. But I misuse them by treating them as reliable touchstones for imme-diate judgments about the essential nature of the bodies located outside us;yet this is an area where they provide only very obscure information.32

This, in a way, is the most that we can hope for from our hypothesesabout the inner nature of particular things as well, that they will provideus predictions about what to expect in the world, and in that way, helpus to survive in this uncertain world; this is just what it means for a beliefto have moral certainty. But perhaps even this is too much to expect.Descartes infers from his account of sensation that since it is given tous as a guide of life, we know that it is, in a sense, trustworthy: “I knowthat in matters regarding the well-being of the body, all my senses reportthe truth much more frequently than not.”33 But we can’t even say this about our conjectures about hidden natures; for all we know theymay be genuinely false. Nor does it really matter to us, as long as the phenomena they entail constitute a reliable guide to life. Regarded in this way, the hidden mechanism, the corpuscular substructure, the real nature of a body has become a mere calculating device for pre-dicting future phenomena, and lost the status of even being a candi-date for knowledge or ignorance; all that really seems to count are thephenomena.

The progression here is very significant. We began in the earlyRegulae and the later Discours with certain knowledge, and progressedin the Principia to mere moral certainty and genuine ignorance; from the certainty of intuition to the lesser grade of certainty associ-ated with the senses, good enough for guiding life, but not for findingtruth. But, at the same time, we also passed from a certain knowledgeof hidden natures, obtained through a hopelessly obscure cognitiveprocess (intuition), to a clear and manifest cognitive process (analogyand hypothetical argument) that claims to give us not truth, but onlyutility.

Why the change? In defending his use of hypothetical argument, itis important to remember that Descartes can’t appeal to the need tohide his views any more; he has made the decision to go public withthe foundations of his physics, and cannot use the need to hide themas a justification for his use of this apparently inferior way of arguing.One can only suppose that Descartes adopts hypothetical argument as


32 AT VII 83. 33 AT VII 89.

a way of finding particular natures because he genuinely believes thatthis is the best that he can do. Perhaps he came to appreciate the obscu-rity of the appeal to intuition, his earlier conception of how corpuscu-lar substructures are to be found, and saw in hypothetical argument theclarity he sought, even if it meant sacrificing certainty. Perhaps in actu-ally working out and defending his views on the inner nature of things,he came to appreciate the sheer complexity of nature, and saw in hypo-thetical argument a better way of coming to grips with the world. Butfor whatever reason, Descartes was led to give up his earlier extravagantclaims about what we can know and how, in favor of the relatively moremodest claims in the Principia.


PART III

MIND, BODY, AND THE LAWS OF NATURE

7

MIND, BODY, AND THE LAWS OF NATUREIN DESCARTES AND LEIBNIZ

One of the central doctrines of Descartes’ metaphysics was his divisionof the created world into two kinds of stuff: mental substance whoseessence is thought and material substance whose essence is extension.And one of the central problems that later philosophers had withDescartes’ doctrine was understanding how these two domains, themental and the material, relate to one another. Descartes’ solution wasto claim that these two domains can causally interact with one another,

133

ABBREVIATIONS

Books and Collections

C Couturat, L., ed., Leibniz: Opuscules et Fragments Inédits (Paris: 1903).G Gerhardt, C. I., ed., Leibniz: Philosophischen Schriften (Berlin: 1875–1890).GM Gerhardt, C. I., ed., Leibniz: Mathematische Schriften (Berlin: 1849–1855).L Loemker, L., ed. and trans. Gottfried Wilhelm Leibniz: Philosphical Papers and

Letters, 2nd ed. (Dordrecht: 1969).M Mason, H. T., ed. and trans. The Leibniz–Arnauld Correspondence (Manchester:

1967).NE Langley, A. G., ed. and trans. Leibniz: New Essays Concerning Human Understand-

ing (La Salle, IL, 1949).

Individual Works

DM Leibniz, Discours de métaphysique. Found in G IV 427–63 and translated in L303–28.

Mon. Leibniz, Monodologie. Found in G VI 607–23 and translated in L 643–52.PA Descartes, Les passions de l’âme. Found in AT XI 291–497.Theod. Leibniz, Essais de theodicée. Found in G VI 21–471.

References to books and collections are given by volume (when appropriate) and page. References to individual works are given by part (in the case of Pr) and section number. Original language citations are given first, followed by an English translation in parentheseswhen available.

that bodily states can cause ideas, and that volitions can cause bodilystates. But this claim raises a number of serious questions. The mostobvious problem arises from the radical distinction that Descartes drawsbetween the two domains and from our difficulty in conceiving how twosorts of things so different could ever interact with one another. As the Princess Elisabeth complained to Descartes, “it is easier for me to concede matter and extension to the mind than [it is for me toconcede] the capacity to move a body and to be affected by it to animmaterial thing.”1 Though the story is complex, it is generally heldthat this problem led later in the century to the doctrine of occasion-alism, in which the causal link between mind and body was held to be not a real efficient cause but an occasional cause. Thus, it was claimed, it is God who causes ideas in minds on the occasion of appropriate events in the material world and events in the materialworld on the occasion of an appropriate act of will.2 The causal linkbetween mind and body remains but is reinterpreted as an occasionalcausal link, a causal link mediated by God. But Descartes’ interaction-ism raises another problem as well. For the seventeenth century, thematerial world was thought to be governed by a network of physicallaws. But, it would seem, if the material world is governed by law, thenthere can be no room for minds to act; if mind can be either the effi-cient or the occasional cause of changes in the material world, then, itwould seem, physical laws must fail to hold in any system that containsanimate bodies, bodies under the influence of minds.3 Particularly vul-nerable to such violations are the conservation laws, laws that stipulatethat certain physical quantities must remain constant over time, sinceit is difficult to see how a mind could influence the course of the mate-rial world, either by itself or with the intermediation of God, withoutaltering some physical magnitude. Leibniz seizes upon just this featureof Descartes’ position in an argument intended to persuade us to reject

134 mind, body, and the laws of nature

1 AT III 685.2 The most prominent adherent of this position is, of course, Nicolas Malebranche. See his

The Search after Truth and Elucidations of the Search after Truth, ed. and trans. by ThomasLennon and Paul J. Olscamp (Columbus, Ohio: Ohio State University Press, 1980), pp.446–52 and 657–85; or Dialogues on Metaphysics, ed. and trans. by Willis Doney (New York:Abaris Books, 1980), pp. 144–69.

3 In this essay, the term “animate body” will be used to designate any body related in anappropriate way to a mind or soul, as, for example, the human body is for both Descartesand Leibniz. This has the unfortunate consequence that on my somewhat special use ofthe term Cartesian animals must be considered inanimate. But I could find no more naturalway of designating the special class of bodies with which I will be concerned in this essay.

interactionism and accept his doctrine of pre-established harmony.Leibniz argues:

M. Descartes wanted . . . to make a part of the action of the body depend onthe mind. He thought he knew a rule of nature which, according to him,holds that the same quantity of motion is conserved in bodies. He did notjudge it possible that the influence of the mind could violate this law ofbodies, but he believed, however, that the mind could have the power tochange the direction of the motions which are in bodies. . . . [But] two impor-tant truths on this subject have been discovered since M. Descartes. The firstis that the quantity of absolute force which, indeed, is conserved, is differentfrom the quantity of motion, as I have demonstrated elsewhere. The seconddiscovery is that the same direction is conserved among all of those bodiestaken together which one supposes to act on one another, however they maycollide. If this rule had been known to M. Descartes, he would have renderedthe direction of bodies as independent of the mind as their force. And I believe that this would have led him directly to the hypothesis of pre-established harmony, where these rules led me. Since beside the fact that thephysical influence of one of these substances on the other is inexplic-able, I considered that the mind cannot act physically on the body withoutcompletely disordering the laws of nature.4

Leibniz’s argument is elegant and straightforward. The claim is thateven though Descartes thought that he could reconcile the causal inter-action of mind and body with the universality of physical law, he wasmistaken. The true laws of nature block Descartes’ solution, Leibnizargues, and lead us away from causal interactionism and directly to thehypothesis of pre-established harmony as the true account of the appar-ent relations that hold between the mental and the material.

In this paper, I shall explore this argument of Leibniz’s in somedetail. I shall begin with a careful exposition of the argument, sketch-ing in some of the details of his position and Descartes’ that Leibnizleaves out. I shall then examine the extent to which the position Leibnizattacks is the position that Descartes actually held and argue thatDescartes’ actual position allows him a plausible answer to Leibniz’sattack on interactionism. In the end, I shall argue that the oppositionbetween Cartesian interactionism and Leibnizian harmony is only asymptom of a much deeper difference, a difference between two

descartes and leibniz 135

4 Theod. 60–61. See also Mon. 80; G II 94 (M 117–18); G III 607 (L 655); G IV 497–98; G VI 540 (L 587). The argument in these passages concerns only the metal causation ofphysical events. Consequently, I will not discuss the problems raised by the physical causa-tion of mental events.

opposing conceptions of the laws of nature and of the place of mindin the physical world.

1. Motion, Momentum, and Pre-Established Harmony

Cartesian physics is a physics of geometrical bodies, bodies all of whoseproperties are modes of extension, acting on one another throughdirect impact. Basic to such a physics, of course, are the laws of motionand impact, the laws that govern the only kinds of change allowed inthe world of material things. And basic to the laws of motion and impactfor Descartes is his conservation law, derived directly from the activityof God. As Descartes wrote in his Principia:

God . . . in the beginning created matter along with motion and rest, and now,through His ordinary concourse alone, conserves just as much motion andrest in the whole of it [i.e., the material world] as He put there at that time.For although that motion is only a mode of moving matter, it has a certaindeterminate quantity which can easily be understood to remain always thesame in the totality of things, even though it is changed in the individual parts.And so, for example, we believe that when one part of matter moves twice asfast as another, and the latter is twice as big as the former, there is as muchmotion in the smaller as in the larger; and as much motion as is lost by onepart slowing down is gained by another of equal size moving more quickly.5

Descartes’ example suggests that his conservation principle can be sum-marized by a simple quantitative law: The total quantity of motion, as mea-sured by the mass of each body multiplied by its speed, remains constantfor the whole of the material world.6

It is tempting, but wrong, to assimilate Descartes’ conservation law tothe modern principle of the conservation of momentum. The conser-vation of momentum, a law that entered classical physics only later in theseventeenth century, holds that the total quantity of momentum remainsconstant, where momentum is understood as mass times velocity and


5 Pr II 36. The conservation law is first stated in the ill-fated Le Monde. See AT XI 43.6 This is the standard reading of Descartes’ law. It should be noted that my use of the term

“mass” here is anachronistic. Although it helps one to see the relations between Descartes’incorrect law and later conservation principles, such as Leibniz’s, Descartes himself wouldhave given his law in terms of “size” rather than “mass.” For a discussion of some of thefurther intricacies in interpreting Descartes’ conservation law, see Pierre Costabel, “Essaicritique sur quelques concepts de la mécanique cartésienne.” Archives internationales d’his-toire des sciences, Vol. 20 (1967), pp. 235–52, esp. pp. 240–51. None of these questions ofinterpretation are relevant to the use Leibniz makes of Descartes’ conservation law in theargument under discussion, though.

where velocity is understood as a vector quantity, speed and its direction.Thus, the law of the conservation of momentum governs both the speedand the directions that bodies have. So, for example, if a body movingfrom right to left were to reverse its direction (because of a collision withanother body, say), then the conservation of momentum would requirethat some other body or bodies (say, the body that had been hit) wouldhave to begin moving at an appropriate speed from left to right in orderto preserve the total momentum in the world.

Descartes’ conservation law is quite a different matter, though. Basicto Descartes’ physics is a strict distinction between the motion or quan-tity of motion a body has, and its determination as he calls it, roughly speak-ing, the direction in which that body is moving.7 Now, even though thisdistinction between (quantity of ) motion and determination does notexplicitly appear in any statement of Descartes’ conservation law, it isclear both from the lack of any mention of determination in that law andfrom the way Descartes actually applies the conservation law that it ismeant to govern the motion alone. Thus, for example, when discussingimpact, Descartes quite carefully separates out the two factors in the phys-ical situation, using the conservation law only to determine the postcol-lision speeds of the bodies in question.8 So, if in a system of bodies onebody changes its direction, then, as long as it maintains its original speed,there is no change in the total quantity of motion; no compensatorychange in the direction of another body is required to satisfy Descartes’law, as is the case with the conservation of momentum.9 In holding that


7 The distinction is most clearly drawn in Pr II 41. Once again, this is the standard reading.Though it is sufficient for our purposes here, Descartes’ notion of determination is muchmore complex than the simple equation of determination and direction would suggest. Onthis, see Pierre Costabel, op. cit., 236–40; J. Ohana, “Note sur la théorie cartésienne de ladirection du mouvement,” Les Etudes philosophiques, Vol. 16 (1961), pp. 313–16; OleKnudsen and Kurt Pedersen, “The Link between ‘Determination’ and Conservation ofMotion in Descartes’ Dynamics,” Centaurus, Vol. 13 (1968–1969), pp. 183–86; A. I. Sabra,Theories of Light from Descartes to Newton (London: Oldbourne Press, 1967), pp. 116–27; andAlan Gabbey, “Force and Inertia in the Seventeenth Century: Descartes and Newton,” inStephen Gaukroger, ed., Descartes: Philosophy, Mathematics and Physics (Sussex: HarvesterPress, 1980), pp. 230–320, esp. pp. 248–61.

8 See, e.g., Pr II 41; AT IV 185–86; AT VI 94, 97.9 This is exactly the situation envisioned in Descartes’ infamous fourth rule of impact, given

in Pr II 49. According to that rule, if C is larger than B and if C, at rest, is hit by B, thenB will reverse its direction and rebound from the collision with exactly the speed with whichit originally approached C. Strictly speaking, though, even this very simple case wouldrequire innumerable changes in the speeds and directions of other bodies in the system,since the Cartesian world is a plenum.

the conservation law does not govern the directions in which bodiesmove, Descartes is not saying that direction is completely arbitrary. Both(quantity of ) motion and direction are modes of body, and, as such,neither will change without an appropriate cause.10 The point is just thatwhatever causes might result in changes in direction, such changes indirection are, by themselves, irrelevant to the law of the conservation ofmotion. One can alter the directions in which bodies in the world moveas much as one like, and as long as the speeds remain unchanged, thetotal quantity of motion will remain unaltered.

This feature of Descartes’ conservation law opens an obvious possi-bility with respect to his account of mind-body interaction. Descartesclearly held that minds can cause events in the physical world. And it is also at least initially plausible to suppose, as Leibniz did, thatDescartes wanted such interaction to take place without violating hisconservation law. These two commitments can be easily reconciled,given the particular conservation law that Descartes adopted. If wesuppose that mind acts on body by changing the direction in which somepiece of matter is moving without changing its speed, then the problemis solved: mind can act on body without violating the conservation law.Mind can thus fit into the gap left open in Descartes’ conservation lawand help to determine what that law makes no pretense of governing.We will have to examine the textual evidence there is for attributingthis line of reasoning to Descartes. But it is a position that he could have taken, and it is clearly the position that Leibniz thought that hedid take.

However, it is just as clear that this is a position that Leibniz does notthink Descartes is entitled to take. As the passage quoted above suggests,Leibniz’s argument depends crucially on his refutation of Descartes’conservation law and its replacement by two somewhat different con-servation principles. The arguments are complex, and a full examina-tion of them would take us far beyond the scope of this paper. Putbriefly, though, Leibniz was able to show that Descartes’ conservationlaw has the absurd consequence that if it were the only law that bodiesin motion were constrained to observe, then it would be possible tobuild a perpetual motion machine. More generally, he showed that inbody-body interactions (collisions, for example) governed only by theprinciple of the conservation of quantity of motion, it is possible for


10 See Pr II 41; AT III 75; AT IV 185; AT VI 94, 97.

the system to either gain or lose the ability to do work (the ability toraise a body of a given weight a given height, for example). This situa-tion violates the principle of the equality of cause and effect, a meta-physical principle that, Leibniz held, governs this best of all possibleworlds. According to that principle:

The entire effect is equal to the full cause, and therefore, there is no mechan-ical perpetual motion, nor can a cause produce an active effect which can domore than the cause itself, but neither can there be an entire effect that cando less than the cause itself.11

Leibniz argues that if the equality of cause and effect is to be main-tained, we must conserve not quantity of motion, mass times speed, buta different physical magnitude, living force (vis viva), which, he argues,is measured by mass times the square of the speed.12 This new law is animprovement over Descartes’ to be sure. But by itself it does not seemto constrain directionality any more than Descartes’ conservation lawdid. In a system of bodies, each of which is governed only by the con-servation of living force, it seems as if one could change the directionsof the bodies without changing the living force in the system. However,from this basic conservation law Leibniz is able to derive a second con-servation law, a new law that constrains directionality in a way thatDescartes’ law does not.

Consider an aggregate of bodies in motion that constitutes a closedsystem, i.e., one in which no force is being added from the outside. Thissystem contains living force in two different respects. First of all, eachbody in the aggregate has its own force, as measured by the mass ofeach body times the square of its speed. The sum of all these individ-ual forces is what Leibniz calls the “respective or proper force” of theaggregate. But in addition, the aggregate has what Leibniz calls “direc-tive or common force . . . , that by which the aggregate can itself act


11 GM VI 437. See also G III 45–46.12 This argument is implicit in the critique of Descartes’ conservation law given in Leibniz’s

important “Brief Demonstration of a Notable Error in Descartes . . .” of 1686, the first ofLeibniz’s mature publications in physics. The next of this is given with a later appendixin GM VI 117–23 (L 296–302). For other presentations of the same basic argument, see,e.g., DM 17, G IV 370–72 (L 393–95); GM VI 243–46 (L 442–44); GM VI 287–92; etc.For discussions of the argument, see, e.g., Carolyn Iltis, “Leibniz and the Visa Viva Con-troversy,” Isis, Vol. 62 (1971), pp. 21–35; George Gale, “Leibniz’ Dynamical Metaphysicsand the Origin of the Vis Viva Controversy,” Systematics, Vol. 11 (1973), pp. 184–207; and Martial Gueroult, Leibniz: Dynamique et Métaphysique (Paris: Aubier-Montaigne, 1967),pp. 28–34.

externally.”13 This force is the force that the aggregate considered as awhole has, and it is measured by the total mass of the aggregate timesthe square of the speed of the center of mass of the aggregate. Now, justas the force in each individual body remains unchanged if nothingexternal affects it, so should the directive force of the aggregate remainunchanged if no force is added. But, Leibniz shows, this entails thatwithin the aggregate any change in the direction of one body (througha collision with other bodies in the aggregate, say) must be compen-sated for by a change in the direction of some other body or bodies in the aggregate (say, the body or bodies hit), or else the speed of thecenter of mass of the aggregate as a whole will change, changing thedirective force of the aggregate. Using reasoning like this, Leibniz estab-lishes that if the total force of an aggregate is to be conserved, then notonly must the respective force be conserved, the mass times the squareof the speed of each individual body in the aggregate, but also the totalquantity of momentum, mass times velocity, speed and direction! And sincethe universe as a whole constitutes such an aggregate, the conservationof momentum must govern the universe as a whole.14 Thus, Leibnizargues, the principle of equality of cause and effect governs not onlythe speeds bodies have but their directions as well; a change in either thespeed or the direction of a given body not compensated for by appropriate changes in other bodies is not permitted in the best of allpossible worlds.15


13 For the distinction between these two kinds of force, see GM VI 238–39 (L 439); GM VI462; GM VI 495.

14 The theorem is stated in numerous places. See, e.g., Theod. 61; G II 94 (M 117–18); GIV 497–98; GM VI 216–17 (NE 658); GM VI 227 (NE 667). A detailed argument is givenin the Dynamica, GM VI 496–500. The crucial lemmas are given on GM VI 440, whereLeibniz argues that “the same power [potentia] remains in any system of bodies not com-municating with others” and concludes that, since the universe is such a system, “the samepower always remains in the universe.” This kind of argument is somewhat problematicfor Leibniz when applied to momentum, since it is difficult to see what sense he couldmake of the speed of the center of mass of the universe as a whole. It should be notedthat “momentum” is not Leibniz’s term for the quantity at issue. Leibniz uses a numberof terms, sometimes “quantity of nisus” (GM VI 462), sometimes (quantity of ) “progress”(GM VI 216–17 [NE 658]; GM VI 227 [NE 667]) but most often “direction,” “total direc-tion,” or the like (Theod. 61; Mon. 80; G II 94 [M 117–18]; G III 607 [L 655]; G VI 540[L 587]; G IV 497; etc.).

15 It seems as if this general kind of argument could have been used directly against Descartes’conservation law to show that it, too, ought to govern directionality and not just speed.Thus, Leibniz’s replacement of quantity of motion by vis viva as the physical magnitudeconserved is not, strictly speaking, relevant to the argument against interactionism.

This argument quite effectively blocks the reasoning that Leibnizattributed to Descartes. There is no room in Leibniz’s conception ofthe material world for Cartesian minds to act. Cartesian interactionismis impossible without a violation of what were for Leibniz the basic meta-physical and physical laws that govern our world. This, Leibniz claims,led him and would have led Descartes, if he had grasped the true lawsof nature, to reject interactionism and adopt the hypothesis of pre-established harmony. The hypothesis of pre-established harmony is, ofcourse, one of Leibniz’s proudest inventions. In its strictest formula-tion, it posits a perfect correspondence among the perceptions of allmonads. As such, it is intimately connected with Leibniz’s conceptionof the world as a collection of monads that are, by their nature, inca-pable of any genuine causal interaction.16 But Leibniz also formulatesthe doctrine of pre-established harmony in a somewhat different way,a way that can be understood, argued for, and adopted independentlyof Leibniz’s idiosyncratic views about the ultimate nature of the worldand the ultimate reduction of material bodies to well-founded phe-nomena grounded in a world of monads. In this version, the doctrineof pre-established harmony is less a claim about the interrelationsamong all created substances than it is a claim about two very specialones, the human mind and the human body. In its less rigorous formulation, the doctrine states simply that events in the mind andthose in the body correspond to one another not because of anygenuine causal link between the two, as Descartes held, and not becauseof the intervening action of God, as the occasionalists would have it, but because God, in the beginning, created mind and body inde-pendent of one another in such a way that there would always be anappropriate correspondence between what was going on in the one andwhat was going on in the other. As Leibniz succinctly summarized histheory:

If we posit the distinction between mind and body, their union can beexplained without the common hypothesis of influence, which cannot beunderstood, and without the hypothesis of occasional causes, which summonsa deus ex machina. For GOD from the beginning so constituted both the mind and the body at the same time, with such wisdom and such skill thatfrom the first constitution and essence of each, everything that comes about


16 This conception of the doctrine of pre-established harmony is found in G I 382–83; G II68–70 (M 84–86); G IV 518 (L 493); G VII 412 (L 711–12).

through itself in the one corresponds perfectly to everything that happensin the other, just as if [something] passed from the one into the other.17

This hypothesis, of course, deals neatly with the problem that hadworried so many about how things as different as minds and bodiescould be causally connected with one another. On Leibniz’s theory theyaren’t. But, in this respect, Leibniz’s theory is at best a small improve-ment over occasionalism, substituting one large divine labor in creat-ing mind and body in harmony with one another for numerous lesserdivine actions in coordinating the moment-by-moment states of the two.The deeper differences between pre-established harmony and occa-sionalist interactionism become clearer when we examine the problemsraised by physical law. Although occasionalism addresses the problemof the mechanism of interaction, there is nothing in the occasionalistposition that bears on the problem of interactionist violations of phys-ical law. For the occasionalist, just as for the direct interactionist, everyvoluntary action would seem to violate some law of nature. Not so forLeibniz’s pre-established harmony. If God can create a world in whichevents in minds and bodies can correspond with one another in anappropriate way without the necessity for either real or occasionalcausal links, He can also create things in such a way that this corre-spondence can take place without violating any of the laws that holduniversally in the physical realm. Thus, Leibniz wrote:

Minds follow their laws, which consist in a certain development of percep-tions in accordance with goods and evils, and bodies also follow theirs, whichconsist in the laws of motion. But these two things entirely different in kindjoin together and correspond like two time-pieces perfectly well regulated tothe same time, even though perhaps of entirely different construction.18

Or, even more graphically, Leibniz wrote to Arnauld:

It is thus infinitely more reasonable and more worthy of God to suppose thatHe created the machine of the world from the beginning in such a way thatwithout violating at any moment the two great laws of nature, those of forceand direction, and instead in following them perfectly (excepting the case ofmiracles), it happens that the springs of bodies are ready to act of themselves,as is necessary, just at the moment that the soul has a volition, . . . and thus


17 C 521 (L 269). For other statements of this version of pre-established harmony, see, e.g., DM 33; G II 57–58 (M 64–65); G II 112–14 (M 144–46); G IV 483–85 (L 457–58);G IV 498–500 (L 459–60); G IV 520 (L 494); G VII 410–11 (L 710–11); etc.

18 G VI 541 (L 587).

that the union of the mind with the machine of the body and the parts whichit contains and the action of one on the other consist only in that concomi-tance which marks the admirable wisdom of the creator much better thandoes any other hypothesis.19

Given this particular statement of the doctrine, it is clear why Leibniz’sreflections on mind-body interaction and physical law might have ledhim to pre-established harmony. Pre-established harmony seems to bean attractive way in which a dualist could account for the posited cor-respondence between acts of will in a nonmaterial mental substanceand appropriate events in a nonmental body without violating any ofthe laws of nature that, Leibniz held, govern every event in the mater-ial world.

2. Interaction and Conservation in Descartes

Leibniz’s argument is an elegant one, a paradigmatic example of theinterconnection between physics and metaphysics that characterizesrationalist science. And Leibniz seems to have focused on one of the central questions raised by any dualist interactionist philosophy ofmind. Now, as a purely philosophical argument. Leibniz’s attack onDescartes is worthy of serious consideration, to be sure.20 But what inter-ests me here is a somewhat more historical question: Is the position thatDescartes actually held open to this kind of attack?

There is no question but that Descartes held the conservation law towhich Leibniz alludes in his statement of the argument, and there is noquestion but that Descartes’ law is wrong and the laws that Leibniz sub-stitutes for it correct, at least within the world of classical physics. ButLeibniz’s attack on Cartesian interactionism makes at least one furtherassumption, the assumption that the laws of nature must, miracles aside,hold universally, without exception for all bodies in the material world,including animate bodies like our own. Leibniz certainly believed in the


19 G II 94–95 (M 118). See also Mon. 78; Theod. 62; G II 71 (M 87); G II 74 (M 92); G II205–6; G IV 484 (L 458); G IV 559–60 (L 577–78); G V 455 (NE 553); G VI 599 (L 637);G VII 412 (L 712); G VII 419 (L 716–17). These passages make it evident just how deeplyLeibniz was influenced by the materialism of Hobbes and the dual aspect theory ofSpinoza. In these passages, Leibniz emphasizes that every event in the material world hasan explanation in terms of the laws of physics alone.

20 For the classic examination of this objection to dualist interactionism from a purely philo-sophical point of view, see C. D. Broad, Mind and Its Place in Nature (London: K. Paul,Trench, Trubner and Co., 1925), pp. 103–9.

universality of natural law in this sense and attributed the same beliefto Descartes, claiming that this commitment forced Descartes to holdthat minds can change only the directions in which bodies move andnot their speeds. But curiously enough, even though Leibniz was wellversed in the Cartesian corpus, he refers to no passages from Descartes’writings to support those attributions. Nor could he have. For a closeexamination of Descartes’ writings gives us good reason to believe thathe never held the positions that Leibniz attributed to him, either thechange-of-direction account of mind-body interaction or the universal-ity of the laws of motion.21

Let us begin with the change-of-direction account of mind-bodyinteraction. The most striking evidence against the claim that Descartesheld such a position is the simple fact that nowhere in what currentlysurvives of Descartes’ writings do we find anything like a clear statementof the account that Leibniz attributed to him; nowhere did he ever saythat he held that minds can only change the direction in which bodiesmove. Typically when presenting his position he is content to assertsimply that mind can cause motion in bodies. For example, Descarteswrote the following passage in a letter to the Princess Elisabeth in thecontext of an explanation of the primitive notion we have of the unionof mind and body:

As regards mind and body together, we have only the notion of their union,on which depends our notion of the mind’s power to move the body [la forcequ’a l’ame de mouuoir le corps], and the body’s power to act on the mind andcause sensations and passions.22

Similarly, Descartes wrote to Arnauld:

Moreover, that the mind, which is incorporeal, can set a body in motion[corpus possit impellere] is shown to us every day by the most certain and most


21 Although not generally recognized, this feature of Cartesian thought has been pointedout from time to time, only to be forgotten and then rediscovered by successive genera-tions of scholars. On this, see Octave Hamelin, Le Système de Descartes (Paris: Librairie FélixAlcan, 1911), pp. 372–73; Jean Laporte, Le Rationalisme de Descartes (Paris: Presses Uni-versitaires de France, 1950), pp. 245–48; Norman Kemp Smith, Studies in the CartesianPhilosophy (London: Macmillan, 1902), p. 83 n.2; Geneviève Rodis-Lewis (ed.), Descartes:Passions de l’Ame (Paris: J. Vrin, 1970), p. 92 n.1. The most recent rediscovery is in PeterRemnant, “Descartes: Body and Soul,” Canadian Journal of Philosphy, Vol. 9 (1979), pp.377–86. Needless to say, there is substantial overlap between my argument in this sectionand the arguments presented in the other commentaries cited. However, the continuedunfamiliarity of this point plus the new bits of evidence I have found make it worthwhileto review the case for this interpretation once again.

22 AT III 665.

evident experience, without the need of any reasoning or comparison withanything else.23

And finally, consider a passage that Descartes wrote to Henry More:

The force moving [a body] [vis . . . mouens] can be that of God Himself . . .or also that of a created substance, like our mind, or that of some other thing to which He gave the force of moving a body [cui vim dederit corpusmouendi].24

There is no mention of directionality in these passages. Descartes iscontent to say only that our minds have the ability to move our bodies.But these remarks are, admittedly, casual and were given in the contextof nontechnical and almost off-the-cuff explanations of his position.However, it is significant that this casual lack of attention to the ques-tion of change of speed versus change of direction is also found in thestrict and more technical accounts of mind-body interaction thatDescartes gave.

Consider, for example, the discussion of interaction that Descartesgives in the Passions de l’âme, a sort of auto mechanic’s manual for themind-body union, where Descartes outlines in rather specific ways thenuts and bolts of how the mind acts on the part of the body to whichit is most directly connected, the pineal gland.25 Some of Descartes’most careful discussions of the direct action of the mind on the pinealgland there do indeed suggest that at least sometimes the mind acts onthe human body by changing the direction in which the pineal glandis moving. Thus, Descartes writes in the Passions that “when the mindwants to remember something, this volition makes the gland incline


23 AT V 222.24 AT V 403–4. This passage will be discussed in greater detail below.25 On the direct connection between the mind and the pineal gland, see, e.g., PA 31; AT

VII 86, AT XI 176–77, 183. It should also be noted that, in addition to the direct con-nection between mind and body, Descartes also holds that by virtue of being directly con-nected to the pineal gland the mind is indirectly connected to the human body as a whole.See, e.g., PA 30. Margaret Wilson sees these as two opposing conceptions of mind-bodyunity. See her Descartes (London: Routledge and Kegan Paul, 1978), pp. 204–20. I see thetwo conceptions as perfectly consistent and, in fact, complementary, as their juxtapositionin PA 30–31 suggests. Though I quote exclusively from the PA in discussing the action ofthe mind on the pineal gland, Descartes also discusses this question in the earlier Traitéde l’Homme. But the discussions there are much less useful for our purposes. Most of thediscussions that deal with the pineal gland deal with its role in sensation. See, e.g., AT XI143–46, 176–77, 181, 183. And when volition is discussed in l’Homme, Descartes givesalmost no detail as to how mind actually manipulates the pineal gland. See, e.g., AT XI131–32, 179.

successively in different directions [vers divers costez].”26 Similarly, intalking about the opposition between the mind and the animal spirits,a bodily substance also capable of moving the pineal gland and, in sodoing, causing both passions and involuntary movement of the body,Descartes notes that the pineal gland “can be pushed in one direction[poussée d’un costé ] by the mind, and in another by the animal spirits.”27

But there is nothing to suggest that the only way that the mind acts inthe pineal gland is by changing the direction of its motion. In the Pas-sions, Descartes often says simply that the pineal gland “can be movedin different manners by the mind [diversement meuë par l’ame]” or thata volition of the mind can “make the small gland to which it is closelyjoined move in the manner [façon] that is required to produce theeffect which corresponds to that volition.”28 These passages suggest thatthe mind can alter the state of the pineal gland in ways other than bychanging its direction.

Descartes’ casual talk of mind simply moving body, both in strict andtechnical writings and in looser, nontechnical writings, together withthe lack of any clear positive statement of the change-of-directionaccount is evidence enough against Leibniz’s attribution. But, in addi-tion, there are some passages among Descartes’ writings whose senseseems to run directly contrary to the account that Leibniz attributes toDescartes. Consider, for example, some passages of the Passions inwhich the mind is said to act on the pineal gland in ways that appeardifficult to reconcile with the change-of-direction account of interac-tion. Descartes discusses in the Passions the circumstance in which theanimal spirits are moving the gland in such a way as to cause in themind a desire for something that the mind wants to avoid, as, forexample, when the animal spirits, stirred up by the sight and smell ofa glass of fine wine, cause the gland to move in such a way as to implantthe passion of desire for the wine in the mind at the same time that themind wills that the body abstain. Descartes analyzes this familiar situa-tion as a struggle (combat) “between the effort by which the [animal]spirits push the gland to cause the desire for something in the mind,and that by which the mind pushes it back by the volition it has to avoidthat same thing.”29 Descartes gives a similar account of the conflict


26 PA 42. 27 PA 47. 28 PA 34, PA 41. See also PA 43.29 PA 47. It is important to note here the distinction between the passion of desire and a

volition.

between the natural tendencies of the pineal gland and the volition ofthe mind in his account of how it is that we fix our attention: “Thuswhen one wants to hold one’s attention to consider the same object forsome time, this volition holds the gland inclined to the same sidethroughout that time.”30 In both these passages, Descartes representsthe mind as resisting the movement that the pineal gland would have,left to purely mechanical causes; our minds are preventing the glandfrom having motion that it would otherwise have. It is difficult to seehow this can be reconciled with the change-in-direction account ofmind-body interaction, and it seems unlikely that Descartes would haveallowed such passages to creep into his most careful account of themind’s action on the pineal gland if he genuinely held the account thatLeibniz attributed to him.

Or consider, for instance, the comparison that Descartes drawsbetween the action of mind on body and the scholastic account of heav-iness (gravity). According to that theory, at least as Descartes under-stood it, the heaviness of a body is taken to be a real quality, somethingreal and distinct from the body itself that causes the body to movetoward the center of the earth.31 Although Descartes rejects this accountof heaviness in favor of a purely mechanical account of the phenome-non in terms of the laws of motion and impact and the size, shape, andmotion of the particles that make up the heavy body and its ambientmedium, the scholastic theory, still familiar in his day, was of some useto Descartes in explaining his own account of mind-body interaction.For, Descartes claims, if one can understand the scholastic account ofheaviness, then one ought to be able to understand how an immater-ial substance can cause changes in a material substance. Thus, Descarteswrote to Arnauld:

Many philosophers who think that the heaviness of a stone is a real qualitydistinct from the stone believe that they understand well enough how sucha quality can move the stone toward the center of the earth, since they thinkthat they have a manifest experience of it. I, who have persuaded myself thatthere is no such quality in nature, nor thus is there any true idea of it in thehuman intellect, believe that they use the idea which they have of incorpo-real substance to represent that heaviness to themselves. Thus, it is no more


30 PA 43.31 For a discussion of the scholastic theory of gravity and Descartes’ rejection of it, in the

context of his rejection of substantial forms, see Etienne Gilson, Etudes sur le Rôle de laPensée Médiévale dans la Formation du Système Cartésien (Paris: J. Vrin, 1930), pp. 141–90.

difficult for us to understand how mind moves body than it is for them [tounderstand] how this heaviness bears a stone downwards.32

This example is intended to take away some of the mystery surround-ing the question as to how a nonbodily thing can act on a body by givingan example of a nonbodily thing (the real quality of heaviness) thatDescartes’ contemporaries had no trouble accepting as a cause ofmotion. But this would be a curious example to use if Descartes thoughtthat mind could change only the direction in which a body was moving.In the case of a body falling toward the center of the earth, there is nomere change in direction. Rather, the quality of heaviness is thought toproduce new motion in the heavy body where there was none before.The implication is that mind acts on body in the same way.

This implication is clearest of all in another passage relating the actionof mind on body to heaviness, this time comparing the action of mindon body not with the scholastic theory of heaviness but with Descartes’own theory. On Descartes’ account of heavy bodies and free fall, thefalling body is impelled downward toward the center of earth by meansof collisions between that body and other smaller and more quicklymoving bodies in the surrounding medium.33 Thus he wrote in a passage,ironically enough, preserved only in a copy Leibniz made:

If a body is pushed or is impelled to motion by means of a uniform force[semper aequali vi], of course imparted to it by mind (for there can be no suchforce otherwise), and if it is moved in a vacuum, then it would always takethree times longer to travel from the beginning of the motion to the mid-point than from the mid-point to the end. However, there can be no suchvacuum. . . . But suppose that the body were impelled by heaviness. Since thatheaviness never acts uniformly like mind, but [acts by] some other bodywhich already is in motion, is can never happen that a heavy body is impelledmore quickly than that which moves it.34


32 AT V 222–23. See also AT III 667–68; AT VII 441–42.33 For this account of heaviness and free fall, see, e.g., Pr IV 20. Matters are complicated by

a somewhat different account of heaviness that Descartes offers in Le Monde and mentionslater in the Principia, in accordance with which heaviness is due to the centrifugal forcethat pushes the small particles of the subtle matter turning quickly around the earth awayfrom the center of the earth. On this account, heavy bodies are pushed to the center ofthe earth to take the place of the subtle matter that is receding, in accordance withDescartes’ claim that there can be no vacuum. For this account, see AT XI 72–80 and PrIV 23. It is not clear how these two accounts of heaviness are related to one another.

34 AT XI 629–30. This interesting passage comes from a manuscript entitled “Problemata,”preserved only in a copy Leibniz had made. Though one must use these documents withsome care, the passage seems unquestionably authentic. The (mistaken) formula for the

Descartes’ main point in this passage is the contrast between the uniformacceleration due to the activity of mind, and the nonuniform accelera-tion due to heaviness. But it is clear from this passage that Descartesthought that the action of mind on bodies does not result in a merechange in direction. Rather, Descartes quite clearly thought, mind canproduce a real change in the speed of a body, if fact, that mind is the onlynatural means by which a uniform change in speed is possible.

It is, of course, possible that all the passages I have presented can bereconciled with the change-of-direction thesis or that Descartes thoughtthey could or that he actually rendered them consistent with that thesisin some now lost fragment, perhaps even one that Leibniz saw in Pariswhen Clerselier showed him Descartes’ literary remains. But the pas-sages I have cited, together with the lack of any clear and positive state-ment of the change-of-direction account in any of the numerouswritings that do survive, make it likely that Descartes just did not holdthe account of mind-body interaction that Leibniz attributed to him. Atthe very least, the burden of proof is on anyone who wants to claim thatLeibniz’s account of Descartes is correct.35 This by itself leaves us in the


acceleration of a body in a vacuum given a uniform force is uniquely Cartesian andappears in a number of documents as the law of free fall for heavy bodies from 1618 to1629 and is mentioned as a law that Descartes once held in a letter of 1634. See AT X75f, 219; AT I 71–73, 304–5. For an account of Descartes’ struggles with the problem offree fall, see Alexander Koyré, Galileo Studies (Atlantic Highlands, N.J.: Humanities Press,1978), pp. 79–94. Dating the fragment, though, is problematic. In this fragment,Descartes is clearly distinguishing the problem of acceleration given a uniform force fromthat of free fall. But until at least 1629 Descartes identified the two problems. See AT I71–73. This suggests that the passage dates from later than 1629. It is also unlikely thatthe passage dates from later than 1640, the last date in which we have evidence ofDescartes worrying about the derivation of the laws of free fall. See AT III 164–65. But itis hard to date the fragment more closely than that. It may be associated with a letter of1631 in which Descartes claims that “I can now determine the proportion by which adescending stone increases its speed, not in vacuo, but in this air” (AT I 231). But it couldjust as well be associated with a letter of 1637 in which Descartes asks Mersenne to excusehim from answering a question “concerning the retardation which the movement of heavybodies receives from the air where they move,” claiming that such an account involves hiswhole physics and is inappropriate for a letter (AT I 392). External factors suggest a thirddate from the mid-1630s. One fragment in the “Problemata” is dated 5 February 1635and corresponds to material in the Météores of 1637. See AT XI 626.

35 This, of course, raises the question as to why Lebniz attributed the position to Descartes.The best conjecture is that the change-of-direction account of mind-body interaction wascommon among later Cartesians, and Leibniz just assumed that it must have beenDescartes’ position as well. Norman Kemp Smith (op. cit., p. 83 n. 2) cites Clauberg inthis connection. Alan Gabbey has also called my attention to a letter written afterDescartes’ death by Claude Clerselier, Descartes’ friend, translator, and editor, in whichClerselier argues that mind can change only the direction in which bodies move but

dark about the relations between mind-body interaction and Descartes’conservation law, however. Even if Descartes did not hold the change-of-direction account of mind-body interaction, perhaps he had someother way of rendering interactionism consistent with a universal con-servation law. Perhaps he would have argued that whenever a mind putsa body into motion, something somewhere else in the material worldloses the requisite quantity of motion, so that mind serves only to redis-tribute motion in the world, for example.36 Although such a move isopen to Descartes, there is no textual evidence that he as much as con-sidered it. The overwhelming impression that one gets from the textsis that Descartes just was not very concerned about reconciling his inter-actionism with his conservation law. Now, the apparent lack of atten-tion to this problem may be explained in a number of ways. There isalways the possibility that Descartes simply neglected to see the seriousproblem that his position raises. But there is another, better explana-tion for this apparent gap in Descartes’ argument. The case can bemade, I think, that, from Descartes’ point of view, there just is noproblem reconciling interactionism with the laws of nature. That is,there is reason to believe that Descartes may never have been commit-ted to the position that his conservation law holds universally and mayhave allowed for the possibility that animate bodies lie outside the scopeof the laws that govern inanimate nature.

Many versions of the conservation law do, indeed, suggest that the law is intended to hold universally. For example, when introducingthe conservation law in the Principia, Descartes writes: “God . . . in the


cannot add motion. See Clerselier to de La Forge, 4 December 1660, in Clerselier, ed.,Lettres de Mr. Descartes, Vol. III (Paris: 1667), pp. 640–46. I have not been able to examineKemp Smith’s citation. But it is interesting to note that in the letter Gabbey cites Clerselier does not explicitly attribute the change-of-direction account to Descartes. Fur-thermore, the grounds on which Clerselier advances the claim involve a significant depar-ture from Descartes’ thought on motion and determination. Clerselier’s argumentdepends on the claim that to create a motion requires as much power as to create matteritself, whereas determination “n’adjoûte rien de réel dans la Nature” and can thus be manip-ulated by finite minds (Clerselier, loc. cit., pp. 641–43). But this contradicts what Descarteswrote to Clerselier in a letter 15 years earlier, a letter that Clerselier published in VolumeI of his edition of Descartes’ correspondence. Descartes wrote:

It is necessary to consider two different modes in motion: one is the motion alone, orthe speed, and the other is the determination of this motion in a particular direction,which two modes change with equal difficulty. (AT IV 185)

Thus, the Clerselier letter of 1660 gives us no grounds for attributing the change-of-direction account to Descartes himself.

36 This, in essence, is Broad’s response to the objection. See C. D. Broad, op. cit., pp. 107–9.

beginning created matter along with motion and rest, and now, throughHis ordinary concourse alone, conserves just as much motion and restin the whole of it as He put there at that time.”37 It is hard to see howGod could conserve “just as much motion and rest” as He initiallycreated if minds are allowed to add and subtract motion from the worldliterally at will. But when Descartes is being especially careful, he seemsto allow that his conservation law may admit of some exceptions. As Iwill discuss in some detail below, Descartes’ conservation law followsfrom the immutability of God. Thus Descartes writes just a few lines fol-lowing the passage just quoted:

Therefore, except for changes [in quantity of motion] which evident experience ordivine revelation render certain, and which we perceive or believe to happenwithout any change in the Creator, we ought not to suppose that there areany other changes in His works, lest from that we can argue for an incon-stancy in Him.38

Here Descartes clearly admits that there can be violations of the conser-vation law, circumstances in which motion is added or taken away. Thereference to divine revelation suggests that some such violations mightarise from miracles. But Descartes also makes reference to violations that“evident experience . . . renders certain.” an obvious suggestion as towhat Descartes has in mind here is the ability that the human mind hasto set the human body in motion, which, as he told Arnauld, “is shownto us every day by the most certain and most evident experience.”39 Thisnatural reading is confirmed a few pages later in the Principia, whereDescartes is discussing his third law of motion, a law explicitly governedby the conservation law, in which Descartes sets out the general featuresof his account of impact. Descartes writes:

And all of the particular causes of the changes which happen to bodies arecontained in this third law, at least insofar as they are corporeal; for we are notinquiring into whether or how human or angelic minds have the force [vis]to move bodies.40

This is, to be sure, something less than a clear and positive statementthat minds can cause violations in the laws of nature. But, together withthe lack of any attempt to reconcile interactionism with his conserva-tion law, these passages suggest that in the Principia Descartes, at very


37 Pr II 36. 38 Ibid. Emphasis added.39 AT V 222. Emphasis added. 40 Pr II 40. Emphasis added.

least, left open the possibility that the activity of minds is not con-strained by the laws of nature that hold for bodies.41

At this point we can return to the questions raised at the beginningof this section. The passages cited earlier strongly suggest that Descartesdid not hold the change-of-direction account of mind-body interactionthat Leibniz attributes to him. Even more radically, although the textsare not completely decisive on the question, they do suggest thatDescartes at least left open the possibility that his conservation law maybe violated by animate bodies. The philosophical point should be clear.Descartes might have answered Leibniz’s attack on interactionism bysimply denying that the conservation laws must hold for animate bodies.If this were Descartes’ answer, as I suspect it would have been, then evenif Leibniz were to convince him of the falsity of his own conservation law,Descartes would not have been forced to reject interactionism. There isno reason to think that Descartes would have held Leibniz’s conserva-tion laws to be any more universal than he seems to have held his ownto be. And if Leibniz’s conservation laws are not taken to govern thebehavior of animate bodies, then they pose no obstacle at all to the claimthat minds can alter the course of events in the material world.

3. God and the Laws of Nature

In the previous section I outlined one answer that Descartes could havegiven to Leibniz’s argument. I have claimed that, given what he saysabout mind-body interaction, it is open to Descartes to deny the uni-versality of physical law and to deny that antimate bodies are con-strained by the same laws that govern the purely material world.


41 There is one passage in Le Monde that seems to contradict this interpretation. In chapterVII of that work, after having given the laws of motion and having claimed that these lawssuffice for an “a priori demonstration of everything that can be produced” in the newworld that Descartes is building in Le Monde (At XI 47), Descartes says:

And finally, so that there will be no exceptions which prevent [such a priori demon-strations], we shall add to our assumptions, if it pleases you, that God will produce nomiracles, and that the intelligences or rational minds, which we might assume below[in the Traité de l’Homme], will not disrupt the ordinary course of nature in any way.(AT XI 48)

This might be read as a denial that God can perform miracles or that minds can interferein the “ordinary course of nature” in any way. But given what Descartes says about mind-body interaction elsewhere, it is more reasonable to read this as a simplifying assumptionknown to be false but helpful in simplifying the initial presentation of the mechanist worldthat Descartes intended to give in Le Monde.

Thus, it seems, the difference between Descartes’ interactionism andLeibniz’s pre-established harmony comes down to a more basic differ-ence with respect to the scope of physical law. This, however, raises stilldeeper questions. First of all, there is the question of the coherence ofDescartes’ own position. Is the position that the texts suggest consistentwith Descartes’ otherwise mechanistic world view? Can the exclusion ofanimate bodies from the laws of the material world be anything butarbitrary? And, second, there are arguments of Leibniz’s to deal with.Leibniz took it for granted that the laws of nature apply to animatebodies. Are Leibniz’s reasons for holding this position binding onDescartes as well? In the argument I presented at the beginning of thisessay, Leibniz attempts to trace Descartes’ interactionism to a relativelyuncontroversial and straightforward mistake about the true laws ofmotion. The argument I offered in the previous section suggests thatLeibniz’s argument may not be applicable to the position that Descartesactually held. But Descartes’ position may still rest on a mistake, amistake different from the one that Leibniz attributes to him, to besure, a mistake about the scope of physical law rather than its content,but a mistake nevertheless. We must, then, explore whether there issome unobjectionable way for Descartes to exclude animate bodiesfrom the scope of physical law.

One place we might begin is with Descartes’ discussion of the unionof mind and body. In an interesting essay, the only discussion of thisquestion that I known of in the literature, Peter Remnant attempts to link the exclusion of animate bodies from the laws of motion to thediscussion of mind-body unity and interaction found in Descartes’ celebrated correspondence with Elisabeth.42 Remnant notes that forDescartes the world of created things is understood through three dis-tinct primitive notions, the notions of extension, thought, and theunion of mind and body. Descartes writes to Elisabeth that

there are in us certain primitive notions which are as it were models on whichall our other knowledge is patterned. . . . As regards body in particular, wehave only the notion of extension which entails the notions of shape andmotion; and as regards soul in particular, we have only the notion of thought.. . . Finally, as regards soul and body together, we have only the notion of theirunion, on which depends our notion of the soul’s power [force] to move thebody, and the body’s power to act on the soul and cause sensations and passions.43


42 See op. cit. 43 AT III 665. Quoted in Remnant, op. cit., p. 382.

These notions are primitive in the sense that they must be grasped oneby one, apart from all other notions, and cannot be explicated in termsof one another. As Descartes wrote:

If we try to solve a problem by means of a notion that does not apply, wecannot help going wrong. Similarly, we go wrong if we try to explain one ofthese notions by another, for since they are primitive notions, each of themcan only be understood by itself.44

Thus, Remnant claims, “each of these primitive notions defines anautonomous sphere of knowledge.”45 We must understand mind interms of its primitive notion and the laws that follow from it, and bodyin terms of its primitive notion and the laws that follow from it. And,most important, we must understand the animate body, the thing com-posed of the union of mind with body in terms of its primitive notionand the laws that follow from it. To impose the laws of inanimate matteron animate bodies, unions of mind and body, is for Descartes, onRemnant’s reading, a basic mistake that can lead only to confusion andmisunderstanding; it is an instance of attempting to apply one primi-tive notion (that of extension and the laws it obeys) to an object towhich it does not apply. Thus Remnant concludes:

On Descartes’s view there is a system of principles which applies to all purelyphysical interactions among bodies (including most biological processes) and another system which describes intellectual processes. But there is also a third realm, that of animated bodies. Animated bodies can participatein purely physical interactions and when they do their behavior conforms tothe laws of motion. . . . But when they are behaving qua animated the laws of motion do not apply to them – their behavior conforms to a different setof principles, falling under the primitive notion of the union of soul andbody. . . . If all the activities of bodies consisted in animated behavior thenthe laws of motion would have no application; similarly, if all the activities ofthe soul involved its union with its body . . . the principles of intellectionwould have no application; it is only because bodies also behave purely quabodies and minds purely qua minds that these two sets of principles haveapplication. But this is consistent with the occurrence of another sort ofbehavior, subject to another set of principles, namely that of animatedbodies.46


44 AT III 665–66. Quoted in Remnant, op. cit., p. 383. 45 Remant, op. cit., p. 383.46 Ibid., pp. 384–85. Remnant, like most commentators, is too quick to trust Descartes’

answer to Elisabeth here. On this point, see my essay, “Understanding Interaction: WhatDescartes Should Have Told Elisabeth,” essay 8 in this volume.

Remnant’s account of the matter has the ring of truth. Descartesdoes, indeed, treat the union of mind and body almost as if it were aseparate substance, and it is plausible to suppose that he thought of theanimate body as satisfying laws different from the ones that inanimatebodies satisfy.47 But this cannot be the whole story. Surely, some of thelaws applicable to inanimate bodies are also applicable to bodies unitedto minds. Surely, the geometrical properties of the pineal gland are thesame, whether that gland is connected to a mind or not. Surely, a livinghuman being can no more be in two places at the same time than cana corpse. And surely, although the mind enables us to do much thatcannot be done in inanimate nature, it does not allow us to create avacuum in Descartes’ world. Thus, even though animate bodies may beexempt from the laws of motion, there are many other laws that allbodies must obey, even those that are behaving qua animated, to useRemnant’s phrase. And this raises a basic question: What specifically isit about the laws that govern motion that exempts the union of mindand body from their scope? Why are the laws that govern shape, forexample, one mode of extension, greater in scope than the laws thatgovern motion, another mode of extension? The arbitrariness stillremains on Remnant’s account; there still seems no reason whyDescartes can exclude animate bodies from the laws of motion. If thereis any reason why animate bodies can violate the laws that hold for inanimate nature, it must concern not only the doctrine of primitivenotions that Descartes expounds to Elisabeth but also his conceptionof the laws of motion. And if there is any way that Descartes can sustain his position against Leibniz’s claims, it must be found in the dif-ferent accounts of those laws that the two philosopher-scientists offer.Thus, we must for the moment turn away from minds and bodies andinvestigate the ways in which Descartes and Leibniz treat the laws ofmotion.

For Leibniz, the laws of motion, like every other contingent featureof this world, are grounded in God. In particular, they are grounded inGod’s ends, in his decision to create the best of all possible worlds.Leibniz writes:


47 On the mind-body union as a substance distinct from mind and body, see, e.g., GenevièveRodis-Lewis, L’Oeuvre de Descartes (Paris: J. Vrin, 1971), Vol. I, pp. 352–54, and the refer-ences cited in Vol. II, p. 543 n. 29. Rodis-Lewis is quite correct to reject the claim thatDescartes thought of the union of mind and body as a distinct substance, but Descartes’frequent use of the notion of “substantial union” in connection with the mind and body(AT VII 228; AT III 493; AT III 508; etc.) does suggest something of the sort.

. . . The true physics should in fact be derived from the source of the divineperfections. It is God who is the ultimate reason of things and the knowledgeof God is no less the source of sciences [principe des sciences] than His essenceand His will are the source of beings. . . . Far from excluding final causes andthe consideration of a being who acts with wisdom, it is from these that every-thing must be derived in physics. . . . I agree that the particular effects ofnature can and ought to be explained mechanically, though without forget-ting their admirable ends and uses, which providence has known how to con-trive. But the general principles of physics and mechanics themselves dependon the action of a sovereign intelligence and cannot be explained withouttaking it into consideration.48

Leibniz’s physics, then, begins with a consideration of God as the finalcause of the world. Leibniz’s position is, of course, that God acts in accordance with the principle of perfection, that God chose ourworld from among an infinity of other possible worlds because it is the most perfect, the one that has the most order consistent with the greatest variety in phenomena. Now, the order that Leibniz attributes to the world God creates is complex and involves a number of important metaphysical principles. But among these prin-ciples are the laws of nature in general, and among the laws of natureare the laws of motion and the more general metaphysical principleson which they rest. Thus Leibniz wrote in the Principles of Nature and Grace :

The supreme wisdom of God has made Him choose especially those laws of motion which are best adjusted and most fitted to abstract or metaphysicalreasons. There is conserved the same quantity of total and absolute force, or of action; also the same quantity of relative force, or of reaction; and finally, the same quantity of directive force. Furthermore, action is alwaysequal to reaction, and the entire effect is equivalent to its full cause. It is surprising that no reason can be given for the laws of motion which have been discovered in our own time . . . by a consideration of efficient causes or of matter alone. For I have found that we must have recourse to final causes and that these laws do not depend upon the principle of neces-sity, as do the truths of logic, arithmetic, and geometry, but upon the prin-ciple of fitness [principe de la convenance], that is to say, upon the choice ofwisdom.49


48 G III 54–55 (L 353). 49 G VI 603 (L 639–40).

The laws of motion, then, are intertwined with the order that God hasimposed on our world as a consequence of His decision to create thebest of all possible worlds.50

These basic laws governing nature are not without exception, though.God, acting in accordance with some higher principles of order, prin-ciples of supernatural order that, Leibniz thought, lie beyond our com-prehension, can violate the laws that He set down for finite things toobserve. As Leibniz wrote in the Discourse on Metaphysics:

Now, since nothing can happen which is not according to order, it can besaid that miracles are as much subject to order as are natural operations andthat the latter are called natural because they conform to certain subordi-nate maxims which we call the nature of things. For we may say that thisnature is merely a custom of God’s with which He can dispense for any reasonstronger than that which moved Him to use these maxims.51

However, it is important to note, such violations of the subordinatemaxims that constitute the laws of nature are miracles, happenings that,Leibniz argues, must lie beyond the capability of finite beings to bringabout if miracles are to be genuinely distinct from the ordinary courseof nature. Thus Leibniz explained to Clarke:

If a miracle differs from what is natural only in appearance and with respectto us, so that we call a miracle only that which we seldom see, there will beno internal real difference between a miracle and what is natural, and at thebottom every thing will be either equally natural or equally miraculous. Willdivines like the former, or philosophers the latter? . . . In good philosophyand sound theology we ought to distinguish between what is explicable bythe natures and powers of creatures and what is explicable only by the powersof the infinite substance. We ought to make an infinite difference betweenthe operation of God, which goes beyond the extent of natural powers, andthe operations of things that follow the law which God has given them, and


50 For a discussion of the contingency of the laws of nature in Leibniz, see Margaret Wilson,“Leibniz’s Dynamics and Contingency in Nature,” in P. K. Machamer and R. G. Turnbull,eds., Motion and Time, Space and Matter (Columbus, Ohio: Ohio State University Press,1976), pp. 264–89; reprinted in R. S. Woolhouse, ed., Leibniz: Metaphysics and Philosophyof Science (Oxford: Oxford University Press, 1981), pp. 119–38.

51 DM 7. See also Theod. 207; G II 41 (M 44–45); G II 51 (M 57); G II 92–93 (M 115–16).Leibniz claims that the supernatural order that governs miraculous violations of the lawsof nature is beyond our comprehension in DM 16 and in G III 353.

which He has enabled them to follow by their natural powers, though notwithout His assistance.52

So, even though God can violate natural law for the sake of a higherorder, for the sake of supernatural law, nothing in nature can. These sub-ordinate laws govern nature as a whole and without exception, save forthe extraordinary (and infrequent) interference of God.

This conception of natural law and its place in the order that Godimposes on nature has important consequences for Leibniz’s accountof mind and its relation to body. By the argument sketched in section1, if mind could act on body, either directly or through the intermedi-ation of God, then bodies animated by rational minds would violate the laws that govern inanimate bodies. Now, such violations are by nomeans impossible, even if the laws that God imposed on matter are uni-versal in scope and make no distinction between animate and inani-mate matter. But, if God’s laws are universal in that sense, as Leibnizalmost always assumes, then any such violations would be miraculous,even if such violations occurred in an entirely lawlike and regular way.Thus Leibniz writes:

. . . The common system [i.e., direct interactionism] has recourse toabsolutely inexplicable influences, while in the system of occasional causesGod is compelled at every moment, by a kind of general law and as if bycompact, to change the natural course of the thoughts of the soul to adaptthem to the impressions of the body and to interfere with the natural courseof bodily movements in accordance with the volitions of the soul. This canonly be explained by a perpetual miracle.53

Though such a world of perpetual miracles is possible, Leibniz rejectssuch an account of the matter for both methodological and metaphys-ical reasons. Methodologically, the appeal to God that is required toaccount for the constant violation of natural law is an ad hoc appeal toa deus ex machina in quite a literal sense of the phrase. Leibniz writes:

Problems are not solved merely by making use of a general cause [i.e., God]and calling in what is called the deus ex machina. To do this without offering


52 G VII 416–17 (L 715). See also G II 93 (M 116); G IV 520 (L 494). Leibniz sometimesalso suggests a more epistemic definition of a miracle as “a divine act which transcendshuman comprehension.” See C 508–9; G III 353.

53 G VI 541 (L 587). See also Theod. 207; G II 57–58 (M 65); G II 94 (M 117–18); G III 354. It should be noted that Leibniz recognizes a number of senses in which interactionism, particularly of the occasionalist variety, involves perpetual miracles. See M. Gueroult, Malebranche (Paris: Aubier-Montaigne, 1955–1959), Vol. II, pp. 241–53.

any other explanation drawn from the order or secondary causes is, properlyspeaking, to have recourse to miracle. In philosophy we must try to give areason which will show how things are brought about by the Divine Wisdom,in conformity with the notion of the subject in question.54

And metaphysically, the perpetual miracle that interactionism requiresis objectionable insofar as it attributes an imperfection to God’s work.Thus Leibniz writes to Clarke:

But they who fancy that the soul can give a new force to the body, and thatGod does the same in the world in order to mend the imperfections of Hismachine, make God too much like the soul by ascribing too much to the souland too little to God. For none but God can give a new force to nature, andHe does it only supernaturally. If there was a need for Him to do it in thenatural course of things, He would have made a very imperfect work.55

So, if the laws of motion that God decreed are universal and makeno distinction between human being and stone, then order and per-fection, not to mention good scientific method, require that we rejectthe hypothesis of interaction as miraculous. But, one might ask, howdoes Leibniz know that the laws of motion are universal? Surely, Godcould have set things up in such a way that animate bodies followed different laws from bare matter, so that it would be a law of nature thatwhen a mind has an appropriate volition, the animate body to which itis attached is exempted from laws that otherwise govern its behavior.One might suggest, for example, that the laws of nature are hierarchi-cal, as it were, that the laws of physics are dominated by the psy-chophysical laws of mind-body interaction in the same way that, forLeibniz, the totality of laws of nature are dominated by the supernat-ural laws that govern God’s activity and in accordance with which Hecan suspend the laws of nature to satisfy higher laws.56 What is wrongwith such a conception of natural law? Although Leibniz usually takesthe universality of physical law for granted, rarely arguing the pointexplicitly, Leibniz has an answer to this question. From Leibniz’s pointof view, though such a hierarchical world is possible, such a world is less


54 G IV 483–84 (L 457). 55 G VII 375–76 (L 689).56 The position sketched here is Malebranche’s. On the hierarchy of laws, see, e.g., Nicholas

Malebranche, Dialogues on Metaphysics, ed. and trans. Willis Doney (new York: AbarisBooks, 1980), pp. 320–21. On the ability of the mind-body laws to cause suspensions ofthe laws of physics, see Nicholas Malebranche, The Search after Truth and Elucidations of theSearch after Truth, ed. and trans. by Thomas Lennon and Paul J. Olscamp (Columbus, Ohio:Ohio State University Press, 1980), pp. 580–81, 594.

perfect than a world governed by pre-established harmony and, thus,would not have been created. Consider two possible worlds, wg, a worldin which there is direct or occasional interaction, a world that thusembodies a hierarchy of “gappy” laws and a world wh that is governedby pre-established harmony, a world governed by universal and excep-tionless laws. Suppose, first, that wg and wh contain exactly the samephenomena: Sensation and bodily state, volition and action correspondin exactly the same way in each. But, despite the agreement on the phe-nomena, it is obvious that wh, the world of universal and exceptionlesslaws, is considerably simpler and more orderly than wg, the world gov-erned by the hierarchy of gappy laws.57 So, from Leibniz’s point of view,wh must be preferable to wg. But what if wg and wh differ in the varietyof their phenomena? One might argue, in fact, that they must differ insome phenomena if they are to have genuinely different laws. Here theargument is more difficult. But, even in this case, Leibniz seems to holdthat wh is the more perfect world. Leibniz’s position is that simplicity ismore important than variety of phenomena, so that even if the varietyof phenomena in wg were greater than that in wh, the simplicity of thelaws in wh would tilt the balance in favor of that world. The argumentI have sketched is presented most explicitly in a passage from the Theod-icée. Leibniz writes:

Thus, it is necessary to judge that among the general rules which are notabsolutely necessary, God chooses those which are the most natural, thosewhich are the easiest to account for and which also serve to account for otherthings. This is doubtless most beautiful and pleasing, and were the system ofpre-established harmony not otherwise necessary to eliminate superfluous mir-acles, God would have chosen it, since it is the most harmonious [system].The ways of God are the most simple and the most uniform: They are to choose therules which limit one another least. They are also the most fruitful with respectto the simplicity of means. . . . One can, indeed, reduce these two conditions,simplicity and fruitfulness, to a single advantage, which is to produce as muchperfection as is possible. . . . But even if the effect were supposed greater, butthe means less simple, I think that one could say that all and all, the effectitself would be less great, counting not only the final effect but also themediate effect. Thus those who are wisest act, as much as possible, so thatthe means are, in a way, ends as well, that is to say, desirable not only for what


57 It is, of course, a commonplace observation in contemporary philosophy of science thatany statement can be presented as a universal statement. But the distinction between uni-versal and “gappy” laws is clear enough for our purposes here.

they do, but for what they are. Complicated ways occupy too much ground,too much space, too much place, too much time that could have been better used.58

Leibniz thus concludes that the doctrine of pre-established harmony,in which the laws that govern bodies and the laws that govern minds“limit one another least,” is “infinitely more reasonable and worthy ofGod”59 than is any variety of interactionism. Leibniz’s principle of per-fection, the principle in accordance with which God creates the best ofall possible worlds, demands that the laws that God decrees for inani-mate nature hold for human beings as well. Human beings, complexbodies animated by rational minds, must, by the principle of perfec-tion, be an integral part of the world of finite things governed by thesimple and uniform principles that God decrees as the laws of nature,principles that only He can violate, principles whose violation can onlybe miraculous. And if the scope of natural law is to include humanbeings as well as tables, chairs, and potted palms, then, unless we arewilling to embrace the odious hypothesis of perpetual miracle, inter-actionism of any sort must be out of the question.

Leibniz’s position on the scope of physical law is, thus, grounded in some of his most basic metaphysical commitments, the connectionbetween perfection and order and the principle that God creates thebest of all possible worlds. Because of these principles, Leibniz musthold that the laws of nature are universal, and because of these princi-ples, supplemented with some commonsense scientific methodology,Leibniz must reject the perpetual miracles that interactionism entailsfor him. But, for all that, Leibniz’s position is by no means invulnera-ble. There are, to be sure, any number of gaps in Leibniz’s argumentsthat a clever Cartesian might well be able to exploit in defense of amore limited scope for physical law and in support of an interactionistdualism. One might, for example, point out the ad hoc way in whichLeibniz favors order over variety of phenomena in arguing for pre-established harmony over its alternatives. But Descartes himself wouldhave found Leibniz’s claims vulnerable to attack on the most basic level.The considerations of perfection, order, and God’s ends in construct-ing the best of all possible worlds, considerations that led Leibniz toinclude animate bodies within the scope of the laws of physics, and that


58 Theod. 208; emphasis added. The argument is also suggested in G II 94–95 (M 118) andG III 340–41.

59 G II 94 (M 118).

led him from interactionism to pre-established harmony, would havemoved Descartes little, if at all. For Descartes, the immensity and incom-prehensibility of God preclude any appeal to such reasoning to estab-lish the laws that govern the material world. Thus Descartes wrote inresponse to Gassendi:

Although in Ethics, where it is often permissible to use a conjecture, it issometimes pious to consider what end we can conjecture for God to have setout for Himself in ruling the universe, this is certainly out of place in Physics,where everything ought to shine with the firmest reasons. Neither can wepretend that some of God’s ends are better displayed to us than others; forall [of God’s ends] are hidden in the same way in the abyss of His inscrutablewisdom.60

In fact, given Descartes’ radical voluntarism with respect to the eternaltruths, God has no aims or goals, strictly speaking. His volitions are freewith a freedom of complete indifference. God did not set out to createthe world that would be the most perfect; God did not create this worldbecause it is the most perfect one. Rather, it is the most perfect onebecause God created it.61

The rejection of final causes in physics marks a basic differencebetween Cartesian and Leibnizian physics. But this does not mean thatDescartes rejects Leibniz’s grounding of physics in the activity of Godor Leibniz’s claim that true knowledge of the physical world must bederived from our knowledge of God. Neither does it mean that the lawsof physics are inaccessible to rational argument or demonstration.Rather, Descartes claims, they are to be derived not from God as a finalcause but from God as an efficient cause. Thus he wrote:

And finally, we shall not seek the reasons for natural things from the endswhich God or nature propose for themselves in making them, since we oughtnot to be so arrogant as to think that we participate in their counsels. Butconsidering Him as the efficient cause of everything, we must see what canbe concluded from those attributes of which He allows us some notion, aboutthose of His effects which the senses make apparent to us, by means of thelight to nature which is innate in us.62

The laws of nature, then, are to be derived not from considerations oforder, perfection, and God’s ends in creating this world, as they are for


60 AT VII 375. See also AT VII 55.61 See AT VII 432. For Leibniz’s remarks on this claim, see, e.g., DM 2.62 Pr I 28. For Leibniz’s comments on this, see, e.g., G IV 360–61 (L 387).

Leibniz, but from His nature and the way in which He operates in theworld. The laws of nature are not chosen by God and imposed on theworld. Rather, they follow directly from the way in which God acts onthe world. To use a distinction familiar from recent moral theory,whereas Leibniz’s God is a teleologist, acting for the end of order andperfection, Descartes’ God is a deontologist, doing the right thingfrom moment to moment, whatever might come of it. Consequently,for Descartes, one cannot appeal to order and perfection to justify oneconception of the world over another.

This strategy for deriving the laws of nature is apparent in the argu-ment that Descartes offers for his conservation law. The law is resentedin the context of a discussion of the “universal and primary” cause of motion, that which is the “general cause of all motions which are inthe world.” This general cause is, of course, “none other than GodHimself,” who

in the beginning created matter along with motion and rest, and now,through His ordinary concourse alone, conserves just as much motion andrest in the whole of it [i.e., the material wold] as He put there at that time.. . . We also understand God to be perfect not only insofar as He is, inHimself, immutable, but also in that He works [operetur] in as constant andimmutable a way as possible. Therefore, except for those changes [in quan-tity of motion] which evident experience or divine revelation render certain,and which we perceive or believe to happen without any change in theCreator, we ought not to suppose that there are any other changes in Hisworks, lest from that we can argue for an inconstancy in Him.63

The precise intuitions behind Descartes’ proof are illuminated by otherpassages in which Descartes discusses the operation of God in the world.Descartes notes that the nature of time is such that:

its parts do not depend on one another, and never exist simultaneously; andtherefore from the fact that we exist now, it does not follow that we will alsoexist in the next following time unless some cause, indeed the same onewhich produced us at first, continually re-creates us, that is, conserves us.64

Thus, Descartes claims, God must continually re-create the world atevery moment, or else it would pass into nonexistence. This providesan obvious way of seeing how God’s immutability results in the conser-vation law for Descartes. Descartes argues: “[God] conserves [motion]


63 Pr II 36. 64 Pr I 21. See also AT VII 48–49.

just as it is at the moment in which it is being conserved, without regardto what it was a bit before.”65 God’s immutability requires that when Here-creates the world from one moment to the next, He must recreateit as much as possible as it was the previous moment. In part, He mustre-create the world with the same quantity of motion it had the momentbefore.

In this argument Descartes is quite explicitly following the strategyhe set out for deriving “reasons for natural things.” He is consideringGod as an efficient cause, the cause of motion in the beginning, and thecontinuing cause of motion in the moment-by-moment conservation of the world.66 He then considers God’s attributes, the fact that God’sperfection involves constancy of operation and argues from that to theconservation law. Descartes’ reasoning is not without its problems here. The derivation is obscure, complex, and the conclusion ultimatelywrong, as Leibniz successfully showed. But it is the strategy that I aminterested in here, what Descartes thought he was doing, and that isclear enough. The conservation law for Descartes is not a law that Godimposes on the world to further some end; it is intended to be a con-sequence of the constraints that God’s nature imposes on God as anefficient cause of motion in the material world.

Descartes’ conception of the conservation law and its ground in the


65 Pr II 39. See also the parallel passage in Le Monde, AT XI 44. The argument is somewhatmore complex than the brief exposition I have given suggests. Since each moment is without duration, there can be no motion, strictly speaking, at any given moment, asDescartes fully realized. See, e.g., Pr II 39; AT II 215. What is preserved from one momentto the next, then, cannot be motion itself but the tendency or inclination to motion. And,Descartes would have had to have held, in order to preserve the tendency to motion fromone moment to the next, God would have to create the moving body at a somewhat dif-ferent place from one moment to the next if this tendency is ever to result in any actualmotion. On the notion of momentary tendency to motion, Descartes’ need for such anotion, and the problems it raises for his metaphysics, see, e.g., F. Alquié, ed., OeuvresPhilosophiques de Descartes (Paris: Garnier Frères, 1963–1973), Vol. I, p. 359 n. 1; ThomasL. Prendergast, “Motion, Action, and Tendency in Descartes’ Physics,” Journal of the Historyof Philosophy, Vol. 13 (1975), pp. 453–62; and Martial Gueroult, “The Metaphysics andPhysics of Force in Descartes,” trans. in Stephen Gaukroger, ed., Descartes: Philosophy, Math-ematics and Physics, pp. 196–229. Gueroult’s final judgment is that “instantaneous movingforce, the distinction between the instant of motion and the instant of rest, . . . pose[s] aninsoluble problem for Cartesian metaphysics” (p. 222).

66 Peter Machamer argues that, whatever Descartes’ intentions were, final causes inevitablycreep into his derivation of the laws of nature. See his “Causality and Explanation inDescartes’ Natural Philosophy,” in P. K. Machamer and R. G. Turnbull, eds., op. cit., pp.168–99. Although I think that Descartes can be defended on this point, it is beyond thescope of this essay to do so. What is important in this context is simply how Descartes con-ceived of his enterprise.

immediate activity of God has important consequences for the way inwhich he conceives of mind in the context of the order of nature. Theconservation law is, for Descartes, a law that follows out of the way inwhich God acts as an efficient cause of motion. As an efficient cause of motion, He must, by virtue of His nature, act in such a way as to preserve the same quantity of motion from moment to moment. But,Descartes says, although God is the “universal and primary” cause ofmotion,67 He is not the only cause. As he wrote to More:

The translation which I call motion, is a thing of no less entity than shape: itis a mode in a body. The force moving [a body] can be that of God Himselfconserving the same amount of translation in matter as He put in it in thefirst moment of creation; or also [it can be] that of a created substance, likeour mind, or that of some other thing to which He gave the force of movinga body.68

Now, when God causes motion, the motion He causes must observe theconservation law. But there is no reason at all to impose similar con-straints on finite and imperfect causes of motion. That is, even thoughfinite, imperfect minds may act in some law like way, deriving from theirfinite and imperfect natures, the motion they cause need not satisfy theconservation principle. They may add or subtract motion from the world,even if God cannot. To suppose that they do argues for no change in GodHimself and does not give us grounds for imputing an “inconstancy inHim.”69 Thus, it seems, there is nothing arbitrary or inconsistent withDescartes’ principles to suppose that animate bodies, bodies capable ofbeing acted upon by minds, can violate the conservation principle. Suchbodies stand, as it were, outside the world of purely mechanical nature.The conservation principle governs only purely material systems in nature,systems in which God is the only cause of motion.70


67 Pr II 36.68 AT V 403–4. This position is not without its problems. This passage puts the activity of

mind in causing motion on a par with that of God. But, surely, however minds causemotion, they do not do it as God does, by way of a continual re-creation. In fact, it seemsdifficult to see how the mental causation of motion could be reconciled with the contin-ual recreation pricture at all. Malebranche seizes on exactly this problem, using it to pushDescartes to occasionalism in the seventh of his Dialogues on Metaphysics. There is no reasonto believe, though, that Descartes was aware of this difficulty with his position.

69 Pr II 36.70 The precise wording in the letter to More quoted above (“the force . . . can be that of God

Himself conserving the same amount of translation in matter as He put in it in the firstmoment of creation”) suggests a somewhat different conclusion from the one I have

It should be clear by now that Descartes’ interactionism rests on nosimple mistake, either about the content or the scope of physical law.Because of his general rejection of final courses in physics, he has adefense against the arguments from the principle of perfection that ledLeibniz to pre-established harmony.71 And because of his conception of


drawn. Read literally, it seems to say that what is conserved from moment to moment isprecisely the quantity of matter that God put into the world at the beginning, implying that,even if minds could add motion in one moment, God would simply fail to preserve it inthe next. If this were Descartes’ position, then even though minds could, in a sense, causemotion, the motion would not persist; the conservation principle would govern all bodies,animate and inanimate, with the exception of momentary lapses. But there is no reasonto attribute such a strange position to Descartes. The position that the literal reading ofthat sentence suggests is inconsistent with the account of God’s continuous re-creation ofthe world given in the context of Descartes’ derivation of the laws of motion, in accor-dance with which “[God] conserves [motion] just as it is at the moment in which it isbeing conserved, without regard to what it was a bit before” (Pr II 39; see also AT XI 44).For God to destroy motion added by mind would require Him to “remember” how muchmotion there was at the beginning in deciding how much to create at the next moment.Given the central role that this conception of continuous re-creation plays in the deriva-tion of the laws of motion, it seems most likely that Descartes’ remarks to More are notmeant to be read so literally.

71 There is reason to believe that Descartes may have been explicitly aware that there is someconnection between the admission of final causes, the claim that God created the mostperfect world, and a position much like Leibniz’s pre-established harmony. In a remark-able but almost entirely unnoticed passage, Descartes wrote:

It is a strong conjecture to affirm anything which, if assumed, would make God under-stood as being greater or the world as being more perfect: as, for example, that thedetermination of our will to local motion always coincides with a corporeal cause deter-mining motion; that miracles are always consistent with natural causes, etc. (AT XI654)

The passage is found in a series of gleanings from Descartes’ manuscripts preservedamong Leibniz’s papers. This portion of the manuscript is entitled “Annotations whichDescartes seems [videtur] to have written in [or, on] his Principia Philosophiae” and may, Isuspect, have been marginalia in Descartes’ own copy. For a brief account of the manu-scripts and their history, see AT X 207–10. The remark quoted is the second in a seriesof discrete paragraphs. The paragraph preceding the quote can plausibly be read as acomment on Pr I 26, and the paragraphs succeeding the quote link up naturally with PrI 30, Pr I 30, Pr I 31, Pr I 33, Pr I 37, and so on in order. This suggests that the text quotedmay well be a comment on Pr I 28, a passage quoted above in which Descartes explicitlyrejects the appeal to God’s purposes in particular and final causes in general. This, inturn, suggests that Descartes thought that if his strictures against final causes were lifted,then pre-established harmony would be a reasonable position to adopt. Although thispassage indicates that Descartes may have been aware of some connection between aversion of pre-established harmony and the appeal to God as the creator of the best of allpossible worlds, it gives us no reason to believe that Descartes was aware of the full posi-tion, as Leibniz develops it, nor does it give us any indication as to how precisely Descartessaw the connection between the claim that the world is perfect and the claim that “thedetermination of our volition to local motion always coincides with a corporeal cause

the laws of motion as deriving from the action of God as an efficientcause of motion, Descartes can exempt animate bodies from the lawsthat govern inanimate bodies in motion in a coherent and nonarbitraryway and allow mind to affect the behavior of body. Descartes’ interac-tionism thus rests reasonably secure against Lebniz’s attack. This is aninteresting conclusion in and of itself. But, I think, the defense I havesketched gives something even more interesting, an insight into the real differences that separate Descartes’ and Leibniz’s positions. Whatforces Leibniz to reject interactionism and to adopt pre-establishedharmony is the fact that for him mind is an integral part of a world governed by principles of order, overarching metaphysical principlesdecreed by a wise and benevolent God. In Leibniz’s best of all possibleworlds, simplicity and tidiness dictate that the laws of nature that Goddecreed must, miracles aside, govern all bodies, both animate and inanimate, thus ruling out any variety of interactionism. For Descartes,though, the wisdom of God is beyond our reach; simplicity and orderare just not at issue. The laws of motion are not, for Descartes, prin-ciples of order that God imposes on the world but, rather, a direct consequence of the laws that God Himself obeys as one of a number of possible causes of motion in the world. Because mind is a cause ofmotion that lies outside the scope of the laws that govern God’s activity,Descartes can maintain his interactionism in spite of Leibniz’s argu-ment. What explains Leibniz’s rejection of interactionism, then, can be no simple discovery that Descartes’ conservation law is wrong, asLeibniz seems to have believed. Rather, what separates Leibniz’saccount of the relation between mind and body from Descartes’ issomething much deeper and more significant, a change in the place ofmind in the natural order of things, a change motivated by a funda-mental shift in the very conception of what a law of nature is and howit derives from God.


determining motion.” However, the fact that this passage was preserved in a copy Leibnizmade during his crucial stay in Paris in 1672–1676, before Leibniz’s mature systememerged, suggests that Leibniz’s contact with Descartes’ thought may have played somerole in the formulation of the doctrine of pre-established harmony.

8

UNDERSTANDING INTERACTION

What Descartes Should Have Told Elisabeth

A typical textbook account of the philosophy of mind in the seventeenthcentury goes something like this. Descartes believed in two kinds of stuff,mental stuff and material stuff, substances distinct in nature that gotogether to constitute a single human being. But Descartes also took itfor granted that these two substances were capable of genuine causalinteraction, that minds can cause bodily events, and that bodies cancause mental events, i.e., that acts of will can genuinely cause changes inthe state of the human body, and that the state of the sensory organs andthe brain can cause sensation and imagination in the mind. But, the storygoes, Descartes went astray here and vastly underestimated the philo-sophical problems inherent in his position. Descartes, it is claimed,repressed, or even worse, simply ignored the central question his posi-tion raises: How is it even possible that an immaterial substance, like themind, could conceivably act on an extended substance like the humanbody? According to the standard account, later philosophers recognizedthe inherent unintelligibility of Descartes’ position and started one ofthe largest cottage industries in the history of philosophy, the attempt toprovide satisfactory solutions to the mind-body problem, intelligibleaccounts of how mental and physical events are related to one another.Realizing the unintelligibility of the doctrine of causal interactionism,this cottage industry produced such noteworthy products as occasional-ism, dual-aspect theory, pre-established harmony, and so on, all in theattempt to fill in the gap in Descartes’ dualist program.1

168

1 This standard account dates back to the seventeenth century. For an account of this readingin the texts of Spinoza, Leibniz, and Malebranche, see, e.g., Jean Laporte, Le Rationalisme

This general outline can (and has) been challenged; the actualhistory of philosophy is much richer than any of its rationalized reconstructions. Sympathetic commentators usually call attention to animportant pair of letters that Descartes wrote to the Princess Elisabethin 1643,2 where Descartes takes up just this question, the intelligibilityof mind-body interaction, and offers a philosophically interesting andsophisticated account of why he thinks that the notion of mind-bodyinteraction is perfectly intelligible on its own terms, and why it neitherneeds nor admits of clarification.3

Now, the letters to Elisabeth are carefully thought out responses tothe very questions that troubled later philosophers about Descartes’view, and as such, they deserve careful study. But there is a curious difficulty in using these letters as the key to Descartes’ position. No oneseems to have noticed that Descartes is just not entitled to the answer he gives Elisabeth; despite Descartes’ clear endorsement, the answer he gives Elisabeth is blatantly inconsistent with other well entrenchedaspects of the Cartesian system.

The defense of this claim will be the central task of this essay. I shall begin with an exposition of the account Descartes gives of mind-body interaction in the letters he wrote to Elisabeth in May and Juneof 1643, letters that form the first line of defense for Descartes’ inter-actionism among those commentators who are committed to defend-ing Descartes’ position. After a short digression on a curious analogyDescartes makes between his position and the Scholastic account ofheaviness and free fall, I shall examine Descartes’ answer to Elisabethin some detail, and argue that it is inconsistent with the foundationsDescartes gives to his theory of motion. Finally, I shall attempt to sketchout an answer that Descartes could have given to Elisabeth in 1643, ananswer that seems both philosophically interesting, and consistent withthe rest of his writings.

understanding interaction 169

de Descartes (Paris, 1950), pp. 220–25. Richard Watson discusses similar themes in lesser known Cartesians of the late seventeenth century in his book, The Downfall of Carte-sianism (The Hague, 1966). The claim that interaction is the scandal of Descartes’ philos-ophy is still commonplace in the standard commentaries. See, e.g., Anthony Kenny, Descartes(New York, 1968), pp. 222–26; and Bernard Williams, Descartes (New York, 1978), pp. 287–88.

2 Descartes to Elisabeth, 21 May 1643, AT III 663–68; Descartes to Elisabeth, 28 June 1643,AT III 690–95.

3 For instances of this more sophisticated reading, see, e.g., Jean Laporte, op. cit., pp. 220–54;Henri Gouhier, La Pensée Métaphysique de Descartes (Paris, 1962), pp. 321–44; and RobertRichardson, “The ‘Scandal’ of the Cartesian Interactionism,” Mind 91 (1982).

Before entering into the argument proper, though, I would like tomake a few prefatory remarks concerning the issues I intend to takeup, and the issues I don’t. The issue that I intend to focus on is that ofthe intelligibility of mind-body interaction. The issue is, admittedly, afuzzy one, as fuzzy as the notion of intelligibility itself. But historicallyspeaking, it is an important one, as the reaction of Descartes’ contem-poraries and successors shows. To make the question a bit more precise,I shall construe it, as Descartes and his contemporaries often seemedto do, as the problem of whether the notion of mind-body interactionis somehow intelligible on its own terms, or whether its intelligibilityrequires an explication, analogy, or analysis in terms of some other dis-tinct variety of causal interaction, itself more basic, or, at least, betterunderstood. To be more precise still, given the prominence of thenotion of impact in the then modish mechanistic world view, the ques-tion of the intelligibility of mind-body interaction quickly becomes aquestion of whether mind-body interaction can be understood withoutsomehow relating it to the way in which bodies cause changes in oneanother through impact.4 The question of intelligibility should be dis-tinguished from the closely related question of whether or not the mindand body do, as a matter of fact, actually interact with one another.Though Descartes and his correspndents and critics often link the twoquestions for obvious reasons, they are really somewhat independent.One can hold that despite the intelligibility of mind-body interaction,minds and bodies do not, as a matter of fact, interact with one another.Philosophically, some reason must be given over and above the bareintelligibility of interactionism for adopting that position. Descartesdoes have an answer to this question, and an interesting one: It is ex-perience, he claims, “the surest and plainest everyday experience,”5 ashe writes to Arnauld, that convinces us of the truth of interactionism.


4 There are, of course, other ways in which the question of the intelligibility of mind-bodyinteraction could be raised. One could take it to be a question about how interaction canbe reconciled with certain commonsense notions about causality, in particular with the socalled “reality principle” (the cause must, in some sense, contain everything that is in theeffect), or with the intuition that causal relations can only hold among things that are suf-ficiently similar. On this question see, e.g., Richard A. Watson, op. cit., passim, but espe-cially pp. 33–36; Louis Loeb, From Descartes to Hume (Ithaca, 1981), pp. 134–43; andchapters I and II of Eileen O’Neill’s unpublished Ph.D. dissertation, Mind and Mechanism(Princeton, 1983). Another kind of incoherence involves the question as to how mind-bodyinteraction can be reconciled with a law-governed conception of the material world likeDescartes’. On this question see, e.g., Louis Loeb, op. cit., pp. 143–48, and Daniel Garber,“Mind, Body, and the Laws of Nature in Descartes and Leibniz,” Midwest Studies in Philoso-phy 8 (1983), essay 7 in this volume.

5 Descartes for [Arnauld], 29 July 1648, AT V 222.

But as important as this question is, it will not interest me here. Myconcern will be with bare intelligibility.

Even more specifically, my main focus will be the bare intelligibilityof the causal link in only one direction. Descartes’ interactionism hastwo aspects: the mental causation of bodily events (volition) and thebodily causation of mental events (sensation and imagination). Whileboth aspects are important, I shall be concerned mainly with theformer, mind-body rather than body-mind causation. In part this is tonarrow the range of the discussion. But more important, the accountof body-body causation that, I shall argue, runs through Descartes’ writ-ings on physics makes it, to my mind at least, virtually impossible tounderstand how he conceived of body-mind causation. The reasons forthis will become clearer as the argument progresses, I hope, and I shallpoint them out when the time comes. But this is an issue that I wouldlike to sidestep in this essay.

And finally, there is one last issue I would like to sidestep. It will become apparent that mind-body interaction is closely connected withthe question of the unio substantiale, as Descartes called it, the substan-tial or real union between the mind and body. As a consequence of thisdoctrine, strictly speaking, one should not talk about a causal interactionbetween two different things, a mind and a body; one should talk aboutthe causal explanation of certain behavior or states of a single thing, themind-body union, in terms of mental acts of will or the physical statesof the body.6 But while I recognize that an understanding of Descartes’doctrine of the unio substantiale is important to a full understanding ofDescartes’ position on sensation and voluntary action, I shall try asmuch as possible to avoid this tangled issue. And, consequently, I shallfollow Descartes’ usual practice, and that of his correspondents, andconsider the problem as one of making intelligible the interactionbetween two substances.

I. The Doctrine of the Three Primitive Notions

Any attempt to come to terms with Descartes’ thought on mind-bodyinteraction must begin with a few short letters exchanged betweenDescartes and the Princess Elisabeth, the most explicit discussion of the


6 For an account of the substantial union of mind and body and some aspects of its relationto the problem of interaction, see, e.g., Geneviève Rodis-Lewis, L’Oeuvre de Descartes (Paris,1971), vol. I, pp. 351–65, and the numerous references cited there; and Henri Gouhier,op. cit.

problems raised by Descartes’ interactionism in the corpus of his writings. The exchange begins with a question Elisabeth raises. She asksDescartes to explain:

how the mind of a human being can determine the bodily spirits [i.e., thefluids in the nerves, muscles, etc.] in producing voluntary actions, being onlya thinking substance. For it appears that all determination of movement isproduced by the pushing of the thing being moved, by the manner in whichit is pushed by that which moves it, or else by the qualification and figure ofthe surface of the latter. Contact is required for the first two conditions, andextension for the third. [But] you entirely exclude the latter from the notionyou have of the body, and the former seems incompatible with an imma-terial thing.7

Or, as Elisabeth put the question when, unsatisfied with Descartes’ firstanswer, she wrote for further clarification:

And I admit that it would be easier for me to concede matter and extensionto the mind than it would be for me to concede the capacity to move a bodyand be moved by one to an immaterial thing.8

The problem Elisabeth has is an obvious and understandable one; shefinds it impossible to conceive of how a non-extended mind can causechanges in an extended body. On the other hand, she finds the mech-anist’s conception of how one body can change the motion of anotherbody at least reasonably unproblematic. There appears to be no mysteryfor Elisabeth with the phenomenon of impact that constitutes the basicconcept in a mechanist physics like Descartes’ own. What she seeks issome connection between the two domains, a way of understanding theseemingly incomprehensible mechanism of mind-body interaction interms of the relatively more intelligible phenomenon of body-bodyinteraction.

Descartes’ reply is reasonably clear. Put briefly, Descartes denies thatthe mechanical explanation of change in terms of impact is relevant tothe question as to how mind acts on the body. The claim is that we havea special notion in terms of which we understand mind-body interac-tion, a notion distinct from the notions in terms of which we under-


7 Elisabeth to Descartes, 6/16 May 1643, AT III 661.8 Elisabeth to Descartes, 10/20 June 1643, AT III 685. Other contemporary critics and cor-

respondents made the same point to Descartes. See, e.g., Gassendi’s remarks in his FifthObjections, AT VII 341; Arnauld to Descartes [ July 1648], AT V 215; More to Descartes, 11December 1648, AT V 238–39.

stand things that pertain to the mind or to the body taken separately.Descartes argues as follows in his first reply to Elisabeth:

First I observe that there are in us certain primitive notions which are, as itwere, the originals on the pattern of which we form all of our other thoughts.. . . First, there are the most general ones, such as being, number, and dura-tion. . . . Then, as regards body in particular, we have only the notion of exten-sion, which entails the notions of shape and motion; as regards mind inparticular, we have only the notion of thought, which includes the concep-tions of the intellect and the inclinations of the will. Finally, as regards themind and body together, we have only the notion of their union, on whichdepends our notion of the mind’s power to move the body, and the body’spower to act on the mind and cause sensations and passions. I observe nextthat all human science consists solely in clearly distinguishing these notionsand attaching each of them to the things to which it applies. For if we try tosolve a problem by means of a notion that does not apply, we cannot helpgoing wrong. Similarly, we go wrong if we try to explain one of these notionsby another, for since they are primitive notions, each of them can only beunderstood by itself. The use of our senses has made the notions of exten-sion, shape, and movement more familiar to us than the others; and the maincause of our errors is that we commonly want to use these notions to explainmatters to which they do not apply. For instance, we try to use our imagina-tion . . . to conceive the way in which the mind moves the body after themanner in which one body is moved by another. . . . So I think that we havehitherto confounded the notion of the mind’s power [force] to act on thebody with the power one body has to act on another.9

Descartes’ full answer to Elisabeth is what might be called the doctrineof the three primitive notions. General notions aside, we have within usthree basic ideas, that of mind, that of body, and that of their union. Eachis separate, each is distinct, and each has its own domain of application;each is per se intelligible, and cannot be explained in terms of otherprimitive notions. Elisabeth’s mistake is that of trying to explain onenotion, that of mind-body interaction, which pertains to the primitivenotion of the union of mind and body, in terms of impact, which per-tains to another primitive notion, that of extension or body, somethingthat is neither necessary, since each notion is per se intelligible, nor pos-sible, since the notions are completely distinct. Mind-body interactioncan be grasped only by grasping the unity of mind and body. Since theprimitive notion of mind-body units is made “familiar and easy to us”


9 Descartes to Elisabeth, 21 May 1643, AT III 665–66. See also Descartes to Elisabeth, 28June 1643, AT III 691–92; and Pr I 48.

only through the senses, Descartes recommends that the young Princessabstain from philosophy, and re-enter everyday life.10 We have a notionthat is per se intelligible in terms of which to understand interaction,and if anyone, like Elisabeth (or Arnauld, or Gassendi, or More, orReguis . . .) fails to see this, it must be because their minds are confusedand cluttered. What is called for is a bit of therapy, not argument or expla-nation. Go about your daily life, and you will find the appropriate notion,just as the unreflective man in the street does.

This is how Descartes tries to explain himself. It can, admittedly, looksomewhat suspicious, as if Descartes is simply declining to deal with aserious problem, claiming to understand something that is just unin-telligible. Worse than that, Descartes looks as if he is patronizing thesincere but penetrating young Princess who, many later readers havejudged, actually got the better of the older and more distinguishedDescartes in this exchange.

But I don’t think that this is fair. I agree with Descartes’ sympatheticcommentators in seeing Descartes as offering a philosophically sophis-ticated answer to Elisabeth’s serious question. The doctrine of the threeprimitive notions is an interesting and not implausible claim about whatis going on in the mind, about our native endowments. It is, further-more, a claim that coheres well with the epistemology and account ofour mental faculties that Descartes already worked out in the unpub-lished Regulae and the then recently published Meditations.

Descartes’ answer is a philosophically serious answer. While it maynot ultimately hold up under philosophical scrutiny (what answer towhat problem, alas, has?), it cannot be dismissed as begging the ques-tion or patronizing the questioner. On this much I agree with a numberof friends of Descartes’. But the defense of the intelligibility of Carte-sian interactionism cannot end here. For the answer Descartes gave toElisabeth, while interesting and, perhaps, defensible, is flawed in animportant way; it is, I claim, not the answer that should have beenoffered by the author of Le Monde and the Principia.

II. The Heaviness Analogy

Before making good on my claim, though, I would like to digress for afew pages, and point out one comparison that Descartes does think


10 This is the general theme of the letter, Descartes to Elisabeth, 28 June 1643, AT III690–95.

illuminates the account of mind-body interaction, a comparison thatinvolves the Scholastic account of free fall or heaviness. In part, I wantto deal with an obvious question that this raises: How is this compari-son different from the one that Elisabeth suggests? How is the use ofthis comparison consistent with Descartes’ apparent claim that com-parisons can be of no use in illuminating mind-body interaction? Butin addition to dealing with these questions, I want to point out some-thing that this discussion of Descartes’ suggests, a way of looking atDescartes’ conception of mind-body interaction that will be helpful inunderstanding the account of that notion that, I shall argue, better suitsDescartes’ system than the one he offered.

On the Scholastic account of heaviness, at least as Descartes under-stood it, the heavy body is impelled to the center of the earth by thereal quality of heaviness, something distinct from the body itself, some-thing incorporeal.11 This account, which Descartes thinks is intelligibleand generally understood,12 can be helpful in getting his correspon-dents to understand his conception of mind-body union and interac-tion. Thus, Descartes writes to Elisabeth:

When we suppose that heaviness is a real quality of which all we know is thatit has the power [force] to move the body that possesses it towards the centerof the earth, we find no difficulty in conceiving how it moves the body orhow it is united to it. We do not suppose that the production of this motiontakes place by a real contact between two surfaces, because we experience inourselves that we have a specific notion to conceive it by. I think that wemisuse this notion when we apply it to heaviness, which as I hope to show inmy physics [i.e., the yet to be published Principia Philosophiae], is not anythingreally distinct from body; but it was given us for the purpose of conceivingthe manner in which the mind moves the body.13

It is important here to appreciate the difference between the analogythat Descartes appeals to, and the comparison Elisabeth makes between


11 For an account of the Scholastic theory of form and quality as Descartes understood it,and one of his principal lines of attack against it, see Etienne Gilson’s classical essay, “Dela critique des formes substantielles au doute méthodique” in his Etudes sur le rôle de lapensée médiévale dans la formation du système cartésien (Paris, 1930), pp. 141–90.

12 At least he usually concedes this. Descartes takes a different position in his letter to Regius,January 1642, AT III 506, 507.

13 Descartes to Elisabeth, 21 May 1643, AT III 667–68. Descartes uses similar comparisonsin other writings as well. See, e.g., Descartes to Hyperaspistes, August 1641, AT III 424;Descartes for [Arnauld], 29 July 1648, AT V 222–23; and the Letter of Mr. Descartes to Mr.C.L.R. [i.e., Clerselier], AT IXA 213.

mind-body and body-body interaction, a comparison that Descartesrejects. Descartes’ criticism of Elisabeth is that she is attempting tounderstand one primitive notion in terms of another, something that canonly lead to grief. But the situation is altogether different with theScholastic analogy to which Descartes appeals. As Descartes claimed inhis reply to the Sixth Objections, in a passage to which he calls Elisabeth’s attention, the common idea of heaviness, the idea theScholastics and the common man and the idea that Descartes himselfhad in his naive and sense-bound youth, is, in fact, derived from the idea we have of mind. Descartes writes:

The chief sign that my idea of heaviness was derived from that which I hadof the mind is that I though that heaviness carried bodies toward the centerof the earth as if it contained some cognizance [cognitio] of this center withinit. For it could not act as it did without such cognizance, nor can there beany such cognizance except in the mind.14

Thus, Descartes can claim, as he did to Elisabeth in the passage I quotedearlier, our notion of how the real quality of heaviness acts on the bodyto which it is attached must be derived from the notion we have of howthe mind acts on the body. Now, since Descartes assumed that hisreaders were conversant with the Scholastic account of heaviness, hethought that he could use this familiar doctrine to call his skepticalreader’s attention to the notion of mind-body union and interaction,and point out that, despite their claims of not being able to conceivehow an incorporeal mind could act on an extended body, they reallydo have the notion in question. This is what he explained to Arnauld,to whom he offered the same analogy in 1648, five years after the lettersto Elisabeth:

So, it is no harder for us to understand how the mind moves the body thanit is for them [i.e., the Scholastics] to understand how such heaviness movesa stone downwards.15

Whether or not this explanatory device was successful,16 it is clear thatDescartes is entitled to use it. Unlike the comparison Elisabeth presses,the comparison between mind-body causation and mechanical causa-tion, in Descartes’ comparison there is no real analogy, no comparison


14 Sixth Replies, AT VII 442. 15 Descartes for [Arnauld], 29 July 1648, AT V 222–23.16 We don’t know Arnauld’s reaction, but the tactic wasn’t particularly successful with

Elisabeth. See Elisabeth to Descartes, 10/20 June 1643, AT III 684.

between two different notions. Rather, Descartes claims, there is an identity: The same notion, that of mind-body union and interaction is atissue in both contexts. Only in one of those contexts it is misapplied.

This is all a fairly straightforward and unproblematic exposition ofwhat Descartes was up to, of why Descartes thought the analogy drawnfrom Scholastic science was helpful, and, unlike the analogy Elisabethtries to draw from mechanist science, unproblematic. But I would liketo point out an interesting aspect of Descartes’ use of the heavinessanalogy. The account that Descartes gives of the Scholastic theory ofheaviness makes the primitive notion of mind-body unity and the correlative notion of mind-body interaction conceptually basic in anextremely interesting sense. Descartes’ claim is that the Scholastic sci-entist is just projecting his innately given conception of his own com-posite nature onto the inanimate world;17 unless the Scholastic scientisthad this primitive notion pertaining to the union of mind and body, hecouldn’t understand the explanations he gives of phenomena in theinanimate world. That is, as Descartes understands it, our comprehen-sion of Scholastic explanations in terms of substantial forms and realqualities is parasitic on the notions we have of mind-body unity andinteraction. The notion we have of the interaction between mind andbody is a kind of paradigm notion, a notion that is intelligible on its ownterms (i.e., through the closely related notion of mind-body unity), butone in terms of which at least some other seemingly distinct varietiesof causal explanation are intelligible. Two things are worth notingabout this paradigm. For one, it should be pointed out that thoughmind-body interaction is a paradigm with respect to Scholastic explanations, Descartes is unambiguous in thinking that Scholasticexplanations in terms of forms and qualities are bad explanations. TheScholastic projection of mind and mental activity onto the materialworld is an illicit projection, in Descartes’ judgment. And second, andmore important, it should be noted that although mind-body interac-tion is a paradigm for causal explanation, it is not the only paradigm, itis not universally applicable. There are, Descartes seems to claim in hisreply to Elisabeth, some causal explanations, those that involve themechanical interactions of bodies with one another, that cannot beunderstood through our understanding of mind-body interaction; ourunderstanding of voluntary action in animate beings can no more


17 This is exactly parallel to the account Descartes often gives of the common belief thatmaterial things are really red, or hot, or sweet. See, e.g., Pr I 66–71.

clarify mechanical explanations than vice versa. Or so, in any case,Descartes tells Elisabeth.

III. Motion, Impact, and God

Let us return now to the main thread of my argument. In section I Ipresented Descartes’ answer to Elisabeth’s worries about the intelligi-bility of mind-body interaction. However, I suggested there that thereis something radically wrong with the answer that Descartes gave Elisabeth; it is an answer, I claim, that goes against some of Descartes’most deeply held beliefs about the foundations of physics. Now I mustmake clear just what I have in mind.

I would like to begin by focusing in on the comparison Elisabethattempts to draw between mind-body interaction and body-body inter-action, i.e., impact. Elisabeth finds body-body interaction perfectlyintelligible. What she is asking Descartes, in effect, is to explain the onein terms of the other; she wants Descartes to explain how a nonex-tended and incorporeal mind can literally make contact with and impelan extended body. Descartes’ answer is to say that body-body and mind-body interaction are both intelligible, but on their own terms, that eachmust be comprehended through its own primitive notion, body-bodyinteraction through the notion of extension, and mind-body interac-tion through the notion of the unity of mind and body.

Let us examine these claims of Descartes’. Since we are dealing withclaims that relate to primitive notions and the notions that derive from,are comprehended through, fall under, etc., these primitive notions, wemust first inquire into how it is that the primitive notions are relatedto the less primitive notions that fall under them. Descartes, if youremember, characterizes the relation as follows:

First I observe that there are in us certain primitive notions which are, as itwere the originals [comme des originaux] on the pattern of which [sur le patrondesquels] we form all of our other thoughts [connoissances].18

Descartes is none too clear in this passage. But at very least, I think thatDescartes means to say that if a given idea Q falls under a primitivenotion P, then having P is in some sense necessary for having Q, andthat no primitive notion distinct from P is necessary for having Q. P isthe original of and pattern for Q in at least this minimal sense.


18 Descartes to Elisabeth, 21 May 1643, AT III 665.

The problem I see with Descartes’ answer to Elisabeth relates to theclaims that he seems to make about precisely what ideas fall under theprimitive notions he enumerates. Now, it is tempting to suppose thatthe real problem must arise in connecting mind-body interaction to theprimitive notion of mind-body unity that Descartes claims we have, tosuppose that Descartes’ answer must break down there if it breaks downanywhere at all. But this is not what worries me. Although Descartes’conception of mind-body unity has its obscurities, I am reasonably con-fident that one can concoct a plausible account of mind-body unity thatmakes comprehensible just why Descartes saw mind-body interaction asfalling under the primitive notion of mind-body unity.19

I certainly concede that working out this account may involveDescartes in some unforeseen difficulties. But be that as it may, theobvious problems lie not with unity and interaction, but with the primafacie more plausible account of the idea of body-body interaction as itrelates to the primitive notion of extension. Descartes’ answer seems tosuggest that impact, body-body interaction falls under the primitivenotion of extension. But does it?

In answering Elisabeth, Descartes gives only two examples of ideasthat derive from the primitive notion of body: shape, and motion. It isclear why shape is included there. Shape is a mode of extension, inDescartes’ technical vocabulary.20 And it is plausible to suppose that wecannot have an idea of a mode, like shape, without having an idea ofthe kind of substance of which it is a mode, i.e., extended substance,and that no other idea is required for us to be able to have an idea ofshape. That is, the idea of shape falls under the primitive notion ofextension in the appropriate sense.

The same case can be made for the idea of motion. Though there aresome complexities here, Descartes was clear in considering motion to bea mode of body, a mode of extension, just like shape. Descartes wrote inhis Principia in offering a formal definition of the notion of motion:


19 One might even suggest that when Descartes says that mind and body are united, thisclaim simply means that they are capable of appropriate causal interaction. See, e.g., HenriGouhier, op. cit., p. 335. For a contrary view, that the mind-body union results in a thirdsubstance, a substance over and above the mental and material substances that make itup, see, e.g., G. Rodis-Lewis, op. cit., vol. I, p. 353 and the references cited in vol. II, p. 543, note 29; or Janet Broughton and Ruth Mattern, “Reinterpreting Descartes on the Notion of the Union of Mind and Body,” Journal of the History of Philosophy 16 (1978),pp. 23–32.

20 See, e.g., Pr I 53, 61.

If we consider how motion must be understood . . . in accordance with thetruth of the matter, we must say that it is the translation [translatio] of onepart of matter, or of one body from the vicinity of those bodies which imme-diately touch it and which are regarded as being at rest, into the vicinity ofothers. . . . And I say . . . strictly speaking that it is a mode [of body], not some-thing substantial, just as shape is a mode of a thing with shape, and rest is amode of a thing at rest.21

Similarly, Descartes wrote to Henry More:

The translation that I call motion is a thing of no less entity than shape: It isa mode in a body.22

Consequently, one can say that motion is understood through the prim-itive notion of extension in roughly the same way as shape is.23

But, it should be noted, Elisabeth’s question didn’t deal with motionper se. The comparison she is attempting to press is not a comparisonbetween mind-body interaction and motion, i.e., the translation a bodyundergoes with respect to other bodies, but between the way in whicha mind can cause motion in bodies, and the way in which bodies cancause motion in other bodies. That is, the comparison is not betweeninteraction and motion, but between two purported ways of causingmotion. And while motion itself may be a mode of body, somethingcomprehended through the notion of extension, change in motion andits causes are something altogether different.

Now, how are we to understand body-body interaction, the way in whichone body can change the speed or direction of another body’s motionthrough impact? Elisabeth takes this to be intelligible in and of itself andto be in need of no further explanation. And although Descartes seems toconcur with this in his answer to her, quite a different answer emergesfrom his more careful writings on physics from early to late. A way intoDescartes’ position is through the question: What are the laws that governthe behavior and interaction of bodies, and why do bodies obey the lawsthey do? One might, as some of Descartes’ contemporaries tried to do,answer this question either through empirical studies24 or through an


21 Pr II 25. See also Pr II 27.22 Descartes to More, August 1649, AT V 403.23 Despite what Descartes says, this cannot be quite right, since motion, unlike shape, involves

time or, at very least, change.24 Though the question is hotly debated, this is at least one reading of what Galileo thought

he was up to. For this reading, see, e.g., Stillman Drake, Galileo Studies (Ann Arbor, 1970),especially Drake’s polemical introduction. For Descartes’ nutshell assessment of Galileo’sphysics, see Descartes to Mersenne, 11 October 1638, AT II 380.

analysis of the nature of body and motion.25 But for Descartes, the lawsthat govern bodies in motion and impact must derive from the causes ofmotion.

But what are the causes of motion for Descartes? He answers thisquestion in very general terms in a letter to More that I quoted earlier.Descartes writes:

The translation which I call motion is a thing of no less entity than shape: itis a mode in body. The force causing motion [vis . . . mouens] may be that ofGod Himself conserving the same amount of translation in matter as He putin it the first moment of creation; or it may be that of a created substance,like our mind; or of any other thing to which He gave the force to move a body.26

The causes of motion, then, are God, or minds.27 Now, the mental cau-sation of motion is something of great importance to Descartes, as wehave seen already. But in physics, it is the divine causation of motionthat is mostly at issue. And it is from an understanding of how Godcauses motion that the laws of motion are derived.

Descartes begins his discussion of the causes of motion and the lawsit obeys with the following statement:


25 This seems to be the strategy Thomas Hobbes adopts, e.g., in De Corpore, chapter 15. Thisis also the strategy that Leibniz sometimes attributed to his own youthful works in physics,the Theoria Motus Abstracti and the Hypothesis Physica Nova. See, e.g., Leibniz’s remarks atthe time these works were being written, in Leibniz to Oldenburg, 13/23 July 1670, inLeibniz, Sämtliche Schriften und Briefe (ed. Preussischen Akademie der Wissenschaften),series II, vol. I (Darmstadt, 1926), p. 59; or Leibniz’s later remarks on this early programin Part I of his Specimen Dynamicum (1695), in G. W. Leibniz (ed. C. I. Gerhardt), Mathe-matische Schriften, vol. VI (Halle, 1860), p. 240, translated in P. P. Weiner (ed.), LeibnizSelections (New York, 1951), p. 128. In some of his polemical writings against the Carte-sians, Leibniz gives the misleading impression that for Descartes, too, the laws of motionare to be derived from the nature of body. See, e.g., the essay that Weiner has entitled,“Whether the Essence of a Body Consists in Extension,” in Leibniz (ed., C. I. Gerhardt),Die Philosophischen Schriften, vol. IV (Berlin, 1880), pp. 464–66, translated in Weiner, op. cit., pp. 100–2.

26 Descartes to More, August 1649, AT V 403–4.27 P. H. J. Hoenen has suggested that the “other things” to which God gave the ability to

cause motion in bodies are just other bodies. See the excerpt from his Cosmologia, trans-lated as “Descartes’s Mechanicism” in Willis Doney, ed., Descartes (Garden City, 1967), pp. 353–68, esp. p. 359. But it is interesting that in the sections of Principia II that deal with the causes of motion, properly speaking, sections 36 and following, bodies arenever mentioned as genuine causes. However, in Pr II 40 Descartes does mention, in addi-tion to human minds, angelic minds as possible causes of motion. Angelic minds as causesof motion also come up in the letter Descartes wrote to More that immediately precedesthe one from which I quoted. See Descartes to More, 15 April 1649, AT V 347. This suggests that the “other things” in question in the August 1649 letter are not bodies, but angels.

After having considered the nature of motion, we must consider its cause,and that is twofold: first, indeed the universal and primary cause, which isthe general cause of all motions in the world; and then the particular cause,by which it happens that individual parts of matter acquire motions whichthey did not have before. And it seems obvious to me that the general causein question is nothing else but God Himself.28

The distinction between the universal and particular causes thatDescartes announces here makes it look as if he is dealing with a dis-tinction between a prime-mover God who is the first cause, setting theworld in motion, and other corporeal causes, which result in the worldchanging from moment to moment. But this is not the picture at all.The universal and general cause, God, not only sets the world inmotion, but preserves motion in the world; the secondary causes to whichDescartes refers, as it turns out, are not causes of motion over and aboveGod, but rather three laws in accordance with which God Himself pre-serves motion in the world from moment to moment.

In order to understand just how this works, we must remember thatfor Descartes, the world must be preserved from moment to momentby God, if it is not to pass out of existence. But since preservation andcreation are the same thing, Descartes argues, this is to say that Godmust continually re-create the world for it to persist.29 So, Descartes’God is not merely the prime mover; He is the general cause of motioninsofar as it is His continual activity, His changing of the relative placesof bodies from moment to moment while keeping them in existencethat constitutes motion in the world.30 Consequently, the laws that bodiesin motion obey must derive from the way in which God continuouslyre-creates the world. And this, indeed, is just how Descartes derivesthose laws. The first general principle Descartes notes is his famous con-servation of motion law. This law is derived from the immutability ofGod. Descartes argues that:


28 Pr II 36. See also Descartes to [the Marquis of Newcastle], October 1645, AT IV 328.29 See Meditation III, AT VII 48–49; Second Replies, AT VII 165; Pr I 21.30 The continual re-creation account of God’s activity creates a curious difficulty for the

mental causation of events in the material world. When God is re-creating the materialworld from moment to moment, He must put each material thing somewhere when He re-creates it. But if it is God who determines the position of bodies from moment to moment,how is it possible for minds to affect the momentary position of a body? There seems tobe no room for minds to act on Descartes’ continual re-creation picture. Nicolas Malebranche develops this difficulty into an argument for occasionalism in the seventhof his Entretiens sur la métaphysique.

We must understand God to be perfect not only insofar as He is immutable,but also insofar as He works with the greatest constancy and immutability.. . . Whence it follows that it is most consistent with reason that we think thatfrom this alone, that God moved the parts of matter in different ways whenHe first created them, and now conserves all that matter in the same way andfor the same reason He created it before, that He would also conserve thesame amount of motion in it always.31

This is Descartes’ “master law” of motion. But the secondary laws arealso derived, as the master law was, from God’s activity. Descartes writes:

And from this same immutability of God, certain rules or laws of nature canbe understood, which are secondary and particular causes of the differentmotions which we notice in individual bodies.32

The dependence of the first two of the secondary laws on God’simmutability as a cause of motion is evident. These laws, the so-calledCartesian laws of inertia (laws of persistence would be more accurate)mandate that certain states in bodies, the state of motion itself in thefirst law, and the state of moving in a particular direction in the second,persist. These follow directly from the immutability of God, who,Descartes writes, “preserves motion precisely as it is in that very momentof time in which he conserves it.”33

The third law, the law dealing with impact and the way in which onebody can change the state of another body, is somewhat more difficult.In order to continue to argument, Descartes must argue that theimmutability of God requires that He change the motion of a given bodyunder certain circumstances, when, for example, it is hit by anotherbody of appropriate size and speed (force of going on). And this,indeed, is how Descartes argues. The intuition is this. The fact thatthere is no space devoid of body,34 together with the fact that Godcreated a world of bodies in motion35 entails that if God is to preservemotion in the world, as His immutability requires, He must change themotion of at least some bodies as they encounter one another. ThusDescartes writes in the Principia, in defense of his law of impact:

All places are filled with body, and at the same time the motion of every bodyis rectilinear in tendency; so clearly, when God first created the world, Hemust not only have assigned various motions to its parts, but also have causedtheir mutual impulses and the transference of motion from one to another;and since He now preserves motion by the same activity and according to the


31 Pr II 36. 32 Pr II 37. 33 Pr II 39. 34 Pr II 5–19. 35 Pr II 36. Pr III 46–47.

same laws, as when He created it, he does not preserve it as a constant inher-ent property of given pieces of matter, but as something passing from onepiece to another as they collide. Thus the very fact that creatures are con-tinually changing argues for the immutability of God.36

Descartes’ reasoning here is hardly a model of clarity and distinctness.But at least this much is clear: For Descartes, impact and the changes inbodily motion that result from impact are nothing but the changes thatGod must make in re-creating the world from moment to moment inorder to accommodate the motion of bodies to one another. Strictlyspeaking, bodies in motion are not real causes of change in impact, itwould appear; motion transferred, motion begun, and motion ended inimpact must derive from God himself, shuffling bodies about as part ofthe process of “conserving the same amount of translation in matter asHe put in it the first moment of creation,” as he wrote to More.37

(Here, by the way, is the reason why body-mind causality must beproblematic for Descartes, as I suggested earlier. The picture one getsfrom the physics is one of inert matter being shuffled around frommoment to moment by an active God and, from time to time, by activeincorporeal minds. But given the inertness of matter on this picture, inwhat sense can one say that the body can cause changes in mentalstuff?)

The discussion of the last few pages has taken us a bit out of our way.I started with the claim that Descartes seems to make to Elisabeth, thatbody-body causation must be understood through the primitive notionof extension. I claimed that while this may be true of motion simpliciterwhich is, indeed, a mode of body, the case of body-body interaction orimpact is more complex, at least as analyzed in Descartes’ writings onphysics. An account of impact led us from motion simpliciter to its causes,to God and the way in which He acts on the world in shuffling bodiesabout from moment to moment. So, it seems, a full understanding ofbody-body interaction requires that we understand not only motion, amode of extension, but the way in which God acts on the world. Butunder which of Descartes’ three primitive notions does this fit?

Descartes never takes this question up in quite those terms. But avery similar question does arise in the all too brief correspondence withMore at the end of Descartes’ life. One of More’s deepest criticisms ofDescartes concerns the doctrine that the essence of material substance


36 Pr II 42. 37 Descartes to More, August 1649, AT V 403–4.

is extension. More argues that material substance is not mere exten-sion, but tangible or impenetrable extended stuff. As part of his attack,More makes the claim that spirits and even God are extended.38 In thecase of God, More argues:

Now, the reason why I judge that God is, in His way, extended is that He isomnipresent and intimately fills the whole machine of the universe and eachof its individual parts. For how could He imprint motion on matter, whichHe once did, and which He actually does now, according to you [i.e.,Descartes], unless He now as it were touches the matter of the universe, orat least once did? . . . God is thus in His way extended, and consequently, Godis an extended thing.39

Descartes’ answers to More’s general attack are quite interesting, andbear interesting relations to his responses to Elisabeth’s general worriesabout how incorporeal substances can move extended bodies.40 But mostinteresting is his answer as to how we can conceive of a nonextendedGod as being able to act on an extended world. Descartes writes:

It is no disgrace to a philosopher to believe that God can move a body, withoutregarding God as corporeal; it is no more of a disgrace to Him to think thesame of other incorporeal substances. Of course I do not think that any modeof action [modus agendi] belongs univocally to both God and creatures, but Imust confess that the only ideal I can find in my mind to represent the way[modus] in which God or an angel can move matter is the one which showsme the way in which I am conscious I can move my own body by my ownthought.41

This comes as close as one could like to answering my question. Theway God acts upon the world in sustaining motion and rearrangingbodies in impact must, it seems, be derived from the conception I haveof how I act upon my body; it, too, must be derived from the primitivenotion of the unity of mind and body. Descartes’ answer to Elisabeth,


38 See, e.g., More to Descartes, 11 December 1648, AT V 238–40; More to Descartes, 5 March1649, AT V 301; More to Descartes, 23 July 1649, AT V 379.

39 More to Descartes, 11 December 1648, AT V 238–39. This seems similar to a point Spinozamakes in defense of his claim that God must have the attribute of extension. See EthicsI, prop. 15, scholium, in Spinoza (ed. Carl Gebhardt), Opera (Heidelberg, 1925), vol. II,p. 57.

40 Compare, e.g., the discussion of the sense in which God is extended in potentia in theletters to More (Descartes to More, 15 April 1649, AT V 342; Descartes to More, August1649, AT V 403) with Descartes’ remarks to Elisabeth about the sense in which it is properto say that mind is extended (Descartes to Elisabeth, 28 June 1643, AT III 694).

41 Descartes to More, 15 April 1649, AT V 347.

thus, cannot have been the correct answer, the answer that he shouldhave given, on his own principles. Body-body interaction is not fullyintelligible under the primitive notion of extension. A full under-standing of bodies in impact, of how one body can alter the motion ofanother, requires that we understand how God acts on the world. Andthis, in turn, requires that we be familiar with the way our minds actupon our bodies. So, if there is something wrong with the comparisonthat Elisabeth tries to draw between mind-body and body-body interac-tions, it cannot be what Descartes says it is; it cannot be an illicit inter-mingling of discrete primitive notions. For the same primitive notion isultimately involved with both.

IV. What Descartes Should Have Told Elisabeth

The argument of the previous section undermines Descartes’ answer to Elisabeth. Elisabeth’s attempt to understand interaction throughimpact is not wrong for the reason Descartes says it is; Elisabeth is notconfusing concepts that fall under different primitive notions. Thismuch is clear. But the most interesting question still remains to befaced: how does this observation affect the claim for which Descartes istrying to argue? In responding to Elisabeth, Descartes is attempting toestablish that mind-body interaction is per se intelligible, or, at least,intelligible through the closely related notion of mind-body unity, andthat Elisabeth’s attempt to connect mind-body interaction with body-body interaction is neither possible nor needed. I have shown that theargument he offers for these claims through the doctrine of the threeprimitive notions is not, on Descartes’ own terms, correct. But whatbecomes of the claims themselves? Ironically enough, I think that myCartesian refutation of Descartes’ actual response to Elisabeth, if any-thing, strengthens his position. The considerations concerning motionand impact drawn from Descartes’ writings on the foundations ofphysics suggest a line of defense for the claims in question which ismore consistent with the rest of his works than the one he offered toElisabeth, and which is, I think, philosophically stronger than the onehe actually used. I should point out here that I make no claim thatDescartes ever used, or even saw the argument that I will try to developin this section. All I claim is that it is an argument he could have used,and perhaps, should have used.

Let me begin setting out this new and improved answer to Elisabethby recalling the earlier discussion of the analogy that Descartes appealsto in explaining his position, the analogy with the Scholastic account of


heaviness. I pointed out that what allows Descartes to use that analogy ishis claim that in this case we are not dealing with two different notions,but only one. The claim is that the Scholastic account of heaviness is com-prehensible because it involves a projection of our composite natureonto the inanimate world. The real quality of heaviness is thought of asa kind of mind, united to the heavy body in just the way that the humanmind is united to the human body, and, it is claimed, we conceive of heav-iness acting on the heavy body in drawing it to the center of the earth injust the way we conceive of the mind acting on the body. Thus, theScholastic mode of explanation is parasitic on the idea we have of mind-body interaction in the sense that if we didn’t understand how mindsacted on animate bodies, then we wouldn’t understand how forms orqualities act in inanimate bodies. Furthermore, one can, perhaps, saythat the notion we have of mind-body interaction is a paradigm notionwith respect to the Scholastic account of heaviness, and, more generally,with respect to all Scholastic explanations in terms of form and quality,insofar as our understanding of these modes of explanation involves aprojection of our notion of mind-body interaction onto the world ofinanimate things.

The discussion of motion and impact suggests that something similarcan be said about the relation between mind-body interaction and themechanical conception of explanation in terms of impact. Now, it istrue that the notion of mind-body interaction is not a paradigm notionwith respect to impact in quite the same way as it is with regard to theScholastic conception of heaviness. While the notion of mind-bodyinteraction does enter into a full understanding of interaction, it is nota simple projection of our composite nature onto the inanimate world,as the Scholastic theory is. The notion of mind-body interaction entersin at only the deepest level of analysis of the notion of impact, when weattempt to understand how God, the first and continuing cause ofmotion in the world, the real cause for the changes in the motion ofbodies in impact, can act upon the material world. Consequently,impact cannot be used as Descartes tries to use the Scholastic theory ofheaviness, to call attention to the idea of interaction he claims we allhave. But, the notion we have of impact is like the notion we have ofthe Scholastics account of heaviness in an important respect. Elisabeth,like most of her contemporaries, at least those sympathetic to the newmechanist science, took impact to be per se intelligible, in fact, the verymodel of intelligibility. What Descartes’ analyses of motion and its lawspurport to show is that this is not so. A full understanding of motionin the material world requires reference to God and His action on the


material world, and through this, requires reference to our mind’saction on our bodies. In this way we can say that the notion of impact,like the Scholastic notion of heaviness, is parasitic on the notion wehave of mind-body interaction; for impact as for the Scholastics’ heav-iness, mind-body interaction is a notion without which the notion ofbody-body interaction is, strictly speaking, unintelligible, despiteappearances to the contrary. And though mind-body interaction is notparadigmatic in the easy and obvious way that it is with respect toScholastic science, a full understanding of body-body interactionrequires an appeal to the way our minds can move our bodies.

This suggests an interesting line of defense for Descartes’ positionon the intelligibility of mind-body interaction. Mind-body interactionseems to be, for Descartes, a paradigm for both mechanist and Scholas-tic causal explanation. Since there were the two main competitors atthe time, we can say that, for Descartes, mind-body interaction is theparadigm for all causal explanation, it is that in terms of which all othercausal interaction must be understood. And in this there lies a defensefor the intelligibility of interaction altogether different from the one,based on the doctrine of the three primitive notions, that he offered toElisabeth. Mind-body interaction must be basic and intelligible on itsown terms since if it were not, then no other kind of causal explanationwould be intelligible at all; to challenge the intelligibility of mind-bodyinteraction is to challenge the entire enterprise of causal explanation.Furthermore, we cannot give a simpler or more easily understoodaccount of causal interaction than mind-body interaction because thereare no more basic or more inherently intelligible ways of explaining the behav-ior of anything open to us. We cannot appeal to analogies with impact toclarify mind-body interaction, as Elisabeth does, not because of any con-fusion of primitive notions, but because we must work the other way:body-body interaction must ultimately be understood through thenotion we have of the way in which the mind acts on the body.

I should repeat that, despite suggestions of a position like this in thewritings on motion, Descartes never said anything like this, to the bestof my knowledge. But it is a philosophically interesting answer, one thatis open to him and, I think, more consistent with his conception ofcausal interaction in the physical world than the account that he actu-ally offered. It is, I think, the way of understanding interaction thatDescartes should have offered Elisabeth.


9

HOW GOD CAUSES MOTION

Descartes, Divine Sustenance, and Occasionalism

In his Traité de l’esprit de l’homme (1664),1 Louis de La Forge, one ofDescartes’ early followers, wrote:

I hold that there is no creature, spiritual or corporeal, that can change [theposition of a body] or that of any of its parts in the second instant of its cre-ation if the creator does not do it himself, since it is he who had producedthis part of matter in place A. For example, not only is it necessary that hecontinue to produce it if he wants it to continue to exist, but also, since he cannot create it everywhere, nor can he create it outside of every place,he must himself put it in place B, if he wants it there, for if he were to haveput it somewhere else, there is no force capable of removing it from there.(Traité, p. 240)

De La Forge’s argument is an interesting one. He begins with twopremises. The first is the doctrine of divine sustenance, that God mustsustain the existence of every body, indeed, of every thing, mind orbody, at every moment of its existence. Second, de La Forge assumesas a result, it would seem, that God causes motion in the material worldby re-creating bodies in different places at different times. From this deLa Forge draws the conclusion that only God can move a body. WhenGod sustains bodies, He must sustain them in some place or other; Hecannot sustain them everywhere, nowhere, or in any way independentlyof some place or other. And so causes of motion beside God, causes ofmotion as our own minds are supposed to be, are neither possible nor

189

1 Pierre Clair, ed. Louis de La Forge: Oeuvres Philosophiques (Paris: Presses Universitaires deFrance, 1974). A similar argument is also found in dialogue seven of Nicolas Malebranche,Dialogues on Metaphysics, trans., Willis Doney (New York: Abaris Books, 1980).

needed; if motion and rest are direct results of God’s sustenance of thematerial world, it would seem that there can be no room for othercauses.

The position de La Forge is trying to establish here is a variety ofoccasionalism, and the argument I have sketched is one among manywhich Descartes’ followers used to establish the claim that God is theonly genuine cause in the material world, at least.2 On this view, causalrelations between two bodies, or between a mind and body, are not truecausal links, but only occasional causal links which depend for their effi-cacy on God actually to impart the appropriate motion to the appro-priate body. What is especially interesting is that de La Forge starts fromwhat many commentators assume to be genuinely Cartesian doctrinesto establish his conclusion. Descartes emphasizes in a number of placesthat “we have no force through which we conserve ourselves,” and sofor this we must turn to God, who “continually reproduces us, as it were,that is, conserves us” (Pr I 21).3 Descartes appeals to this doctrine ofdivine conservation in proving his laws of nature, both in Le Monde andin the Principia Philosophiae, arguing that God is the first and continu-ing cause of motion in the world, and that acting with constancy in pre-serving His material creation, He must necessarily sustain the world insuch a way that certain general constraints on motion are satisfied;quantity of motion is thus conserved, as is motion along a straight path(Pr II 36–42). The close connection between God’s sustenance of theworld and His role as cause of motion in the inanimate world have leda number of commentators to see something like de La Forge’s view inDescartes, the view that God’s role as a cause of motion in the world isinseparable from His role as a sustainer of the world, that God causesmotion by creating bodies in different places at different times.4

De La Forge’s premises seem to belong to Descartes as well. But, ifso, then it would appear that, like it or not, Descartes too must be com-


2 For a brief account of occasionalism among seventeenth century Cartesians, see chapter 5of Jean-François Battail, L’Advocat philosophe Géraud de Cordemoy (1626–1684) (The Hague:Martinus Nijhoff, 1973). There are a number of varieties of occasionalism. Here I am onlyconcerned with the claim that God is the only genuine cause of motion in the materialworld.

3 The numerous references to Descartes’ texts will be given in the body, for the most part.4 See, e.g., Jean Wahl, Du rôle de l’idée de l’instant dans la philosophie de Descartes (Paris: Felix

Alcan, 1920); Martial Gueroult, “The Metaphysics and Physics of Force in Descartes,” inStephen Gaukroger, ed., Descartes: Philosophy, Mathematics and Physics (Sussex: HarvesterPress, 1980), pp. 196–229, esp. 218–220; G. Hatfield, “Force (God) in Descartes’ Physics,”Studies in History and Philosophy of Science, X (1979): 113–140, esp. 127.

mitted to de La Forge’s conclusion that God can be the only cause ofmotion in His material world, that, contrary to our “most certain andmost evident experience,” mind cannot really cause motion in theworld (AT V 222). This is the question I would like to examine in thisessay. In the end, I shall argue that, when we understand Descartes’ doc-trine of divine sustenance and of the way God enters the world as acause of motion, we shall see that, wherever de La Forge’s views leadhim, Descartes need not be committed to occasionalism, at least not inthis way. When we understand just how God causes motion, we shall see that Descartes’ God can leave plenty of elbow room for other causesto produce their effects, indeed, produce them as directly as GodHimself does.

I

It will be helpful to begin the story with a brief discussion of Descartes’doctrine of divine sustenance. Descartes writes in Meditation III:

All of the time of my life can be divided into innumerable parts, each of whichis entirely independent of the others, so that from the fact that I existed ashort time ago, it does not follow that I ought to exist now, unless some causeas it were creates me again in this moment, that is, conserves me. [AT VII 49(CSM II 33)]

Now, Descartes argues, “plainly the same force and action is needed toconserve any thing for the individual moments in which it endures aswas needed for creating it anew, had it not existed” [AT VII 49 (CSMII 33)]. Clearly such a power is not in us; if it were, Descartes reasons,I would also have been able to give myself all the perfections I clearlylack [AT VII 168 (CSM II 118)]. And so he concludes that it must beGod that creates and sustains us [AT VII 111, 165, 168, 369–70 (CSMII 80, 116, 118, 254/5); Pr I 21]. This conclusion, of course, holds forbodies as well as for us. It is not just souls, but all finite things thatrequire some cause for their continued existence. And, as with the ideaof ourselves, “when I examine the idea of body, I perceive that it hasno power [vis] in itself through which it can produce or conserve itself”[AT VII 118 (CSM II 84); cf. AT VII 110 (CSM II 79)]. And so we mustconclude that the duration of bodies, too, must be caused by God, whosustains the material world He created in the beginning.

Descartes conceives of God’s continual sustenance of his creaturesas their efficient cause: “I should not hesitate to call the cause that

how god causes motion 191

sustains me an efficient cause” [AT VII 109 (CSM II 79)]. But God’scausality here is in one respect importantly different from other effi-cient causes that we are familiar with from our experience. In reply toGassendi’s Fifth Objections, Descartes distinguishes between two sorts ofefficient causes, a causa secundum fieri, a cause of becoming, and a causasecundum esse, a cause of being. Roughly speaking, as Descartes under-stands the notions, a causa secundum esse is a cause which must continueto act for its effect to continue, unlike a causa secundum fieri, which pro-duces an effect that endures, even after the cause is no longer in oper-ation or even in existence. An architect, thus, is the cause of becomingwith respect to a house, as is a father with respect to his son. ButDescartes claims

the sun is the cause of the light proceeding from it, and God is the cause ofcreated things, not only as a cause of becoming, but as a cause of being, andtherefore must always flow into the effect in the same way, in order to con-serve it. [AT VII 369 (CSM II 254/5)]

And, so just as we ordinarily think that the sun must continue its illu-mination for daylight to persist, so must God continue His activity inorder for the world and its motion to be sustained.5 This continual sus-tenance is also unlike the more ordinary efficient causes insofar as itrequires a kind of power beyond the capacities of created things.Whereas finite things may be able to stand as the efficient causes secun-dum fieri of things in the world, only God, strictly speaking, can standas their cause secundum esse. As we noted earlier, in Meditation IIIDescartes declares that: “plainly the same force and action is needed toconserve any thing for the individual moments in which it endures aswas needed for creating it anew, had it not existed” [AT VII 49 (CSMII 33)]. From this Descartes infers that “it is also one of those thingsobvious by the light of nature that conservation differs from creationonly in reason” [AT VII 49 (CSM II 33)]. That is, the activity and powerneeded to sustain a thing in its existence is identical to the activity andpower necessary to create anything from nothing [cf. also AT VII 165,166 (CSM II 116, 117)]. Elsewhere he puts the point a bit differently,suggesting that conservation is to be understood as the “continual pro-duction of a thing” [AT VII 243 (CSM II 169); cf. Pr II 42], or, more


5 Descartes does concede, under challenge, that the sun may not be an especially goodexample of a causa secundum esse. See AT III 405, 429.

guardedly, suggesting that God as it were (veluti) continually reproducesHis creatures [Pr I 21; cf. AT VII 110 (CSM II 79)].

In the following section, we shall investigate how Descartes’ Godcauses motion while sustaining the world. But, before turning to thatquestion, I would like briefly to discuss an issue closely related to thequestions under discussion here, that of temporal atomism. A numberof commentators take Descartes’ language quite literally when he saysthat God must continually re-create His creatures. On their view, Carte-sian time must, as a result, be a series of discrete timeless instants, createdone after another like the frames of a motion picture.6 Such a view seemsinevitably to lead to a position like de La Forge’s. The cartoonist creat-ing an animated cartoon can cause his creatures to move only by drawingthem in different positions in successive frames; so too for God, it wouldseem, were we to conceive of Him as the grand cartoonist with respectto His creation. In this way, God’s sustenance would seem to be insepa-rable from His role as cause of motion, and all genuine causes of motionother than God would seem to be frozen out.

But it is not at all clear that Descartes held such a position. In a recentstudy, Jean–Marie Beyssade7 has argued that Descartes’ God sustains thecontinuously flowing time of our experience. On Beyssade’s view, timefor Descartes is much like body, infinitely divisible and not composedof any ultimate elements, elements such as the durationless temporalatoms are supposed to be. Beyssade does not deny, of course, thatDescartes is concerned with timeless instants in a number of importantcontexts, and, indeed, that he even talks about God conserving bodiesas they exist at a given instant [AT XI 44 (CSM I 96); Pr II 39]. But,Beyssade argues, such instants are not, strictly speaking, parts of dura-tion. A hunk of extended substance can be divided into innumerableparts. But, for these divisions to be genuine parts of a body, they mustbe extended as well. Points, lines, surfaces, and geometrical objects thatlack extension in length, width, and breadth, are not parts of a body,but limits or boundaries. So, Beyssade suggests:

In the same way, every duration or part of duration contains a before andafter . . . ; the instant is its limit or boundary. If we are not mistaken, Descartesalways takes this word [“instant”] and its Latin original “instans” in the strictsense of a limit. (La philosophie première, p. 348; cf. p. 353)


6 See, e.g., the references cited in note 4.7 La philosophie première de Descartes (Paris: Flammarion, 1979), pp. 129–143.

Durations, no matter how small, can be parts of an enduring world, andthus can be candidates for God’s sustaining activity. But, although theremay be instants in duration as boundaries of finite durations, instants,Beyssade suggests, cannot be parts of an enduring world; they cannotcompose durations, nor can we intelligibly talk about God creating asingle instant by itself without creating the duration it serves to bound,any more than we can talk about God creating a two-dimensionalsurface, a mode of body, without the body that it bounds [AT VII 250/1,433 (CSM II 174, 292)].

With this in mind, it is easy to see that there is really nothing inDescartes’ texts that unambiguously implies temporal atomism. Theidea that all the parts of time are independent, the view we saw earlierin Meditation III, certainly does not; the parts of time in question theremight plausibly be read as genuine parts of time, parts with duration,parts which are independent in the sense that God could create anystretch of time without creating preceding or succeeding portions oftime. One can give similar readings to other passages in which Descartestalks about the independence of the present time from other momentsor moments from one another. Even where Descartes talks of creatingthe things anew at every moment, even where Descartes makes it clearthat God sustains things as they are in a timeless instant, there is noneed to attribute temporal atomism to him.8 To say that God re-createsthe world at every instant is to say that every instant can be regardedas the beginning, as the boundary of a newly created world. But,although every instant can be regarded as a moment of creation, it doesnot follow that what is being created is a bare instant or a sequence ofbare instants, or that God could create an atemporal instant withoutcreating a duration for that instant to bound.

But, just as Descartes was not committed to temporal atomism,neither was he committed to its denial; I know of few passages thatcannot be plausibly interpreted either way. Indeed, I know of nopassage to suggest that Descartes was particularly interested in the ques-tion of temporal atomism, one way or the other. And so it seemsimproper to argue from Descartes’ supposed temporal atomism to theclaim that God causes motion through re-creating bodies in differentpositions at different times. If we want to know how God causes motionfor Descartes, we should face the question directly.


8 For the former formulations see AT VII 49, 109 (CSM II 33, 78/9); for the latter see ATXI 44, 45 (CSM I 96/7), Pr II 39.

II

In presenting his account of God as continual sustainer of the world,Descartes did not think he was telling his readers anything they had notalready heard. As far as he was concerned, he was appealing to an oldand widely accepted doctrine with which his audience could beexpected to be both familiar and generally sympathetic. When in theFifth Objections Gassendi challenged his appeal to a conserving God [ATVII 300 (CSM II 209)], Descartes responded: “When you deny that tobe conserved we require the continual influx of a first cause, you denysomething that all metaphysicians affirm as obvious” [AT VII 369 (CSMII 254); cf. AT VI 45 (CSM I 133)]. And, in defending himself againstGassendi’s criticisms, he seems to have turned directly to his copy of St.Thomas Aquinas.9 God’s sustenance of this world of created things isexplicitly discussed in the Summa Theologiae I, q 104, a 1, and thispassage may be the source of Descartes’ answer to Gassendi. LikeDescartes, Aquinas distinguishes between causes secundum fieri andsecundum esse, and appeals to the same examples Descartes does – thebuilder of a house, the parent of a child, and the sun as illuminator –to clarify the sense in which God is the cause of the world as enduring(ST I, q 104, a 1 c). And, although Aquinas does not say exactly thatGod’s activity in sustaining the world is identical with His activity in cre-ating it, many of Descartes’ contemporaries would have been happy toagree with Descartes that “this conservation is the very same thing ascreation, differing only in reason.”10

In his monumental commentary on the Discours, Étienne Gilson11

noticed this similarity between Descartes and Aquinas. But Gilson alsonoted an apparent (and important) difference:

The being of the things Descartes’ God conserves is so different from thatwhich St. Thomas’s God conserves, that there is a profound difference


9 In December 1639, Descartes tells Mersenne that he owns “vne Somme de S. Thomas,”though it is not altogether clear to me whether this means a copy of Aquinas’s SummaTheologiae or a summary of Aquinas. See AT II 630.

10 Guilelmus Amesius, Medulla theologica (1628), quoted in Heinrich Heppe and Ernst Bizer,ed., Die Dogmatik der evangelisch-reformierden Kirche (Neukirchen, kreis Moers: Buchhand-lung des Erziehungsvereins, 1935), p. 208. For other of Descartes’ contemporaries on the question, see Heppe, Dogmatik, Locus XII; G. T. Thomson, Reformed Dogmatics, trans.,Heinrich Heppe (London: Allen and Unwin, Ltd., 1950), chapter XII; Étienne Gilson,Index Scholastico-Cartesien (Paris: J. Vrin 1979), §§64, 112.

11 Discours de la méthode: texte et commentaire (Paris: J. Vrin, 1967).

between their two notions of continual creation. The Thomist God conservesthe being of a world of substantial forms and essences. . . . But, on the con-trary, in Cartesianism, there are no substantial forms any more. (Commentaire,p. 341)

Gilson goes on to argue that, lacking substantial forms, Descartes,unlike the Scholastic, is doomed to a movie-show world of still frames,mocking the continuity of time and motion that the Scholastic is gen-uinely entitled to. Gilson’s full argument is too complex to enter intohere. But I would like to explore his initial observation a bit.

Descartes does, for the most part, reject substantial forms and thisdoes indeed make a difference, as Gilson emphasizes. But what differ-ence it makes depends on how the notion is understood, and what it isthat takes the place of the absent forms. Now, the notion of a substan-tial form is a basic notion in Aristotelian thought, and there are impor-tant differences of conception among Scholastic thinkers with regardto that notion. But, to understand the importance of the rejection ofsubstantial forms to Descartes’ thought, we must begin with an accountof what the notion meant to him.

In very general terms, a substantial form is that which, joined tomatter (the materia prima of the Scholastics, ultimately) results in a com-plete substance. But, more substantively, substantial form is that fromwhich the characteristic behavior of the various sorts of substancesderives. And, so Descartes notes, writing to his then disciple HenricusRegius in January 1642, “they [i.e., forms] were introduced by philoso-phers to explain the proper action of natural things, of which actionthis form is the principle and the source” (AT III 506). Or, as the Con-imbrian Fathers wrote in a book Descartes likely learned from as aschoolboy,

There are individual and particular behaviors [ functiones] appropriate toeach individual natural thing, as reasoning is to human beings, neighing isto horses, heating to fire, and so on. But these behaviors do not arise from matter. . . . Thus they must arise from substantial form. (Gilson, Index,§209)

More concretely, Descartes views substantial forms as substances of asort: “By the name ‘substantial form’ I have understood a certain sub-stance joined to matter, and with it composing something whole that ismerely material” (AT III 502). And the sorts of substances they are ismental substance, Descartes thinks, “like little souls joined to their


bodies” (AT III 648).12 And so Descartes characterizes the Scholasticaccount of heaviness which he himself once held as follows:

But what especially showed that the idea I had of heaviness was derived fromthat of mind was the fact that I thought that heaviness bore bodies towardthe center of the earth as if it contained in itself some knowledge of it [i.e.,the center of the earth]. For this could not happen without knowledge, andthere cannot be any knowledge except in a mind. [AT VII 442 (CSM II 298)]

If substantial forms are supposed to explain the characteristic behaviorof bodies of various sorts, then we must be thinking of them as inten-tional entities, agents of a rudimentary sort, things capable of formingintentions and exercising volition, little souls joined to matter. And sothe Scholastic doctrine of form and matter is, in a sense, just the imageof the Cartesian human being, an unextended soul united to extendedbody and projected out onto the material world. Indeed, Descartes oftenuses the supposed familiarity of the Scholastic model of heaviness (whicheveryone would have learned at school) to persuade those who havetrouble with mind–body interaction on his view that they already under-stand how interaction is possible; if one can understand the Scholasticaccount of heaviness, then one can understand how the soul can movethe body, Descartes reasons, since the two cases are just the same.13

It should be evident from Descartes’ account of substantial form thathe does not reject forms altogether. Given that the human mind is thevery model of a form, it is not surprising to find Descartes saying fromtime to time that the human soul is “the true substantial form of man”(AT II 505); indeed, it is “the only substantial form” he recognizes [ATIII 503; cf. AT IV 346, AT VII 356 (CSM II 246)]. Descartes from timeto time also uses Scholastic terminology and talks of the soul “inform-ing” the body [AT IV 168; AT X 411 (CSM I 40)]. In this sense, Gilsonperhaps overestimates the difference between the world Descartes’ Godsustains and that which Aquinas’s sustains, insofar as both contain atleast some substantial forms.14


12 The remark in question relates to real qualities, strictly speaking, qualities that followdirectly from forms. But, in his polemics against the Scholastics, Descartes drew no dis-tinction between substantial forms and real qualities.

13 For a fuller account of this, see §II of Daniel Garber, “Understanding Interaction: WhatDescartes Should Have Told Elisabeth,” Southern Journal of Philosophy, XXI supp. (1983):15–32, essay 8 in this volume.

14 For a recent discussion of the human soul as substantial form in Descartes, see MarjorieGrene, “Die Einheit des Menschen: Descartes under den Scholastikern,” Dialectica, XL(1986): 309–322.

But it is significant that Descartes’ world has many fewer forms thanAquinas’s does, that Descartes rejects all forms but those which pertainto human beings.15 This raises something of a problem for Descartes,however. The substantial form was that in terms of which the charac-teristic behavior of a body of a certain sort was to be explained. But,without form, what is to explain why horses neigh and fire heats, whycannon balls fall and smoke rises? In one sense, the replacement forexplanation in terms of form is explanation in terms of size, shape, andmotion – mechanical explanation. Indeed so, but the story does notend there. In order to explain the behavior of a body (say a cannonball) mechanistically, we must know more than just the size, shape, andmotions of its parts and the surrounding medium; we must also knowthe relevant laws of motion, how a body as such can be expected tobehave, what results when two bodies of given sizes, shapes, and motionsencounter one another in collision, etc. Descartes replaces the multi-plicity of Aristotelian substances, each with its own form and distinctcharacteristic behavior, with one kind of body which fills the entire uni-verse and behaves everywhere in accordance with the same laws (cf. PrII 23). But, in the absence of Scholastic substantial forms, Descartesmust find some way of explaining the characteristic behavior of mater-ial substance, the laws of motion. And it is here that God enters as the“universal and primary [cause of motion], which is the general causeof all motions there are in the [physical] world” (Pr II 36). God is thecause of motion, what takes the place of the Scholastic forms Descartesbanished from the inanimate world of nonhuman beings.

But this, of course, leads us back to the question I posed earlier inthis essay: How does God cause motion in the world? And how is God’srole as cause of motion related to His role as sustainer of body?

To answer this first question, we must, I think, reflect on how soulsand the other forms Descartes attributed to the Scholastics werethought to cause motion. We must keep in mind here that the issue isunder a cloud, so to speak, and it may turn out that, because of an argu-ment like the one de La Forge gave, Descartes is not entitled to holdthat the human soul causes motion. But, prima facie and despite thedoubts of a number of his readers, Descartes certainly thought the ques-tion relatively unproblematic. Writing to Arnauld on 29 July 1648,Descartes noted:


15 For the still standard account of Descartes’ rejection of forms, see Gilson, Études sur le rôlede la pensée médiévale dans la formation du système cartésien (Paris: J. Vrin, 1975), pp. 141–190.

That the mind, which is incorporeal, can impel a body, is not shown to us byany reasoning or comparison with other things, but is shown daily by the mostcertain and most evident experience. For this one thing is among the thingsknown per se, which we obscure when we try to explain through other things.(AT V 222)

The mind can cause motion in a body, Descartes holds; it is somethingwe know through experience directly, something we cannot explain inother terms. Insofar as we comprehend it, it is because “we have withinus certain primitive notions, on the model of which we form all ourother knowledge,” Descartes explains to Elisabeth (AT III 665). And,Descartes goes on to explain, we understand the schoolman’s substan-tial forms to work in exactly the same way; indeed, as I noted earlier,the understanding we have of forms derives from the notions we weregiven to understand how human mind works on body.

Descartes is not very informative about just how mind (or form)moves body; on his view, there is not much that can be said, other thanto direct our attention to the experience we all have that is supposedto make it all clear.16 But, although there is not much we can say, thereis no confusing the sense in which mind causes motion in a body withthe way God sustains the body that mind supposedly moves. One of theaxioms Descartes uses in the geometrical presentation of his argumentsappended to the Second Replies reads as follows: “It is greater to createor conserve a substance, than it is to create or conserve the attributesor properties of a substance” [AT VII 166 (CSM II 117)]. The passageis not without its difficulties.17 But the clear sense is that Descartes wantsto distinguish causes that change the modes or properties of a thing(modal causes, as I shall call them) from causes that create or sustainthe very being of a substance (substantial causes, perhaps). God, sus-taining the world, is clearly a substantial cause. But minds are clearlynot; insofar as they cause changes in the motion of bodies, they at bestcan count as modal causes. And, insofar as substantial forms are under-stood on the model of souls acting on bodies, Descartes would have hadlittle trouble classifying them, with minds, as modal causes; they are


16 This, in any case, is what he insisted on in writing to Elisabeth. See AT III 663–668,690–695. What exactly he meant here is not entirely clear.

17 The passage raises an obvious question about the relation between an attribute and a sub-stance, a question the young Burman raised to Descartes in coversation. See AT V 154,and trans. and ed., John Cottingham, Descartes’ Conversation with Burman (New York:Oxford, 1976), pp. 15, 77–80.

causes of motion, a mode in bodies assumed to be sustained by thedivine Sustainer who is the unique substantial cause.

God enters Descartes’ physics to do the business substantial forms didin the Aristotelian system, as he understood it, to cause bodies to behavein their characteristic ways. And, I claim, when doing the business offorms, Descartes’ God is understood to cause motion in just the wayforms were taken to do it, that is, on Descartes’ account, in just the waythat we do it: by way of an impulse that moves matter in a way that wecan comprehend only through immediate experience. This is not at allclear as late as 1644 when, in proving his laws of nature in the Principia,Descartes’ account of God as cause of motion is deeply (and obscurely)intertwined with his account of God as sustainer of the world (Pr II36–42). But, by April 1649, Descartes wrote to Henry More:

Although I believe that no mode of acting belongs univocally to God and toHis creatures, I confess nevertheless, that I can find no idea in my mind whichrepresents the way in which God or an angel can move matter, which is dif-ferent from the idea that shows me the way in which I am conscious that Ican move my own body through my thought. (AT V 347)

And, so Descartes suggests to More, God is conceived to move bodiesin just the way we do, using the same primitive notion we use to under-stand how we move our own bodies.

If this is how we conceive of God as a cause of motion, then, it wouldseem, we are conceiving of Him as a modal cause when it comes tomotion. Conceived as such, there would appear to be a distinctionbetween God as sustainer of the world, a substantial cause keepingthings in existence, and God as cause of motion, a modal cause causingbodies to have the particular motion they have, determining, at least inpart, their modes. The difference between these two roles God playsfor Descartes comes out again in the correspondence with More. In hisletter of 5 March 1649, More asked Descartes if “matter, whether weimagine it to be eternal or created yesterday, left to itself, and receiv-ing no impulse from anything else, would move or be at rest?” (AT V316). Descartes answers in his letter of August 1649: “I consider ‘matter,left to itself, and receiving no impulse from anything else’ as plainlybeing at rest. But it is impelled by God, conserving the same amountof transference in it as He put there from the first” (AT V 404). Thepicture that comes through here is a simple one. Bodies can be con-served with or without the divine impulse. Without the impulse, theyare at rest; with it, in motion. God’s conservation of body seems sepa-


rable from His role as cause of motion, and as cause of motion, Heseems to act as we would in the circumstances; God’s motion seems toresult from a divine act of will, a divine shove.

III

Now that we understand something of how God causes motion, we canreturn to the question originally posed and offer an answer.

As de La Forge construed Descartes’ views on God, continual re-creation, and God’s role as cause of motion, Descartes seems pushedinevitably toward occasionalism and the view that God is the onlygenuine cause of motion in the world; if God causes motion by re-creating bodies in different places at different times, then there seemsto be no room for finite causes to act. But by now it should be clearwhy Descartes need not be committed to such a view.

I have argued that, for Descartes, God enters as a cause of motion inorder to replace the Scholastics’ substantial forms, and, in that role, hecan (and, in the More letters, at least, is) construed as acting in just theway forms were thought to cause motion, that is, in just the way we causemotion. As such, God both sustains bodies in their being and sustainsbodies in their motion. But, it is important to note, these two activitiesseem to be quite distinct; in the one case, God is acting as a modal cause,in the other, as a substantial cause. This is an extremely importantobservation. There is no substantial cause but God, nor can there be,since no other being has the ability to create and sustain the universe.But, although God is a modal cause with respect to motion, there is noreason to hold that God is the only such cause. God is conceived to actas we do in causing motion; just as the finite cause of motion does notexclude others, so the fact that God causes motion does not seem toexclude other causes. This seems true even when we are talking aboutcausing motion in the same body. Just as two human beings can exerttheir contrary impulses on the same bit of matter, so can we impose animpulse contrary to the one God imposes. Indeed, we do so every timewe life a stone, on which God is imposing an impulse to move towardthe center of the earth.

And so Descartes would have to agree with de La Forge that Godcannot sustain bodies that are in no place at all or in indeterminateplaces; the very possibility is absurd. But, I think Descartes might insist,although God sustains bodies that have place, it is not the act of sus-taining them that gives them place. What gives them place and the


motion that puts them in different places at different times is impulseor the lack thereof, a cause quite distinct from that by which bodies aresustained. These impulses may come from God Himself, but they mightcome from other causes, like our own minds [cf. AT V 403/4]. And,when they come from God, they are not to be identified with the causeby which He sustains the bodies He moves.

There are a number of important questions relevant to the topic athand which space will not permit us to discuss. Most important, it wouldbe valuable to discuss the relations between the conception of motionand its divine cause which I have been developing with the discussionof motion and rest and their laws in Part II of the Principles and inchapter 7 of The World – the sense in which motion and rest are distinctand the sense in which they are not, the sense in which motion and restare states, and the way in which motion and rest give rise to forces thatcome into play at the time of collision. My story will not be completeuntil we see how the way in which Descartes’ immutable God causesmotion leads him to the conception of motion (and its associated forcesand laws) which underlies his program in natural philosophy.

But, incomplete as my preliminary sketch of Descartes’ position maybe, it allows us to see one important feature that differentiatesDescartes’ metaphysic of motion and his use of God as cause of motionfrom that of his avowedly occasionalist followers. What lies behind occa-sionalism as advanced by de La Forge and by many Cartesians of hisgeneration is a deep worry about causality in the world of finite things;what comes up again and again is the view that finite things are inca-pable of any genuine causal efficacy, that producing an effect is beyondthe power of any finite thing. God enters as the only being capable ofproducing any change in the world.18 Descartes’ view is quite different.Descartes never rejects finite causes as such; indeed, it is on the modelof one particular finite cause, us, that all causes are understood, con-servation excepted.19 When God enters as a cause of motion, it is simplyon account of the fact that some finite causes needed to do the job arenot available. But, even when God undertakes this task, it seems to methat Descartes can quite well hold that finite causes of motion are inno way squeezed out. Mind, indeed, can remain as direct a cause ofmotion for Descartes as God Himself.


18 See especially Nicolas Malebranche, De la récherche de la vérité, bk. VI pt. II, chapter III, andthe XV e Éclaircissement.

19 See Garber, “Understanding Interaction.”

10

DESCARTES AND OCCASIONALISM

The doctrine of occasionalism was, of course, central to seventeenth-century metaphysics. On this widely held view, the changes that onebody appears to cause in another on impact, the changes that a bodycan cause in a mind in producing a sensation, or that a mind can causein a body in producing a voluntary action are all due directly to God,moving bodies or producing sensations in minds on the occasions ofother appropriate events. And so, on this view, the tickling of the retina and subsequent changes in the brain are only the “occasionalcauses” of the sensory idea I have of a friend in the distance; the realcause is God, who directly moves my sense organs when the lightapproaches them, moves the parts of the brain when the sensory organsare moved, and then produces the sensory idea I have in my mind ofanother person’s face when my sense organs and brain are in an appro-priate state. Similarly, it is God who is the actual cause of my arm’s move-ment when I decide to raise it to wave; my volition is only an occasionalcause.

Now, occasionalism was widely held among many of Descartes’ followers; it can be found in various forms in Clauberg, Clerselier,Cordemoy, La Forge, Geulincx, and, most notably, in Malebranche.1

203

1 For general accounts of occasionalism among the members of the Cartesian school, see,for example, Joseph Prost, Essai sur l’atomisme et l’occasionalisme dans la philosophie cartésienne(Paris: Paulin, 1907); Henri Gouhier, La vocation de Malebranche (Paris: J. Vrin, 1926),chapter III; Jean-François Battail, L’avocat philosophe Géraud de Cordemoy (The Hague: Martinus Nijhoff, 1973), pp. 141–46; and Rainer Specht, Commercium mentis et corporis: über Kausalvorstellungen im Cartesianismus (Stuttgart–Bad Cannstatt: Friedrich FrommannVerlag, 1966), chapters II and III.

And throughout its seventeenth-century career it is closely associatedwith Descartes’ followers.2 But to what extent is it really Descartes’ ownview? To what extent is it fair to attribute this view to the founder of theCartesian school? This is the question that I shall explore here.

I. A Letter to Elisabeth

I will begin my investigation with a passage from a letter that Descarteswrote to the Princess Elisabeth on 6 October 1645:

All the reasons which prove the existence of God and that He is the first andimmutable cause of all the effects which do not depend on the free will ofmen, prove in the same way, it seems to me, that He is also the cause of allof them that depend on it [i.e., free will]. For one can only prove that Heexists by considering Him as a supremely perfect being, and He would notbe supremely perfect if something could happen in the world that did notderive entirely from Him. . . . God is the universal cause of everything in sucha way that He is in the same way the total cause of everything, and thusnothing can happen without His will.3

This passage would seem to be quite clear in asserting that God is thereal cause of everything in the world; if “nothing can happen withoutHis will,” as Descartes tells Elisabeth, then surely it is reasonable to inferthat Descartes was an occasionalist.

He may, in the end, turn out to be an occasionalist, but I think thatthis passage is not so clear as it may look at first. When reading this, itis very important to place it in context, and understand what exactlyDescartes was addressing in the passage. In this series of letters,Descartes is trying to console Elisabeth in her troubles. In a letter of 30September 1645, she wrote:


2 Indeed, when it first appears, it is closely associated with Descartes himself. It is an integralpart of de La Forge’s commentary on Descartes’ Treatise on Man, and it is one of the centralpoints of a letter Clerselier, Descartes’ literary executor, wrote to de La Forge in Decem-ber 1660, a letter that appeals to the authority of “nostre Maistre” on a number of occasions and that Clerselier published alongside Descartes’ own letters in one of his volumes of the philosopher’s collected correspondence. On de la Forge, see Gouhier, La vocation deMalebranche, pp. 93–94; for the Clerselier letter, see Claude Clerselier, Lettres de Mr Descartes. . . [tome III] (Paris, 1667), pp. 640–46. I am indebted to Alan Gabbey for calling the Clerselier letter to my attention.

3 AT IV 313–14. This letter appeared in the first volume of Clerselier’s edition of Descartes’correspondence in 1657.

[The fact] of the existence of God and His attributes can console us in themisfortunes that come to us from the ordinary course of nature and fromthe order which He has established there [as when we lose some goodthrough a storm, or when we lose our health through an infection in the air,or our friends through death] but not in those [misfortunes] which areimposed on us by men, whose will appears to us to be entirely free.4

Descartes’ reply, as quoted above, is that all things, including humanbeings acting freely, are under the ultimate control of an omniscient,omnipotent, and benevolent God. In saying this, Descartes does nottake himself to be saying anything particularly original; it is, indeed, atheological commonplace. While these kinds of theological issues haveled thinkers in various theological traditions to take the issue of occasionalism seriously,5 it is not appropriate to infer the full-blown metaphysical doctrine of occasionalism from this commonplace obser-vation, and conclude that Descartes held that God is the only real cause in nature; his words to Elisabeth are meant as consolation, notmetaphysics.

The question of Descartes’ occasionalism is still open. To settle it wehave to turn to a more detailed investigation of his metaphysical andphysical writings. I will divide the investigation into three parts, dis-cussing first the case of body-body causation (one billiard ball hittinganother), then mind-body causation (voluntary motions in humanbeings), and finally body-mind causation (sensation).

II. The Case of Body-Body Causation

I will not pause (too) long over this case. It seems to me as clear as anything that, for Descartes, God is the only cause of motion in theinanimate world of bodies, that bodies cannot themselves be genuinecauses of change in the physical world of extended substance. Tounderstand why, let me turn for a moment to Descartes’ reflections onmotion and its laws.6

descartes and occasionalism 205

4 AT IV 302.5 For a recent discussion of some of this larger theological debate, see Alfred Freddoso,

“Medieval Aristotelianism and the Case against Secondary Causation in Nature,” in Divineand Human Action: Essays in the Metaphysics of Theism, ed., Thomas V. Morris (Ithaca, N.Y.:Cornell University Press, 1988).

6 For a fuller account of Descartes on the laws of motion, see Daniel Garber, Descartes’ Metaphysical Physics (Chicago: University of Chicago Press, 1992).

Descartes’ conception of physics must be understood as being inopposition to an Aristotelian one, as a substitute for the kind of physicsthat was taught in the schools. Basic to the physics of the schools wasthe notion of a substantial form. According to the Aristotelian physics,each kind of thing had its own substantial form, and it was through thisthat the basic properties of things were to be explained. And so firerises and stones fall because of their forms, for example. In this way,things were thought to have basic, inborn tendencies to behavior;physics consisted in finding out what these basic tendencies were andin explaining the manifest properties of things in those terms.

A basic move in Descartes’ philosophy, something he shared withother contemporary adherents of the so-called mechanical philosophy,was the elimination of these substantial forms, these basic explanatoryprinciples. But how, then, are we to explain the characteristic behaviorof bodies? Descartes’ strategy was simple; instead of locating the basiclaws that govern the behavior of things in these forms, he placed themin God. That is, it is God, not substantial forms, that will ground thelaws that govern bodies.

How God grounds the laws of motion is illustrated in the proofs thatDescartes gives for them. These proofs are grounded in his celebrateddoctrine of continual re-creation. Descartes writes in Meditation III:

All of the time of my life can be divided into innumerable parts, each of whichis entirely independent of the others, so that from the fact that I existed ashort time ago, it does not follow that I ought to exist now, unless some causeas it were creates me again in this moment, that is, conserves me.7

Now, he argues,

plainly the same force and action is needed to conserve any thing for theindividual moments in which it endures as was needed for creating it anew,had it not existed.8

Clearly such a power is not in us; if it were, then, Descartes reasons, “Iwould also have been able to give myself all of the perfections I clearlylack.9” And so, he concludes, it must be God that creates and sustainsus.10 This conclusion, of course, holds for bodies as well as it does forus. It is not just souls, but all finite things that require some cause fortheir continued existence. And as with the idea of ourselves, “when I


7 AT VII 49. 8 Ibid. 9 See AT VII 48, 168.10 See AT VII 49–50, 111, 165, 168, 369–70; and Principles of Philosophy I 21.

examine the idea of body, I perceive that it has no power [vis] in itselfthrough which it can produce or conserve itself.”11 And so, we mustconclude that the duration of bodies, too, must be caused by God, whosustains the physical world He created in the beginning.

This view of divine sustenance underlies Descartes’ derivations of thelaws of motion, both in The World of 1633 and in the Principles of Phi-losophy of 1644. Arguing for his conservation principle in the Principles(for example, the law that God maintains the same quantity of motionin the world), Descartes writes:

We also understand that there is perfection in God not only because He isin Himself immutable, but also because He works in the most constant andimmutable way. Therefore, with the exception of those changes which evidentexperience or divine revelation render certain, and which we perceive orbelieve happen without any change in the creator, we should suppose noother changes in His works, so as not to argue for an inconstancy in Him.From this it follows, that it is most in harmony with reason for us to thinkthat merely from the fact that God moved the parts of matter in differentways when He first created them, and now conserves the totality of that matterin the same way and with the same laws [eademque ratione] with which Hecreated them earlier, He always conserves the same amount of motion in it.12

Similarly, consider his argument for the law that a body in motion tendsto move rectilinearly, as that argument is given in the Principles:

The reason [causa] for this rule is . . . the immutability and simplicity of theoperation through which God conserves motion in matter. For He conservesit precisely as it is in the very moment of time in which He conserves it,without taking into account the way it might have been a bit earlier. Andalthough no motion takes place in an instant, it is obvious that in the indi-vidual instants that can be designated while it is moving, everything thatmoves is determined to continue its motion in some direction, following astraight line, and never following a curved line.13

The picture in both of these arguments is reasonably clear: God standsbehind the world of bodies and is the direct cause of their motion. Inthe old Aristotelian philosophy, the characteristic behavior of bodieswas explained through substantial forms; in Descartes’ new, up-to-date


11 AT VII 118; see also p. 110. 12 Principles of Philosophy II 36.13 Principles of Philosophy II 39.

mechanism, forms are out, and God is in; in Descartes’ new philoso-phy, the characteristic behavior of bodies is explained in terms of animmutable God sustaining the motion of bodies.

I think that it is reasonably clear, then, that in the material world, atleast, God is the only genuine causal agent. There are some further sub-tleties in the argument that I will set aside for the moment, returningto at least one of them later. But before moving on to the somewhatmore difficult cases of mind-body and body-mind causation, I wouldlike to pause a moment and examine one complexity in the case.

Though it is clear that God is the real agent of change, the real causeof motion in the physical world, it is not at all clear how He does it,how He pulls it off. Though it is not appropriate to argue it in full detailhere, it seems to me that there are at least two somewhat differentmodels that one can find in Descartes for this.14 On one model, Godsustains the world by re-creating a succession of discrete, timeless worldstages, one after another, like frames in a movie film. On this view, Godis conceived to cause motion by re-creating bodies in different placesin different frames of the movie, as it were. We might call this the cin-ematic view of how God causes motion. But Descartes sometimes sug-gests something a bit different. On this alternative view, what Godsustains is a world of bodies existing continually in time. Now, in thisworld, some bodies are at rest, while others are in motion. Those inmotion, Descartes sometimes suggests, receive a kind of impulse fromGod. Writing to Descartes on 5 March 1649, More asked if

matter, whether we imagine it to be eternal or created yesterday, left to itself,and receiving no impulse from anything else, would move or be at rest?15

Descartes answered:

I consider “matter left to itself and receiving no impulse from anything else”as plainly being at rest. But it is impelled by God, conserving the same amountof motion or transference in it as He put there from the first.16

On this view, what might be called the divine-impulse view, God causesmotion by impulse, by a kind of divine shove.

It is interesting to try to understand how Descartes thought of Godas a cause of motion. But this distinction I have tried to make between


14 For a fuller development of this idea, see Garber, Descartes’ Metaphysical Physics, chapter 9,or Daniel Garber, “How God Causes Motion: Descartes, Divine Sustenance, and Occa-sionalism,” Journal of Philosophy 84 (1987): 567–80, essay 9 in this volume.

15 AT V 316. 16 AT V 404.

the cinematic view and the divine-impulse view of God as a cause of motionwill come in very handy when we are discussing Descartes’ thoughts onmind-body causation, to which we must now turn.

III. The Case of Mind-Body Causation

The problem of mind-body causation is, of course, a central concernof Cartesian scholarship; there are few issues in his philosophy aboutwhich more ink has been spilled. But my interest in it here is relativelynarrow: To what extent does Descartes think that there can be genuinemental causes of motions in the physical world, and to what extent doeshe believe, with the majority of his followers, that God is the true causeof motion in the world of bodies?

Here, as on the issue of body-body causation, I believe that the caseis reasonably clear: For Descartes, I think, mind can be a genuine causeof motion in the world, indeed, as genuine a cause as God Himself.

But though the case is, in the end, clear, it is not without its com-plications. As a number of later philosophers have noted, Descartes’views on God’s role as continual re-creator, that which underlies thederivation of the laws of motion, as we have seen, would seem to leadus directly to a strong version of occasionalism, where God can be theonly cause of change in the physical world. The argument is formulatedneatly by Louis de La Forge:

I hold that there is no creature, spiritual or corporeal, that can change [theposition of a body] or that of any of its parts in the second instant of its cre-ation if the creator does not do it Himself, since it is He who had producedthis part of matter in place A. For example, not only is it necessary that Hecontinue to produce it if He wants it to continue to exist, but also, since Hecannot create it everywhere, nor can He create it outside of every place, He must Himself put it in place B, if He wants it there, for if He were to haveput it somewhere else, there is no force capable of removing it from there.17

The argument goes from the doctrine of continual re-creation, authen-tically Cartesian, to the conclusion that God can be the only cause ofmotion in the world. When God sustains a body, He must sustain it some-where, and in sustaining it where He does He causes it to move or be at


17 Louis de La Forge, Oeuvres Philosophiques, ed. Pierre Clair (Paris: Presses Universitaires de France, 1974), p. 240. A similar argument can also be found in Dialogue VII of Malebranche’s Dialogues on Metaphysics.

rest. And so, it seems, there is no room for any other causes of motionin the Cartesian world, in particular, mind; if mind is to have a role toplay in where a given body is from moment to moment, it must workthrough God, who alone can sustain a body and who is ultimatelyresponsible for putting a body one place or another.18

This argument is not decisive, I think. First of all, however good anargument it might be, I see no reason to believe that Descartes ever sawsuch consequences as following out of his doctrine of continual re-creation. But, more than that, I do not think that the argument is necessarily binding on Descartes. It is certainly persuasive, particularly if one takes what I called the cinematic view of God as a cause of motion,the view in which God causes motion by re-creating a body in differentplaces in different instants of time. But the argument is considerably lesspersuasive if one takes what I earlier called the divine-impulse view ofGod as a cause of motion. On that view, God causes motion by provid-ing an impulse, much as we take ourselves to move bodies by our ownimpulses. If this is how God causes motion, then His activity in sustain-ing bodies is distinct from His activity in causing motion, and there is noreason why there cannot be causes of motion distinct from God.19

There can be causes of motion for Descartes other than God. But itstill remains to be shown that he thought that there are such causes.The question comes up quite explicitly in Descartes’ last response toHenry More:

That transference that I call motion is a thing of no less entity than shape is,namely, it is a mode in body. However the force [vis] moving a [body] canbe that of God conserving as much transference in matter as He placed in it


18 Though the argument concerns motion, states of body, and their causes, it would seemto hold for the causes of states of mind as well, insofar as the divine Sustainer must sustainminds with the states that they have as much as He must sustain bodies in the places thatthey occupy. To these arguments from continual re-creation, one might also call attentionto the several passages in which Descartes uses the word occasion to characterize particu-lar causal relations (see Prost, Essai). But as argued in Gouhier, La vocation de Malebranche,pp. 83–88, this is hardly worth taking seriously as an argument. See also Jean Laporte, Lerationalisme de Descartes (Paris: Presses Universitaires de France, 1950), pp. 225–26. Forgeneral discussions of the term, see Battail, L’avocat philosophe, pp. 141–46, and Géraudde Cordemoy, Oeuvres philosophiques, ed., P. Clair and F. Girbal (Paris: Presses Universi-taires de France, 1968), p. 322, n. 10; for a general discussion of the language of indirectcausality in Descartes and the later Scholastics, see Specht, Commercium mentis et corporis,chapters II and III.

19 This argument is developed at greater length in Garber, “How God Causes Motion.”

at the first moment of creation or also that of a created substance, like ourmind, or something else to which [God] gave the power [vis] of moving a body.20

Descartes is here quite clear that some created substances, at the veryleast our minds, have the ability to cause motion. Furthermore, thereis no suggestion in this passage that minds can cause motion in bodiesonly with God’s direct help, as the occasionalists would hold. Indeed,our ability to cause motion in the world of bodies is the very model onwhich we understand how God does it, Descartes sometimes argues.Writing to Henry More in April 1649, he remarks:

Although I believe that no mode of acting belongs univocally to God and toHis creatures, I confess, nevertheless, that I can find no idea in my mindwhich represents the way in which God or an angel can move matter, whichis different from the idea that shows me the way in which I am conscious thatI can move my own body through my thought.21

It would then be quite strange if Descartes held that minds are only theoccasional causes of motion in the world. At least two passages in thePrinciples also suggest that he meant to leave open the possibility that,in addition to God, minds could cause motion in the world. In defend-ing the conservation principle, for example, Descartes argues that weshould not admit any changes in nature “except for those changes,which evident experience or divine revelation render certain, andwhich we perceive or believe happen without any change in thecreator.”22 Such a proviso would certainly leave open the possibility that finite substances like our minds can be genuine causes of motion.Similarly, in presenting his impact law (law 3) in the Principles II 40,Descartes claims that the law covers the causes of all changes that canhappen in bodies, “at least those that are corporeal, for we are not nowinquiring into whether and how human minds and angels have thepower [vis] for moving bodies, but we reserve this for our treatise OnMan.”23 Again, Descartes is leaving open the possibility that there maybe incorporeal causes of bodily change, that is to say, motion. And so,I think, we should take him completely at his word when on 29 July1648 he writes to Arnauld:


20 AT V, 403–4. 21 AT V, 347. 22 Principles of Philosophy II 36.23 Principles of Philosophy II 40.

That the mind, which is incorporeal, can set a body in motion is shown to usevery day by the most certain and most evident experience, without the needof any reasoning or comparison with anything else.24

Minds can cause motion in Descartes’ world; there is genuine mind-body causation for him, it would seem. But before going on to examinethe last case, that of body-mind causation in sensation, I will pause fora moment and examine a question raised by the passage from the letterto More that we have been examining: What is the “something else towhich [God] gave the power [vis] of moving a body” to which Descartesrefers? Angels are certainly included, the passage from Principles II 40suggests; angels are also a lively topic of conversation in the earlierletters between Descartes and More. Indeed, when Descartes is dis-cussing with him how we can comprehend God as a cause of motionthrough the way we conceive of ourselves as causes of motion, Descartesexplicitly includes angels as creatures also capable of causing motion,like us and like God.25 It is not absolutely impossible that Descartes meantto include bodies among the finite substances that can cause motion.26

But I think that it is highly unlikely. If Descartes really thought thatbodies could be causes of motion like God, us, and probably angels, I suspect that he would have included them explicitly in the answer toMore; if bodies could be genuine causes of motion, this would be too important a fact to pass unmentioned. As I noted earlier, Descartes’whole strategy for deriving the laws of motion from the immutability of God presupposes that God is the real cause of motion and of change of motion in the inanimate world of bodies knocking up againstone another; this reading of Descartes’ view of inanimate motion seemstoo secure to be shaken on the basis of a possibly oblique remark in a letter.

Before going on to discuss the next case, I will take up one morebrief issue. It is a standard view that, for Descartes, mind cannot causemotion in a body because to do so would violate his conservation law,that the total quantity of motion in the world must always remain con-stant. And so, it is claimed, minds can change the direction with whichbodies move but cannot change the actual motion that they have. This


24 AT V 222.25 See AT V 347.26 P. H. J. Hoenen, “Descartes’s Mechanism,” in Descartes, ed. Willis Doney (New York:

Doubleday, Anchor, 1967), pp. 353–68, esp. p. 359, claims that he did include bodieshere.

is certainly a position that many of Descartes’ later followers held. ButI see no reason to believe that he himself ever maintained such a view.The argument is a bit complex, and I cannot develop the details here.27

But briefly, there is no passage in Descartes that suggests in any but theweakest way that he ever held such a position, and there are other pas-sages that strongly suggest that he did not. Furthermore, Descartes’conception of the grounds of the laws of motion in divine immutabil-ity would seem to impose no constraint on finite causes of motions, likeminds. As I noted earlier, Descartes grounds the laws of motion in God’simmutability; because God is immutable, He cannot add or subtractmotion from the world. But though the conservation principle mayconstrain God’s activity, it does not in any way constrain ours; in ourmutability and imperfection, we are completely free to add or subtractmotion to or from the world.

IV. The Case of Body-Mind Causation

We have established, I think, two reasonably clear cases: For Descartes,God is responsible for all motion in the inanimate world, while in theworld of animate creatures, creatures like us who have souls, minds can cause motion in bodies. The last case we have to take care of is that of body-mind causation, the situation in which the motion of abody causes sensations in a mind. Again, our question is this: Is theregenuine causality in this circumstance, or must God link the cause tothe effect?

Here, unfortunately, I know of no easy way of settling the questionabout Descartes’ views. It seems to me that he should be committed tothe position that the body cannot be a genuine cause of sensation inthe mind. It seems to me that if the motion of bodies is due directly toGod, and if bodies cannot be genuine causes of changes in the statesof other bodies, then it follows that bodies cannot be genuine causesof changes in minds either. This, at least, is the logic of Descartes’ posi-tion. While, to the best of my knowledge, there is no passage in his


27 In Daniel Garber, “Mind, Body, and the Laws of Nature in Descartes and Leibniz,” MidwestStudies in Philosophy 8 (1983): 105–33, essay 7 in this volume, I argue that, in fact, the lawsof motion that Descartes posits for inanimate nature do not hold for motion caused byminds, and that, in this way, animate bodies, bodies attached to minds, stand outside theworld of physics. I argue that the position widely attributed to Descartes, that the mindcan change the direction in which a body is moving but not add or subtract speed (thusapparently violating the conservation principle) is not actually his view.

writings that settles the question with assurance, there is some reasonto believe that this is a view that Descartes may have come to hold bythe late 1640s, at least.

The evidence I have in mind is connected with the proof Descartesoffers for the existence of a world of bodies. The argument first appearsin 1641 in Meditation VI:28 “Now there is in me a certain passive facultyfor sensing, that is, a faculty for receiving and knowing the ideas of sen-sible things. But I could make no use of it unless a certain active facultyfor producing or bringing about those ideas were either in me or insomething else.” So the argument begins. Descartes’ strategy is to showthat the active faculty in question is not in me (i.e., my mind), or inGod, or in anything but bodies.

This [active faculty] cannot be in me, since it plainly presupposes no intel-lect, and these ideas are produced without my cooperation, and, indeed,often involuntarily. Therefore it remains that it is in some substance differ-ent from me. . . . This substance is either body, or corporeal nature, namely,that which contains formally everything which is in the ideas [of bodies]objectively, or it is, indeed, in God, or some other creature nobler than bodyin which it [i.e., corporeal nature] is contained eminently.

To show that bodies really exist, Descartes will eliminate the latter twopossibilities, and show that the active faculty must be in bodies them-selves, or else God would be a deceiver.

The argument in Meditation VI clearly asserts that bodies have an“active faculty” that corresponds to the “passive faculty” of sensation;the clear implication is that the body that exists in the world is the causeof my sensation of it. The same basic argument comes up again, a fewyears later, in Part II, section 1, of the Principles of Philosophy of 1644,where it begins as follows:

Now, it can scarcely be doubted that whatever we sense comes to us fromsome thing which is distinct from our mind. For it is not in our power tobring it about that we sense one thing rather than another; rather, this [i.e.,what we sense] plainly depends on the very thing that affects our senses.


28 The quotations below all come from AT VII 79–80; for fuller treatment of the argument,see Martial Gueroult, Descartes’ Philosophy Interpreted According to the Order of Reasons, trans.Roger Ariew (Minneapolis: University of Minnesota Press, 1984), vol. II, chapter XIV;Daniel Garber, “Semel in vita: The Scientific Background to Descartes’ Meditations,” inEssays on Descartes’ Meditations, ed. Amélie Rorty (Los Angeles and Berkeley: University of California Press, 1986), essay 11 in this volume, pp. 251–53.

As in the Meditations, Descartes goes on to examine the question as towhether the sensation might proceed from me, from God, or fromsomething other than bodies. Talking about that from which thesensory idea proceeds, he says:

[W]e clearly understand that thing as something plainly different from Godand from us (that is, different from our mind) and also we seem to ourselvesclearly to see that its idea comes from things placed outside of us, things towhich it [i.e., the idea] is altogether similar, and, as we have already observed,it is plainly repugnant to the nature of God that He be a deceiver.

And so, Descartes concludes, the sensory idea proceeds from a body.The argument is the Principles is obviously similar to the one in the

Meditations. But there is at least one crucial difference. The argumentin Meditation VI starts with the observation that I have “a certain passivefaculty for sensing”; what we seek is the active faculty that causes thesensations I have, and the ultimate conclusion is that that active facultyis found in bodies. But, interestingly enough, in the argument of thePrinciples there is no appeal to an active faculty. Indeed, the terminol-ogy Descartes uses to describe the relation between our sensation andthe body that is the object of that sensation seems studiously noncausal;we all believe, Descartes tells us, that “whatever we sense comes to us[advenit] from something which is distinct from our mind,” that theidea of body “comes from [advenire] things placed outside of us.” Theconcern I have attributed to Descartes here is suggested further by avariant that arises between the Latin version of Principles II 1, which wehave been discussing, and the French version published three yearslater in 1647. In the Latin, the crucial phrase reads as follows:

We seem to ourselves clearly to see that its idea comes from things placedoutside of us.29

In the French translation, the phrase reads:

it seems to us that the idea we have of it forms itself in us on the occasion ofbodies from without.30

One must, of course, be very careful drawing conclusions from variantsbetween the Latin text and Picot’s French translation; while some alter-natives are clearly by Descartes, it is often unclear whether a given


29 Principles of Philosophy II 1, translation of Latin version.30 Principles of Philosophy II 1, translation of French version; emphasis added.

change is due to the author or to his translator. But this change is con-sistent with the trend already observed between Meditation VI and Prin-ciples II 1, Latin version, and weakens the causal implications furtherstill. Rather than asserting that the idea comes from the thing, the Frenchtext says only that it “forms itself in us on the occasion of bodies fromwithout.” Furthermore, while it is by no means clear how to interpretthe word occasion in Descartes’ vocabulary, the word is certainly sug-gestive of what is to become a technical term in later Cartesian vocab-ulary, that of an occasional cause, a cause whose effect is producedthrough the activity of God.31

It is difficult to say for sure why the two arguments differ in thisrespect, and one should always be open to the explanation that, asDescartes suggests in a number of places, metaphysical issues are takenup in the Principles in a somewhat abbreviated and simplified fashion,and that the Meditations must be regarded as the ultimate source for hisconsidered views in that domain.32 But it is tempting to see in this vari-ation the shadow of an important philosophical question Descartes wasfacing. It is possible that he eliminated the reference to an active facultyprecisely because he was no longer certain that bodies could correctlybe described as active causes of our sensations. The language he sub-stitutes is, of course, consistent with bodies being active causes of sen-sations, as he may well have believed; but it is also consistent with aweaker view, on which our sensations come from bodies, but with the helpof an agent, like God, distinct from the bodies themselves, which, inthe strictest sense, are inert.

There is another place that is sometimes thought to support the attri-bution of occasionalism to Descartes. The passage I have in mind is thecelebrated one from the Notae in Programma (1647):

Nothing reaches our mind from external objects through the sense organsexcept certain corporeal motions. . . . But neither the motions themselvesnor the shapes arising from them are conceived by us exactly as they occurin the sense organs, as I have explained at length in my Dioptrics. Hence itfollows that the very ideas of the motions themselves and of the shape areinnate in us. The ideas of pain, colors, sounds, and the like must be all themore innate if, on the occasion of certain corporeal motions, our mind is to


31 See the reference given in note 18 above in connection with the word occasion.32 On the relations between the Meditations and Part I of the Principles, see, for example,

AT III 233, 259; AT V 291; and AT IXB 16.

be capable of representing them to itself, for there is no similarity betweenthese ideas and the corporeal motions.33

The use of the word “occasion” in this context (as well as in a previoussentence on the same page) does lend some support to the claim thatthe use of the corresponding French word in the French translation ofthe Principles, published in the same year, is no accident, and may besignificant for the way in which Descartes is thinking about body-mindcausality. But it is important to recognize that the claim that the sensoryidea is innate in the mind is, I think, irrelevant to the issue of Descartes’occasionalism. His worry here is not (primarily) the causal connectionbetween the sensory stimulation and the resulting sensory idea; whatworries him is their utter dissimilarity, the fact that the sensory idea isnothing like the motions that cause it. To make an analogy, consider,for example, a computer with a color monitor capable of displayingcomplicated graphics and pictures. Suppose that if I tap in a certainsequence of keystrokes, a picture of the Notre Dame in Paris appearson the screen. One might perhaps want to point out that the actualsequence of motions (i.e., the keystrokes) that causally produce thepicture in no way “resembles” the picture, and one might reason fromthat fact to the claim that the picture must be innate in the machine,that is, stored in its memory. But one probably would not want to reasonfrom that that the keystrokes are not in some sense the direct cause ofthe picture’s appearing, that the keystrokes did not really elicit thepicture; and one certainly would not want to infer that it was God whosomehow connected the keyboard with the screen of the monitor. I think that the situation is similar with respect to Descartes’ point inthe passage quoted from the Notae in Programma; in this case, as in thecomputer case, Descartes’ main point is simply that sensory ideascannot come directly from the motions that cause them, but must, atbest, be innate ideas that are elicited by the motions communicated tothe brain by the sense organs.

But even though this passage does not lend much support to the viewthat Descartes may have come to see God as connecting bodily motionswith sensations, neither does it detract from the evidence I presentedearlier. And so, while the evidence is not altogether satisfactory, it seemsreasonable to think that while Descartes may have seen bodies asgenuine causes of sensations at the time that the Meditations was


33 AT VIIIB 359.

published in 1641, by the publication of the Principles of Philosophy a fewyears later he may have changed his view, holding something closer to what his occasionalist followers held, that God is the true cause ofsensations on the occasion of certain motions in bodies.

V. Was Descartes an Occasionalist?

In the earlier parts of this essay we have examined three different sortsof causal relations as treated by Descartes in his thought. While it seemsclear that mind can be a genuine cause of motion in the physical world,it also seems clear that God is the real cause of change in the inanimateworld of physics, and it seems probable that God is the real causebehind body-mind interaction, the causation of sensations in the mind.It thus seems clear that while Descartes may share some doctrines withthe later occasionalists of the Cartesian school, he is not an occasion-alist, strictly speaking, insofar as he does allow some finite causes intohis world, minds at the very least.

Might we say, on this basis, that Descartes is a quasi-occasionalist, anoccasionalist when it comes to the inanimate world, though not in theworld of bodies connected to minds? The doctrine of occasionalism iscertainly flexible enough to allow this. But even if we choose to viewDescartes in this way, we must not lose sight of an important differencebetween Descartes and his occasionalist followers.

For many of Descartes’ later followers, what is central to the doctrineof occasionalism is the denial of the efficacy of finite causes simply byvirtue of their finitude. Clerselier, for example, argues for occasional-ism by first establishing that only an incorporeal substance can causemotion in body. But, he claims, only an infinite substance, like God,can imprint new motion in the world “because the infinite distancethere is between nothingness and being can only be surmounted by apower which is actually infinite.”34 Cordemoy argues similarly. LikeClerselier, he maintains that only an incorporeal substance can be thecause of motion in a body, and that this incorporeal substance can only


34 Clerselier, Lettres de Mr Descartes . . . [tome III]. p. 642. Clerselier argues that while a finiteincorporeal substance, like our mind, cannot add (or destroy) motion in the world, it canchange its direction, because, unlike motion itself, “the determination of motion . . . addsnothing real in nature . . . and says no more than the motion itself does, which cannot bewithout determination” (ibid.). This, though, would seem to conflict with what Descarteshimself told Clerselier in the letter of 17 February 1645, that motion and determinationare two modes of body that “change with equal difficulty” (AT IV 185).

be infinite; he concludes by saying that “our weakness informs us thatit is not our mind which makes [a body] move,” and so he determinesthat what imparts motion to bodies and conserves it can only be“another Mind, to which nothing is lacking, [which] does it [i.e., causesmotion] through its will.”35 And finally, the infinitude of God is centralto the main argument that Malebranche offers for occasionalism in hismajor work, De la recherche de la vérité. The title of the chapter in whichhe presents his main arguments for the doctrine is “The most danger-ous error in the philosophy of the ancients.”36 And the most dangerouserror he is referring to is their belief that finite things can be genu-ine causes of the effects that they appear to produce, an error that,Malebranche claims, causes people to love and fear things other thanGod in the belief that they are the genuine causes of their happinessor unhappiness.37 But why is it an error to believe that finite things canbe genuine causes? Malebranche argues as follows:

As I understand it, a true cause is one in which the mind perceives a neces-sary connection between the cause and its effect. Now, it is only in an infi-nitely perfect being that one perceives a necessary connection between itswill and its effects. Thus God is the only true cause, and only He truly hasthe power to move bodies. I further say that it is not conceivable that Godcould communicate to men or angels the power He has to move bodies.38

For these occasionalists, then, God must be the cause of motion in the world because only an infinite substance can be a genuine cause ofanything at all.

But, as I understand it, Descartes’ motivation is quite different. Heseems to have no particular worries about finite causes as such. If I amright, he is quite happy to admit our minds and angels as finite causesof motion in the world of bodies. Indeed, it is through our own abilityto cause motion in our bodies that we have the understanding we doof God and angels as causes of motion. When God enters as a cause ofmotion, it is simply to replace a certain set of finite causes, the sub-stantial forms of the Schoolmen, which, Descartes thinks, are unavail-


35 Cordemoy, Oeuvres philosophiques, p. 143.36 See Malebranche, De la recherche de la vérité VI.2.iii, in Malebranche, Oeuvres complètes

de Malebranche, ed., André Robinet (Paris: J. Vrin, 1958–70), vol. I, p. 643, trans. in Malebranche, The Search after Truth, trans., T. M. Lennon and P. J. Olscamp (Columbus,OH: Ohio State University Press, 1980), p. 446.

37 Malebranche, Oeuvres, vol. I, pp. 643–46; Search, pp. 446–48.38 Malebranche, Oeuvres, vol. I, p. 649; Search, p. 450.

able to do the job. He argued that the substantial forms of Scholasticphilosophy were improper impositions of mind onto matter and must,as such, be rejected. But, one might ask, if there are no forms, what can account for the motion that bodies have, for their characteristicbehavior? What Descartes turns to is God. In this way he seems less a precursor of later occasionalism than the last of the Schoolmen, using God to do what substantial forms did for his teachers.39


39 Portions of this essay have also appeared in Garber, Descartes’ Metaphysical Physics.

11

SEMEL IN VITA

The Scientific Background to Descartes’ Meditations

Descartes opens Meditation I with his persona, the meditator, reflect-ing on the project to be undertaken. Descartes writes:

I have observed for some years new how many false things I have admittedas true from my earliest age, and thus how dubious are all of those thingsthat I built on them; and so, I observed that once in life [semel in vita] every-thing ought to be completely overturned, and ought to be completely rebuiltfrom the first foundations, if I want to build anything firm and lasting in thesciences. (AT VII 17)1

And with this, the project has begun. Descartes’ meditator quicklybegins by rejecting the commonsense epistemological principles onwhich everything he formerly believed rested, and quickly sets aboutputting the world back together again. Of course, one of the centralprojects undertaken in this connection must be the replacement of theepistemological principles rejected with new, more trustworthy prin-ciples. Just as Descartes’ meditator undermined his former beliefs by undermining the epistemology on which they were based, he willrebuild his world by rebuilding its epistemology. New epistemologicalprinciples thus seem to be the very “first foundations” on which he willbuild something “firm and lasting in the sciences.” But an obvious ques-tion to raise about this, the opening sentence of the Meditations, andabout the project that follows out of it, is why? Why does Descartesbelieve it necessary even once in life to rebuild all of our beliefs in theway he suggests? Why does Descartes feel called to such an epistemo-

221

1 All textual citations will be given in the body of the essay.

logical project? Why is any genuine knowledge, anything “firm andlasting in the sciences” not possible without entering into such a Herculean labor, cleaning out and rebuilding from the bottom up thecluttered stable-stalls of the mind?

There is an answer to this question that has been put forward by awide variety of commentators, and has become, perhaps, the standardaccount of Descartes’ motivation for taking up epistemology in the Med-itations. In that view, one sees Descartes as engaged in a debate withradical skepticism; the claim is that the call to new foundations is pri-marily a call to find epistemological principles immune to skepticalattack.2 This is a reading for which there is a great deal of support; boththe general intellectual climate, the revival of skeptical thought in thesixteenth and early seventeenth centuries, and the details of MeditationI where Descartes presents a number of arguments derived from theskeptical tradition for later response, point to skepticism as a majorintellectual problem for Descartes in the Meditations.

But, I claim, this is not the whole story. In a letter Descartes wrote tohis close friend Marin Mersenne on 28 January 1641, while he was com-posing the Replies to the Objections submitted to his Meditations, andpreparing the whole work for publication, Descartes confided:

I may tell you, between ourselves, that these six Meditations contain the entirefoundations for my physics. But it is not necessary to say so, if you please,since that might make it harder for those who favor Aristotle to approve them.I hope that those who read them will gradually accustom themselves to myprinciples and recognize the truth in them before they notice that theydestroy those of Aristotle. (AT III 297–298)3


2 For recent developments of this reading, see Harry G. Frankfurt, Demons, Dreamers, andMadmen (Indianapolis: Bobbs-Merrill, 1970), esp. pp. 174–175; Richard H. Popkin, TheHistory of Scepticism from Erasmus to Spinoza (Berkeley, Los Angeles, London: University ofCalifornia Press, 1979), chapter 9; and Alexandre Koyré’s introductory essay in E.Anscombe and P. T. Geach, Descartes: Philosophical Writings. E. M. Curley’s Descartes Againstthe Skeptics (Cambridge, Mass.: Harvard University Press, 1978, pp. 8–9) also suggests sucha view. But in private communication, Curley has emphasized that in stressing the attackon skepticism, he did not mean to deny that the Meditations plays other equally importantroles in Descartes’ philosophy, most prominently in the grounding of his physics. See hischapter 8. For a valuable discussion of Descartes’ attitude towards skepticism, see HenriGouhier, La Pensée Métaphysique de Descartes (Paris: Vrin, 1978), chapter 1.

3 On the importance of the Meditations project as the first step in building his new science,see also the introduction to the French edition of the Principles of Philosophy (AT IXB 13–17).The importance of the program of the Meditations as a foundation for the sciences is alsosuggested by Descartes’ critique of the apostate Henricus Regius. See, e.g., AT IXB 19–20and AT IV 625.

Descartes, thus, is absolutely clear that the program of the Meditationsis not an autonomous philosophical project, but the prelude to a largerscientific program; his remarks to Mersenne suggest that the motiva-tion for the Meditations cannot be merely the refutation of skepticism,a problem that, it would seem, is of no pressing concern to the prac-ticing scientist.4 The Meditations is, as it were, a Trojan horse thatDescartes is attempting to send behind the lines of Aristotelian science.Now, there are a number of ways in which the Meditations can be seento lay the foundations for Cartesian science. One can see, for example,in the discussions of body, its distinction from mind and its nature as extension and extension alone, hints of Descartes’ mechanisticaccounts of the human body, and the world of physics, as we shall latersee. But I think that Descartes meant something deeper still. I shallargue that the Meditations are intended to give the epistemological foun-dations of the new science as much as its metaphysical foundations;5 theaccount of knowledge, of clear and distinct perception, imagination,and sensation that forms the backbone of the Meditations is, I claim,intended to undermine the epistemology that underlies Aristotelianphysics, and lead directly to its replacement by a Cartesian conceptionof the way the world is. It is in this sense, too, that the Meditations con-tains “the entire foundations for my physics.”6

This, then, is what I’ll try to do in this essay – set the epistemologi-cal project of the Meditations into the broader context of the Cartesianprogram for science,7 and show why Descartes thought that such anepistemological project was a necessary preliminary to scientific inves-

SEMEL IN VITA 223

4 For an account that suggests that skepticism was a problem for practicing scientists, see,e.g., Philip Sloan, “Descartes, the Skeptics, and the Rejection of Vitalism in Seventeenth-Century Physiology,” Studies in History and Philosophy of Science 8 (1977):1–28. But see alsomy discussion of this essay in Studia Cartesiana 2 (1981):224–225.

5 I don’t intend, in putting the matter this way, to suggest that there is a radical distinctionbetween metaphysical and epistemological concerns. Descartes’ epistemology stronglydepends on issues relating to the nature of mind, its relation to the body, and its relationto the benevolent God who created it.

6 For similar readings of Descartes’ project, see, e.g., Étienne Gilson, Études sur le Rôle de laPensée Médiévale dans la Formation du Système Cartésien (Paris: Vrin, 1975), Part 2, chapter 1;and Margaret D. Wilson, Descartes (London: Routledge & Kegan Paul, 1978), 3–4; 104.Wilson’s point of view is contrasted with Curley’s in Willis Doney’s “Curley and Wilson onDescartes,” Philosophy Research Archives, Jan. 1, 1980.

7 It should be noted that I am using the term science in an anachronistic way here, and meanit to refer to areas of inquiry that we call scientific, physics, biology, etc. In Latin, science,scientia, means just knowledge. Thus Descartes wrote in Rule II of his early Rules for the Direc-tion of the Mind: “All science is certain and evident cognition” (AT XI 362).

tigation. I shall begin with a brief discussion of the Cartesian programin physics, and the conception of the world against which it was explic-itly directed. I shall argue there that Descartes saw both the Aristotelianand common views of the world as closely connected with certain deeplyheld but very mistaken epistemological views. I shall then try to showhow both the skeptical arguments of Meditation I and the more posi-tive arguments of the succeeding Meditations function in the overthrowof the commonsense epistemology and the Aristotelian metaphysics itsupports, and in the establishment of epistemological foundations forthe Cartesian science. In this way, I hope to show one motivation, overand above any worries about skepticism, for entering into the episte-mological project of the Meditations.

Mechanism, the Vulgar Philosophy, and the Sins of Youth

If we are to read the Meditations as a prelude to Cartesian science, thenwe must begin with a few words about just what Descartes’ conceptionof science was. While there are complications in dealing with Cartesianmedicine and psychology, complications introduced by the mind andits union with body, in physics the program is straightforward: Descartesthe physicist was a mechanist, and held that all physical phenomenawere ultimately explicable in terms of the shape, size, and motion ofthe normally insensible corpuscles that compose the gross bodies ofeveryday experience.8

A full account of Descartes’ physics is far beyond the scope of thisessay.9 But the program as given in Part 2–4 of his Principles of Philoso-phy, its most careful and systematic development, can be summarizedas follows.10 Descartes begins with an account of the nature of body,


8 For general accounts of the so-called mechanical philosophy, see, e.g., Richard Westfall,The Construction of Modern Science (New York: John Wiley and Sons, 1971); Marie Boas,“The Establishment of the Mechanical Philosophy,” Osiris X (1952):412–541; or E. J. Dijksterhuis, The Mechanization of the World Picture (Oxford: Oxford University Press, 1961), esp. Part 4, chapter 3.

9 For some general accounts of Cartesian physics, see, e.g., Paul Mouy, Le Développement dela physique Cartésienne (Paris: Vrin, 1934), 1–71; J. F. Scott, The Scientific Work of Descartes(London: Taylor and Francis, 1952); and E. J. Aiton, The Vortex Theory of Planetary Motions(New York: Neale Watson, 1972).

10 The principal statements of Cartesian physics are the Le Monde of 1632 (in AT XI; trans-lated by Michael Mahoney as René Descartes: The World [New York: Abaris Books, 1979]);the Dioptrics and Meteors of 1637 (in AT VI translated by Paul J. Olscamp in René Descartes:Discourse on Method; Optics, Geometry, and Meteorology [Indianapolis: Bobbs-Merrill, 1965]),and Principles of Philosophy of 1644. The Principles is the only attempt Descartes made at acomplete and systematic exposition of his physics.

which, as he says, “does not consist in weight, hardness, color, or thelike, but in extension alone” (Pr II 4; cf. Pr II 11), and an account ofthe modes of body, shape, size, and motion. Of these, motion, definedas “the translation of one part of body, or of one body, from the neigh-borhood of those bodies which immediately touch it . . . and into theneighborhood of others” (Pr II 25), gets special attention. Descartes iscareful to distinguish motion, a mode of extension, so he claims, fromits principal cause, God, who, in continually re-creating the world frommoment to moment, is responsible for the changes in place bodies areobserved to have.11 And from the activity of God, “not only because Heis a Himself immutable, but also because He acts in as constant andimmutable a way as possible” (Pr II 36), Descartes derives the laws ofmotion, the laws that govern bodies as such, the conservation of quan-tity of motion, the persistence of size, shape, and rectilinear motion,and the laws bodies obey in impact (cf. Pr II 36–42).12 And with this,the mechanist program is off and running. Since all matter is of thesame sort and obeys the same laws, we have no choice but to explainthe special behavior that individual bodies exhibit (the heaviness ofstones and the lightness of air, the color of milk and the attractive prop-erties of lodestones) in terms of the differing size, shape, and motionof the smaller bodies (or corpuscles) that make them up, and the lawsof geometry and motion that govern them (cf. Pr II 22–23). Descartesthus wrote in his Principles II, 64, after this analysis of motion and justprior to the execution of his program in III and IV:

I openly acknowledge that I know of no other matter in corporeal thingsother than that which is divisible, shapable, and movable in every way, andwhich the geometers call quantity, and take as the object of their demon-strations; and that there is nothing in it to consider except those divisions,shapes, and movements; and that nothing concerning these can be acceptedas true unless it is deduced from these common notions, whose truth wecannot doubt, with such certainty that it must be considered as a mathe-

SEMEL IN VITA 225

11 See Pr II 36 and AT V 403–404. This latter passage suggests that God is not the only causeof motion in the world, and that mind can be a genuine cause of at least some motion.On this see my essay 7 in this volume. “Mind, Body, and the Laws of Nature in Descartesand Leibniz”. On the role of God plays in the derivation of the laws of motion in Descartes,see, e.g., Gary Hatfield, “Force (God) in Descartes’ Physics,” Studies in History and Philosophy of Science 10 (1979):113–140; and Martial Gueroult, “The Metaphysics and Physics of Force in Descartes,” in Stephen Gaukroger, ed., Descartes: Philosophy, Mathematics, and Physics (Totowa, N.J.: Barnes and Noble, 1980), 196–229.

12 For an account of the derivation of the laws of motion, see, e.g., Alan Gabbey, “Force andInertia in the Seventeenth Century: Descartes and Newton,” in Stephen Gaukroger,Descartes, 230–320.

matical demonstration. And because all natural phenomena can thus beexplained, as will appear in what follows, I think that no other principles ofphysics should be accepted, or even desired (Pr II 64. Cf. AT II 542; AT III 686)

And thus Descartes observes in his Sixth Replies that, in his physics, allof the sensible properties that bodies seem to have, all color, sound,heaviness, and lightness are to be eliminated from the physical world,leaving only geometry and the laws of motion behind:

I observed that nothing at all belongs to the nature [ratio] of body exceptthat it is a thing with length, breadth, and depth, admitting of various shapesand various motions; that its shapes and motions are only modes which nopower could make to exist apart from it; that colors, odors, tastes, and thelike are merely sensations existing in my thought, and differing no less frombodies than pain differs from the figure and motion of the weapon thatinflicts it; and finally that heaviness [gravitas], hardness, the powers [vires] ofheating, attracting, purging, and all other qualities which we experience inbodies consist solely in motion or its absence, and in the configuration andsituation of their parts. (AT VII 440)

The details of Descartes’ ambitious program, the imaginativeaccounts of light, color, magnetism, gravity, and a host of other phe-nomena that Descartes attempted to explain in these terms, are veryinteresting and well worth the study, even if they tell us more aboutDescartes’ scientific personality than about the world. But for themoment I would like to turn away from Cartesian physics and examinean alternative conception of the physical world.

Descartes’ mechanism is in explicit opposition to a different con-ception of the world, a combination of common sense and ScholasticAristotelianism. In the commonsense view of the world, at least asDescartes imagines it, everything is, for the most part, just as it appearsto us. Things really are colored, they are hot or cold, bitter or sweet,and pains are, for the most part, just where you think they are.13 Bodies,in this view, have some internal property (resistance) by which theyresist motion, and something (heaviness or gravity) by virtue of whichthey move themselves toward the center of the earth. And, Descartesthinks, when we see no body, common sense is inclined to believe thatno body is present, that is, there is vacuum.14


13 See, e.g., the commonsense mistakes that Descartes calls attention to in Pr I, 46, 66–68.14 On the commonsense conception of resistance, see, e.g., Pr II 26: AT II 212–213. It is in

this sense that Descartes denies “inertia or natural tardiness” to bodies, a tendency to come

This much is common sense, what most people take the world to be.But, Descartes thinks, it is this that underlies the principal opponent to his mechanism in the learned world, the Aristotelian worldviewcommon to the Scholastics that Descartes was taught at La Flèche, whathe called on occasion the “vulgar philosophy,” in recognition of its widespread acceptance (cf. AT I 421; AT III 420; AT IV 30). LateScholastic Aristotelianism, the philosophy taught in the universities and colleges in Descartes’ day, was a phenomenon of great complexity,encompassing a number of different schools of thought with importantdifferences on a number of different issues.15 But Descartes was notinterested in the fine points of the Scholastic debates. Descartes writesin a letter to Mersenne from 1640:

I do not think that the diversity of the opinions of Scholastics makes theirphilosophy difficult to refute. It is easy to overturn the foundations on whichthey all agree, and once that has been done all their disagreements over detailwill seem foolish. (AT III 231–232)

What he objected to in Scholasticism was something he saw as commonto all schools, a common conception of the makeup of the physicalworld together with a closely connected pattern of explanation inphysics.

Basic to that view, as Descartes understood it, was the notion of aform or a real quality, and the explanation of the behavior of bodies inthese terms.16 Forms and qualities are, as Descartes put it, “the imme-

SEMEL IN VITA 227

to rest or a resistance to being set in motion from rest (AT II 466–467), although Descartesis perfectly willing to admit as a consequence of his conservation law that one body movinganother will lose some of its own motion (AT II 543; AT II 627). On commonsense conceptions of heaviness, see, e.g., AT III 667; AT VII 441–442. On the commonsenseprejudices that lead to a belief in vacua, see, e.g., Pr II, 17–18.

15 For a survey of some aspects of late Scholasticism relevant to the foundations of physics,see William A. Wallace, “The Philosophical Setting of Medieval Science,” in David C. Lindberg, ed., Science in the Middle Ages (Chicago: University of Chicago Press, 1978),91–119. For an account of the diversity of seventeenth-century Scholasticism on some of the issues about substance relevant to Descartes, see A. Boehm, Le “Vinculum Sub-stantiale” chez Leibniz: ses Origines Historiques (Paris: Vrin, 1962), 33–81. For an account ofDescartes’ relations with late Scholasticism, see Gilson, Rôle, and the extremely valuablecollection of Scholastic texts that Gilson published in his Index Scolastico-Cartésien (Paris:Vrin, 1979).

16 It should be noted that Descartes’ representation of Scholastic doctrine is not always accu-rate. As Gilson notes (Rôle, 163), Descartes’ view is that the Scholastic form is a substance(AT III 502), a conception that is a matter of some controversy among Scholastics. Also,Descartes draws no distinction between the Scholastic conceptions of form and realquality. Cf., e.g., Descartes’ defintion of form in AT III 502 with the conception of real

diate principles of action of things,” introduced “so that through themwe can explain the actions proper to natural things, of which the formis the principle and source” (AT III 503, 506). And so, correspondingto salient qualities or characteristic kinds of behavior, the Scholasticposits a form whose function it is to explain the quality or behaviorobserved. Eustachius a Sancto Paulo, a seventeenth-century Scholasticwhose work Descartes knew and considered representative of the tra-dition, thus wrote:

There are individual and particular behaviors [ functiones] appropriate toeach individual natural thing, as reasoning is to human beings, neighing tohorses, heating to fire, and so on. But these behaviors do not arise frommatter. . . . Thus, they must arise from the substantial form.17

The extent to which forms are linked to specific behaviors is empha-sized in another passage, where Descartes gives his confidant Mersennean account of what he takes to be “the most common explanation ofheaviness of all in the Schools” in preparation for giving his own mech-anistic account:

Most take it [i.e., heaviness] to be a virtue or an internal quality in every bodythat one calls heavy which makes it tend toward the center of the earth; andthey think that this quality depends on the form of each body, so that thesame matter which is heavy, having the form of water, loses this property ofheaviness and becomes light when it happens that it takes on the form of air.(AT II 223)18

Here the observed behavior is so closely linked to a specific form thata change in characteristic behavior from heavy to light requires achange in form.

Insofar as the activity the Scholastic attributes to the body itself is,Descartes thinks, comprehensible only through the category of themental, the Scholastic account of the characteristic properties of bodies


quality expressed, e.g., in AT III 648; AT III 667; AT V 222; AT VII 441–442. Consequently,I shall draw no distinction between form and quality. The account of the Scholastic con-ception of substance given in the text is not intended to be an accurate account of Scholas-tic doctrine. It should be read as a representation of what the Scholastic opponent lookedlike to Descartes.

17 Gilson, Index, sec. 209. See also Eustachius a Sancto Paulo, Summa Philosophica . . . (Cam-bridge: Roger Daniel, 1648), 123–124, 127, 140; and a passage from Suarez, DisputationesMetaphysicae . . . given in Gilson, Index, sec. 211. For Descartes’ judgment of Eustachius,see AT III 232.

18 Cf. Gilson, Rôle, 159–162.

amounts to the attribution of a “tiny mind,” as he put it, linked withspecific behavior, to inanimate bodies, in order to explain that behav-ior (AT III 648; cf. AT VII 441–442). And so, for example, Descartesthinks of the Scholastic notion of heaviness as something mental, a sub-stance linked to body that “bears bodies toward the center of the earthas if it contains some thought of it [i.e., the center of the earth] withinitself ” (AT VII 442). This allows Descartes to appeal to the Scholasticaccount of heaviness to convince confused correspondents that insofaras they find the philosophy they learned in school comprehensible, theyshould have no particular trouble with Descartes’ own conception ofmind-body interaction; the Scholastic account of gravity is, in essence,a misapplication of a notion that “was given us for the purpose of con-ceiving the manner in which the soul moves the body” (AT III 667), aprojection of our dual nature onto the inanimate world.19 And thusDescartes wrote in a letter in 1641:

The first judgments that we have made since our childhood, and since then,the vulgar philosophy [i.e., Scholasticism] have accustomed us to attribute tobodies many things which only pertain to mind and to attribute to mind manythings that only pertain to body. One ordinarily mixes the two ideas of body and mind, and in the compounding of these ideas, one fashions realqualities and substantial forms, which I think should be entirely rejected. (AT III 420)

There are, to be sure, traces of Scholastic ontology in Descartes’ ownmetaphysics, and ways of reconciling Descartes’ own mechanical phi-losophy with the Scholasticism he rejects.20 But there is, from Descartes’

SEMEL IN VITA 229

19 See also the discussion of this question in sec. 2 of my essay, “Understanding Interaction:What Descartes Should Have Told Elisabeth,” Southern Journal of Philosophy, supplement tovol. 21 (1983):15–32, essay 8 in this volume.

20 One trace of Scholastic ontology is the notion of tendency that is essential to Descartes’derivation of the laws of motion. On this and the closely related notion of force, see, e.g.,Gueroult, “Metaphysics and Physics,” and Thomas L. Prendergast, “Motion, Action, andTendency in Descartes’ Philosophy,” Journal of the History of Philosophy 13 (1975):453–462.Another trace of Scholasticism is in his notion of the relation between mind and body.Descartes sometimes claims that mind can be regarded as the substantial form of the body,the only such form he recognizes. See, e.g., AT III 503, 505, AT IV 168, 346. On this see,e.g., Gilson, Rôle, 245–255; Geneviève [Rodis-]Lewis, L’Individualité selon Descartes (Paris:Vrin, 1950), 67–81. It is also possible to assimilate Aristotelian ideas to Cartesian in theother way, by interpreting Aristotle as a Cartesian. This was an idea that attracted the youngLeibniz. See his letter to Jacob Thomasius, April 20/30, 1669, in Leroy Loemker, ed. andtrans., G. W. Leibniz: Philosophical Papers and Letters (Dordrecht: Reidel, 1969), 93–103. Inthat letter, Leibniz cites a number of his lesser known contemporaries in connection withthis view.

point of view, at least, a clear contrast between the two. Thus Descarteswrites in Le Monde, comparing his own theory of combustion with thatof the Scholastics:

When it [i.e., fire] burns wood or some other such material, we can see withour own eyes that it removes the small parts of this wood, and separates themfrom one another, thus transforming the more subtle parts into fire, air, andsmoke, and leaving the grossest parts as cinders. Let others imagine in thiswood, if they like, the form of fire, the quality of heat, and the action whichburns it as separate things. But for me, afraid of deceiving myself if I assumeanything more than is needed, I am content to conceive here only the move-ment of parts. (AT XI 7)

And in more general terms Descartes writes in a 1638 letter:

Compare my assumptions [suppositions] with the assumptions of others.Compare all their real qualities, their substantial forms, their elements, andtheir other countless hypotheses with my single assumption that all bodiesare composed of parts . . . . All that I add to this is that the parts of certainkinds of bodies are of one shape rather than another. (AT II 200)

The contrast is a basic one. For the Scholastic, there is an indefinitelylarge variety of distinct principles of action in the world, one corre-sponding to each kind of characteristic behavior that the Scholasticchooses to recognize in his physics. But for Descartes, while there arean infinite number of ways that extended matter may subdivide intosmaller parts, there is only one kind of stuff in the physical world, and itall behaves in the same way, in accordance with geometry and the lawsof motion; it is in terms of this matter, its laws, and the particular geo-metric configurations it forms in different bodies that all bodily phe-nomena are to be explained. In the Cartesian world, no body is literallyheavy, or hot, or red, or tasty. All these observed properties are a resultof geometry and motion. The mechanistic explanations of such phe-nomena Descartes gives, their reduction to configurations of matter inmotion, may in the end turn out to be every bit as ad hoc as the stageAristotelian’s dormitive virtue explanation of the behavior of opium; itis just as easy to appeal to an unknown corpuscular substructure, anoccult mechanism, as it is to appeal to an occult quality. But there is noconfusing the two kinds of explanations.

Given Descartes’ conception of Scholastic philosophy, it is not diffi-cult to see why he often links the errors of Scholasticism with the errors


of common sense.21 Common sense attributes to bodies the qualities andtendencies to behave in particular ways that bodies appear to have, theproperties our senses tell us bodies have. The Scholastic philosophertakes this one step further, and posits in bodies forms and qualities, prin-ciples of action that are intended to explain the properties that sensetells us are in bodies. Since the qualities that sense attributes to bodiesare largely mental qualities, the sensations and volitions of the minditself, projected onto the physical world as colors, tastes, and tendencies,the forms and qualities must be “tiny minds,” mental substances capableof receiving the properties that common sense attributes to them. TheScholastic world is, thus, nothing but the world of common sense, withsensible qualities transformed into mental substances – forms and realqualities – and embedded in the world of bodies.22 Put briefly, theScholastic world, as Descartes understood it, is simply a metaphysicalelaboration of the world of common sense.23

This, then, is how Descartes sees the matter, his own conception of theworld, the forces of light, against the dark world of common sense andthe obscurities of Scholasticism. But although he thought his opponentswrong, he did not underestimate the attractiveness and virtual inevitabil-ity of their position. For, Descartes thought, the commonsense worldviewand the Scholastic metaphysics it gives rise to is a consequence of one ofthe universal afflictions of humankind: childhood.

Though childhood is a stage through which we all must pass,Descartes finds little to recommend it. It is a time when reason and thesoul are eclipsed by matters corporeal, when we are “governed by ourappetites and by our teachers” (AT VI 13), and when we acquire most

SEMEL IN VITA 231

21 For example, in the quasi-autobiographical account of the origin of his views on the phys-ical world in the Sixth Replies (AT VII 441–442), Descartes makes no real distinctionbetween the commonsense world, and the Scholastic account of the behavior of body. See also AT II 213 where the opinions of the common people are linked to those of “lamauvaise Philosophie,” and AT III 420 where the “vulgar philosophy” is linked to “theearliest judgments of childhood,” where Descartes thinks that the commonsense faith in the senses derives, as we shall see. See also Gilson, Rôle (pp. 168–173), “La psycholo-gie de la physique aristotélicienne.”

22 In seeing the errors of Scholasticism as deriving from the errors of commonsense episte-mology, Descartes does not mean to suggest that the Scholastic metaphysics is a completelyuncritical translation of commonsense sensory beliefs into metaphysics. Contrary tocommon sense, for example, the orthodox Scholastic would deny that there are vacua.

23 For a more explicit development of what is much the same idea, see Nicholas Male-branche, De la Recherche de la Vérité, Bk. 1, chapter 16, and Bk. 6, Part 2, chapter 2, inThomas M. Lennon and Paul J. Olscamp, trans., Nicolas Malebranche: The Search After Truth(Columbus: Ohio State University Press, 1980), 73–75; 440–445.

of the prejudices that cloud the adult mind and make it difficult toapprehend the truth. Descartes writes in the Principles:

And indeed in our earliest age the mind was so immersed in the body thatit knew nothing distinctly, although it perceived much clearly; and becauseit even then formed many judgments, it absorbed many prejudices fromwhich the majority of us can hardly ever hope to become free. (Pr I 47; cf.Pr I 71; AT IV 114; etc.)

Childhood, in Descartes’ view, is the cause of a variety of prejudices.The immersion of the mind in the body, its domination by the imagi-nation and sensation, faculties that, in Descartes’ account, derive fromthe mind’s union with the body, cause us, for example, to confuse theideas we have of the mental with the material, if not ignore the formeraltogether, and make it difficult for us to comprehend the distinctionbetween mind and body.24 This confusion of the mental and the mate-rial is an important prop for the commonsense and vulgar philosophies,the imposition of colors, tastes, and tendencies onto a senseless andunwilling world. But underlying this largely metaphysical confusionand, in a way, leading us directly to it is a basic epistemologic confu-sion. The immersion of the mind in the body, the domination of themind by the corporeal faculties of sensation and imagination, lead usto the unfounded prejudice that those faculties represent to us the waythe world really is. Descartes writes in the Principles:

Every one of us has judged from our earliest age that everything which wesensed is a certain thing existing outside his mind, and is clearly similar tohis sensations, that is, to the perceptions he has of them. (Pr I 66; cf. AT VII 74f.)

Or, as Descartes develops the theme in a later section of the Principles:

In our earliest age, our mind was so allied with the body that it applied itselfto nothing but those thoughts alone by which it sensed that which affectedthe body, nor were these as yet referred to anything outside itself. . . . Andlater, when the machine of the body, which has been so constituted by naturethat it can of its own inherent power move in various ways, turned itself


24 Descartes’ account of sensation and imagination as deriving from the connection betweenthe mind and the body, see, e.g., Pr IV 189–197; Passions of the Soul, 19–26. Sensation andimagination are, for Descartes, both faculties we have by virtue of which we can havemental pictures, and differ only as to whether those pictures derive from the sense organs(sensation) or the brain (imagination). On our early confusion between mind and body,see, e.g., AT III 420; AT III 667; AT VII 441–442.

randomly this way and that and happened to pursue something pleasant orto flee from something disagreeable, the mind adhering to it began to noticethat that which it sought or avoided exists outside of itself, and attributed tothem not only magnitudes, figures, motions, and the like, which it perceivedas things or modes of things, but also tastes, smells, and the like, the sensa-tions of which the mind noticed were produced in it by that thing. . . . Andwe have in this way been imbued with a thousand other such prejudices fromearliest infancy, which in later youth we quite forgot we have accepted withoutsufficient examination, admitting them as though they were of the greatesttruth and certainty, and as if they had been known by sense or implanted bynature. (Pr I 71; cf. Pr I 73)

In our earliest years, then, aware of only what the bodily faculties tellus, but, through our dealings with the world, aware that there are thingsoutside of our immediate control and thus outside of us, we camealmost spontaneously to the belief in a world of external objects similarto our sensations. These judgments became so natural to us, Descartesthinks, that we confused them with the sensations themselves, and wecame to believe that it is our sensory experience itself that gives us thebelief in an external world of sensible properties. As Descartes wrote tothe authors of the Sixth Objections:

In these matters custom makes us reason and judge so quickly, or rather, werecall the judgments previously made about similar things, and thus we failto distinguish the difference between these operations and a simple senseperception. (AT VII 438)25

In this way, we come, in our adulthood, to put our trust in the sensesas an accurate representation of the way the world is.

The prejudice in favor of the senses, the belief that the senses rep-resent to us the way the world of bodies really is, gives rise to a multi-tude of prejudices, as this passage suggests. In an obvious way, it leadsus to think that “seeing a color, we saw something which existed outsideof us and which clearly resembled the idea of that color which we thenexperienced in ourselves” (Pr I 66). Similarly, when we have a painfulor pleasant sensation, this epistemological prejudice leads us to believe

SEMEL IN VITA 233

25 The discussion of sensation in the Sixth Replies from which this passage is excerpted makes it clear that, strictly speaking, the prejudice for the senses which for Descartes ischaracteristic of common sense is a prejudicial judgment about the cause or content ofour ideas of sensation. That is, what is wrong is the judgments we make about sensoryideas; the ideas themselves, Descartes is clear, are neither true nor false.

that it is “in the hand, or in the foot, or some other part of our body”(Pr I 67. Cf. Pr I 47, 68). Such prejudices also lead us to posit vacuawhere we sense no objects (cf. Pr II 18; AT V 271), to think that moreaction is required to move a body than to bring it to rest (Pr II 26), thatbodies in motion tend to come to rest (Pr II 37), and that bodies havean internal resistance to motion (AT II 213–214). And with these prejudices, which make up what I earlier called the commonsenseworldview, we have laid the groundwork for Scholastic physics, the meta-physicalization of this commonsense world of sensible properties andtendencies and the positing of forms and qualities.

At this point we can turn back to the Meditations. These prejudices,grounded in the epistemological prejudices of youth are, I think, chiefamong the “many false things I have admitted as true from my earliestage” that Descartes has in mind in the opening sentence of MeditationI, and one of the chief purposes of the Meditations is to eliminate thoseprejudices and replace them with a true picture of the way the worldis. It is in this sense that the Meditations is intended to lay the founda-tions for Cartesian science and eliminate the foundations of the Aris-totelian. But in order to overturn these prejudices, we must find a wayof setting aside the prejudice for the senses that we have had sinceyouth, and replace our dependence on the senses with an altogetherdifferent epistemological principle; Descartes’ revolution in physicsmust begin with a revolution in epistemology. It is in this sense thatDescartes holds that “once in life [semel in vita] everything ought to becompletely overturned, and ought to be rebuilt from the first founda-tions,” from our epistemology up, “if I want to build anything firm andlasting in the sciences,” if I want to find out how the world really is. Thisproject, as Descartes carries it out in the Meditations, involves two prin-cipal stages. We must first break the hold of the senses and of all theprior beliefs we have held that are based on our faith in the senses; whatis called for is a kind of intellectual infanticide, to use Gouhier’s some-what violent image, the elimination of the child that remains withinus.26 This, I shall argue below, is one of the important functions of Meditation I. And, second, we must carefully reexamine the epistemo-logical foundations of knowledge, and replace our exclusive depen-dence on the senses and the imagination with a more sophisticated viewof knowledge that puts the senses in their proper place and subordi-


26 See Henri Gouhier, Pensée Métaphysique, p. 58.

nates them to another cognitive faculty, which will allow us to establishthe real nature of the world, as opposed to how it appears to us –extended stuff, nonextended rational souls, and God – and will allowus to set aside the prejudices of common sense and the errors of theScholastic philosophy. This, I shall argue, is one of the important func-tions of the remaining Meditations.

Skeptical Therapy

So far I have concentrated on the opening sentence of the Meditations,and offered an interpretation of it in terms of what Descartes wroteoutside of the Meditations itself. I have argued that the Meditations mustbe read not merely as a philosophical project to defeat skepticism but,more generally, as an epistemological preparation for science. It is nowtime to turn to the Meditations themselves and work out some of thedetails of the reading I propose. The first question to be taken up mustbe the skeptical arguments of Meditation I. Meditation I seems toannounce skepticism as the problem of the Meditations. But, I claim, itdoes more than that. Meditation I, I claim, is the first step in buildinga new epistemology, the destruction of the prejudice in favor of thesenses and in favor of the closely related faculty of imagination, whichconstitutes a necessary first step in the construction of an epistemologyappropriate for Cartesian science.

In the previous section, I emphasized Descartes’ account of our intel-lectual development, our initial trust in the senses and the conceptionof the world that grows out of it. These prejudices, Descartes thinks,interfere with our perception of the way things are, and must beremoved before we can find true and certain knowledge; we must, asDescartes puts it, withdraw our minds from the senses, from the body,from the things we formerly believed. And so Descartes wrote in theSecond Replies that, even though the account of the foundations of theworld that he is attempting to outline in the Meditations is, to the openmind, even more obvious than geometry,

yet being contradicted by the many prejudices of our senses to which we havesince our earliest years been accustomed, they cannot be perfectly appre-hended except by those who give strenuous attention and study to them, andwithdraw their minds as far as possible from bodily matters. (AT VII 157; cf. AT I 350–351; AT IV 114; AT VI 37)

SEMEL IN VITA 235

And thus Descartes tells a correspondent in 1638, “those who want todiscover truth must distrust opinions rashly acquired in childhood” (AT II 39).27

It is with this in mind that we should approach the skeptical argu-ments that are the main business of Meditation I. Much of the contentof the Meditations had been made public some four years earlier, in Part4 of the Discourse. But some readers had problems following the argu-ments there. Part of the problem derived from the brevity of treatmentin the Discourse. But Descartes acknowledged another problem as well.In a letter written in 1637, Descartes sympathizes with one such readerand confesses that “there is a great defect in that work you have seen,and I have not expounded the arguments in a manner that everyonecan easily grasp.” But, Descartes continues,

I did not dare to try to do so, since I would have had to explain at length thestrongest arguments of the skeptics to show that there is no material thingof whose existence one can be certain. Thus I would have accustomed thereader to detach his thought from sensible things. (AT I 353)

In the Meditations the defect is corrected, and Descartes rehearses atsome length important skeptical arguments missing in the earlierwork.28 And one prominent reason he gives for doing so is the reasonhe suggested to his earlier correspondent. It is this motivation, the ther-apeutic value that skeptical arguments have in eliminating prejudice,that Descartes emphasizes in the synopsis he gives of Meditation I:

In the First Meditation, I present the reasons why we can doubt generally ofall things, and particularly of material things, at least as we have no otherfoundations of the sciences than those that we have had up until now. Eventhough the utility of such a general doubt is not apparent at first, it is,however, quite considerable, since it delivers us from all sorts of prejudices,and prepares for us a very easy way to accustom our mind to detach itselffrom the senses. (AT VII 12).


27 This claim offers Descartes an interesting reply to objectors not convinced by his argu-ments. Descartes can claim that his objectors are still dominated by the prejudices ofyouth, and for that reason cannot see what is present to their mind’s eye (cf., e.g., AT III267 and AT VII 9–10).

28 In a sense, all of Part 1 of the Discourse can be read as a skeptical argument. But the explicitskeptical arguments that occupy a full Meditation in the later work occupy just a few linesin Part 4 of the Discourse. Furthermore, the Meditations contains two arguments missing inthe Discourse, the deceiving-God argument and the hypothesis of the evil demon.

Similarly, when Hobbes grumbled that he “should have been glad if ourauthor . . . had refrained from publishing these matters of ancient lore”(AT VII 171),29 Descartes replied that the arguments were put therequite deliberately. The skeptical arguments that open the Meditationsset up questions answered in the course of the work, and provide a stan-dard of certainty for later arguments, Descartes explains. But the veryfirst reason Descartes gave Hobbes for rehearsing them at such lengthis that they “prepare the minds of the readers to consider intellectualthings and distinguish them from corporeal things, for which thosearguments always seemed necessary” (AT VII 171–172). The separationof the intellectual from the corporeal is an obvious reference to the dis-tinction between mind and body, which is one of the central conclu-sions of the Meditations. But it also refers to the distinction between theintellectual faculties and the corporeal, between reason and sense,whose confusion underlies the confusion between the mental and thematerial, as noted earlier. The skeptical arguments of Meditation I,then, are to eliminate prejudice and prepare us to see things as theyare, as reason, the intellect sees them, as opposed to the way thingsappear to us through our corporeal faculties.

And it is clear when one reads Meditation I that the trust in our cor-poreal faculties, our senses, the most fundamental prejudice we havefrom youth, is a central focus of Descartes’ attention. The task of Med-itation I, as Descartes puts it, is the “general overthrow of my opinions”(AT VII 18). This task is to be accomplished not by eliminating themone by one, as his later metaphor of the apple basket suggests (cf. ATVII 481), but by eliminating the foundations on which all those preju-dices rest:

Since the destruction of the foundations by itself brings about the downfallof that which is built on it, I shall now attack only those principles on whichall that I once believed rested. (AT VII 18)

And these foundations, these principles, are epistemological, Descartesthinks – the most prominent being the faith we have had in the verac-ity of the senses. Descartes seems to begin by eliminating, first of all,the epistemic principle in accordance with which everything we learnfrom the senses is trustworthy. This epistemic principle is easily set aside

SEMEL IN VITA 237

29 It is interesting to note that when, somewhat later, Hobbes presented his own philosophy,he made use of a device similar to hyperbolic doubt, though without the full battery ofskeptical arguments that Descartes used. See De corpore, chapter 7, sec. 1.

with the observation that the senses sometimes deceive (cf. AT VII 18).Descartes then continues with the consideration of a second epistemicprinciple, again concerned with sensory knowledge:

But it may be that although the senses sometimes deceive us concerningcertain things that are small or remote, there are yet many others about whichwe clearly cannot doubt, although we take them in by their means [i.e., bymeans of the senses]. (AT VII 18)

This second principle, a guarded statement of the first, limiting thetrustworthiness of the senses to middle-sized objects in our immediatevicinity, is eliminated by means of Descartes’ celebrated dream argu-ment, however precisely it is taken to work.30

So far we have been dealing with purported knowledge acquireddirectly from the senses. But at this point, the argument takes an inwardturn: “Now let us assume that we are asleep,” Descartes says, “and thatall these particulars, e.g., that we open our eyes, shake our head, extendour hands, indeed that we have such hands, or such a body, are false”(AT VII 19). Under this assumption Descartes formulates a third anda fourth epistemic principle. The third is suggested in the followingpassage:

At the same time we must at least confess that those things which are seen[visus] in sleep are like certain painted images [imagines] which can only havebeen formed as things similar to real things; and therefore these generalthings, eyes, head, hands, and the whole body, are not [merely] imaginary,but really exist. (AT VII 19)

Here the suggestion is that dream images are formed of componentsthat correspond to things that exist at some indefinite time in a realexternal world, even if they may not be arranged in the real world asthey are in our dreams. So, one would claim on the authority of thisprinciple, one can establish the sorts of things there really are in theworld on the basis of our dream experience. This third epistemic prin-ciple, though, quickly gives way to a fourth:

And for the same reason, although these general things, eyes, head, hands,and the like, may be [merely] imaginary, we must however confess thatcertain other things, yet more simple and universal, are real, from which our


30 See Margaret Wilson, Descartes (chapter 1, secs. 4 and 6) for an account of the main linesof interpretation in the literature and an extremely plausible proposal for reconstructingthe argument.

images [imagines] of things, whether true or false, are made, just as all ofthem are made from true colors. The nature of body in general, its exten-sion, and also the shapes of the extension of things, quantity . . . , the placein which it exists, and the time through which it endures, and the like seemto be of this sort. (AT VII 20)

And, Descartes continues:

That is why we will perhaps not be reasoning badly if we conclude that physics,astronomy, medicine, and all the other sciences which depend on the con-sideration of composite entities are, indeed, uncertain whereas arithmetic,geometry, and the other sciences of this sort, which treat only of the simplestand most general things, without sufficiently concerning themselves [parumcurant] about whether they occur in nature or not, contain something certainand indubitable. For whether I am awake or whether I am asleep, two andthree together will always make the number five, and the square will neverhave more than four sides; and it does not seem possible that truths so evident[perspicuae] can ever be suspected of any falsity. (AT VII 20)

The epistemic principle here is somewhat tricky. The idea seems to bethat even if our dream experience does not inform us about the worldas it is now, or give us access to things like eyes and hands, it does displaythe most general features of the real world, extension, time, and,perhaps, color, the elements “from which our images of things . . . aremade,” just as, in the third principle, eyes, hands, and the like are con-sidered as the elements of our dream images and, for that reason, aresupposed to be real. When the meditator takes this principle to under-lie the certainty of geometry and arithmetic, it is not because these dis-ciplines concern truths that are wholly independent of the real world.Rather, they can be certain in spite of the fact that they don’t give suf-ficient consideration to the question of whether or not circles and tri-angles exist in nature, and they can get away with this lapse in argumentbecause the fourth principle of evidence guarantees that such objectsdo exist. This is why geometry and arithmetic are on better foundationsthan physics, astronomy, and medicine, the existence of whose objectsis doubtful under the assumption that we are dreaming.31

SEMEL IN VITA 239

31 Cf. Frankfurt’s explication of this passage in Demons (pp. 74–76). While I am not certainthat I can agree with everything he says about the treatment of mathematics in Medita-tion I (see Demons, chapters 7–8), I do agree with Frankfurt that, in the particular pas-sages under consideration, the meditator is clearly thinking of mathematics as a sciencethat pertains to certain features of an external world. (In general, I should point out thatmy account of the arguments in Meditation I owes an obvious debt to Part 1 of Demons.)

While the third and fourth principles of evidence may appear to havelittle to do with our faith in the senses, there is, in Descartes’ mind, anintimate connection. What we are dealing with in these two principlesis sensation taken in the very most general way, knowledge about themakeup of the world which we claim to derive from sensory images,considered apart from any assumptions about their immediate causalhistory; what we are dealing with are our mental images of bodies con-sidered irrespective of the question as to whether they are genuine sen-sations, images that derive from the sense organs, or mere imaginations,mental pictures that we have from some other source.32 In this way,then, when the final skeptical argument of this series, the deceiving-God argument, eliminates the third and fourth epistemic principles, iteliminates (among other things, perhaps) what appears to be the lasthope of knowledge from the senses and, more generally, from all thecorporeal faculties, both sensation and imagination, those that arisefrom our connection to our body.33

The skeptical arguments of Meditation I, then, are carefully directedat the youthful prejudice in favor of the senses that, Descartes argues,must be eliminated before we can attain real knowledge. Their func-tion as therapy intended to withdraw the mind from the senses is under-scored in the final pages of the Meditations. Descartes writes:

But it is not sufficient to have made these remarks; we must also be carefulto keep them in mind. For these habitual opinions will still frequently recurin my thoughts, my long and familiar acquaintance with them giving themthe right to occupy my mind against my will. (AT VII 22)


As strange as this account of mathematics may seem, it may well have been the way inwhich Descartes himself thought about mathematics in his earlier years. On this see JohnA. Schuster, “Descartes’ Mathesis Universalis: 1619–28,” in Stephen Gaukroger, Descartes,41–96. Schuster’s suggestion seems to be that in the late 1620s, Descartes thinks of theprimary objects of mathematics as imaginative pictures, mental representations of physical impressions in the brain.

32 Dream images, under consideration in this context, are, for Descartes, a kind of imagi-nation, one that derives from the activity of the brain. (See Passions of the Soul, 21, 26.)But this portion of the argument is intended, I think, to deal with imagination of all varieties, both involuntary (like dreams) and voluntary, all of which derive from the unionof mind with a body, and any knowledge that we may claim to have from imagination.

33 I don’t mean to suggest that this is the only thing that the deceiving-God argument elimi-nates. It is plausible to read it, as most commentators have, as eliminating all the medi-tator’s former beliefs, both those from the senses as well as those beliefs he may have hadfrom reason, thus setting up the problem that is ultimately resolved in Meditation IV withthe validation of clear and distinct perception. But this is a question that relates to thefunction of Meditation I as setting skeptical questions to be answered in the later Medi-tations, a question that goes beyond the scope of my interest in Meditation I here.

It is for this reason that Descartes proposes the famous demon hypoth-esis, to fix the elimination of prejudice that the skeptical argumentsbegan:

I will therefore suppose that not the best God, who is a fountain of truth, butsome malignant demon, no less deceitful than powerful, has bent all hisefforts to deceive me. (AT VII 22)34

The role of Meditation I as skeptical therapy directed against theprejudices of youth is also underscored by the indications he givesreaders as to how the arguments of Meditation I should be read. In theappendix to the Second Replies, the arguments from the Meditationsdrawn up modo geometrico at the request of the objectors, the followingpostulate (postulata – literally a request or demand) is what correspondsto Meditation I:

The first request I press upon my readers is a recognition of the weakness ofthe reasons on account of which they have hitherto trusted their senses, andthe uncertainty of all the judgments that they have based on them. I beg them to turn this over in their minds so long and so frequently that they willacquire the habit of no longer reposing too much trust in them. For I deemthis necessary in order to attain to a perception of the certainty of thingsmetaphysical. (AT VII 162)

This strongly suggests that the arguments of Meditation I are to betreated in the same way, meditations in the truest sense of the term,exercises to practice in order to free the mind from prejudice.35 Wemust take very seriously the remarks Descartes made about MeditationI in the Second Replies when he wrote:

Nothing conduces more to the obtaining of a secure knowledge of realitythan a previous accustoming of ourselves to entertain doubts especially aboutcorporeal things. . . . I should be pleased also, if my readers would expendnot merely the little time which is required for reading it, in thinking overthe matter of which the Meditation treats, but would give months or at least

SEMEL IN VITA 241

34 In presenting the demon hypothesis in this way, as a therapeutic device, I mean to rejectMartial Gueroult’s celebrated reading, in accordance with which the demon hypothesis isintended as a genuine argument, distinct from the deceiving-God argument. Accordingto Gueroult, the demon argument is answered by the causal argument for the existenceof God in Meditation III, whereas the deceiving-God argument is answered by the ontological argument in Meditation V. For a clear and concise statement of Gueroult’sposition, together with the defense of a position much like the one I am advancing here,see Henri Gouhier, “L’Ordre des Raisons selon Descartes,” in Cahiers du Royaumont:Descartes (Paris: Éditions de minuit, 1957), 72–87.

35 Cf. Gilson, Rôle, 186.

weeks, to this, before going on further; for in this way the rest of the workwill yield them a much richer harvest. (AT VII 130)

Descartes is also quite serious when he says in the introduction to theMeditations (to what must have been his publisher’s dismay) that “I don’trecommend reading this to anyone except those who want to meditateseriously with me, and who can detach their minds from the senses, anddeliver them from all kinds of prejudices” (AT VII 9; cf. AT IXA 1, 3 andAT VII 157–159). To fail in this way, to simply read the Meditations ratherthan meditate with Descartes is to miss the point of the book.

The skeptical arguments of Meditation I do, indeed, set up argumentsfor Descartes to answer in the course of the Meditations. But they are alsoexercises we must undertake before beginning science, a necessaryprelude to constructing an epistemology that will lead us to knowledgeof things, not as they appear to us, but as they really are.36

Reason, Sense, and Imagination

The elimination of our infantile prejudices in favor of the senses is anessential first step toward the true science, Descartes thinks. But it is onlya first step. Descartes must then replace the conception of knowledgerejected in Meditation I with a different foundation for knowledge, onethat will allow us to see things for what they really are. The new founda-tion is, of course, clear and distinct perception, the light of reason, a lightcapable of illuminating the mind without the aid of the sensory organs.Reason will show us the true nature of body, extension, and extensionalone, offering us the means to begin rebuilding the world well lost inMeditation I, a brave new world without color or sound, taste or heavi-ness, form or real quality, a world not without sense and imagination,but a world in which the light perceptible to the eye is properly subor-dinated to the light directly perceptible by the mind. The validation ofreason must, then, be a central project in the Meditations and its trust-worthiness as important to the theme I am stressing, the laying of epis-temological foundations for the sciences, as it is for the theme moreoften emphasized, the refutation of skepticism. But, it is important tonote, the defense of reason against skeptical attack is not the only epis-temological project Descartes undertakes in the Meditations. In the body


36 For similar readings of the aim of Meditation I, see Gilson, Rôle, 184–190; Gouhier, PenséeMétaphysique, chapter 2; Wilson, Descartes, chapter 1, sec. 3; and Mike Marlies, “Doubt,Reason, and Cartesian Therapy,” in Michael Hooker, ed., Descartes: Critical and InterpretiveEssays (Baltimore: Johns Hopkins University Press, 1978).

of the Meditations there is, I claim, a series of arguments very carefullyand systematically directed against the commonsense prejudice for thesenses; the disease treated by skeptical therapy in Meditation I is sub-jected to the light of reason in the Meditations that follow. The carefultreatment of the prejudices of the senses throughout the epistemologi-cal discussions of the Meditations demonstrates, I think, that Descarteswas concerned not only with the refutation of skepticism, but with theelimination of a false epistemology and its replacement by the true, withthe elimination of the commonsense dependence on the faculties of sen-sation and imagination that lead toward Aristotelianism, and theirreplacement with a conception of knowledge appropriate to groundingthe new, mechanical philosophy.

Before entering into these arguments, it will be helpful to say some-thing about how the Meditations is written. The authors of the SecondObjections asked Descartes to set out his principal arguments more geo-metrico, with formal definitions, postulates, axioms, and with carefulformal proofs (cf. AT VII 128). While Descartes complied with theirrequest (cf. AT VII 160–170), he was not entirely comfortable doingso. For reasons obvious from the discussion of the previous section, hetold the objectors that, while the Euclidean mode of exposition is finefor geometry, it is unsuited to the material at hand, which requires, forits proper comprehension, the therapeutic withdrawal from the senseswhich, I have argued, is an important function of Meditation I. It is inthis context that Descartes tells his objectors a bit about how the Med-itations themselves were written. “In my Meditations I have followed onlyanalysis, which is the true and best way for teaching,” Descartes wrote(AT VII 156). Analysis, in contrast to the more usual mode of argument(ratio demonstrandi) in geometry, what Descartes calls synthesis, is pre-sented as the mode of argument that shows “the true way by which athing was methodically and, as it were, a priori discovered” (AT VII155).37 Descartes’ conception of analysis and its precise distinction fromsynthesis is obscure and has given rise to much discussion.38 But an

SEMEL IN VITA 243

37 “A priori” seems meant in an epistemic rather than in the usual metaphysical sense. Fora short discussion of the textual problems this sentence raises, see Daniel Garber andLesley Cohen, “A Point of Order: Analysis, Synthesis, and Descartes’ Principles,” Archivfür Geschichte der Philosophie 64 (1982):136–147, esp. n. 5, and the references cited there.This appears as essay 3 in this volume.

38 For discussions of Descartes’ conception of analysis and synthesis, see Garber and Cohen,“Point of Order,” and E. M. Curley, “Analysis in the Meditations,” in Amélie Rorty, ed., Essaysin Descartes’ Meditations (Berkeley and Los Angeles’. University of California Press, 1986).

examination of the six Meditations shows one clear sense in which theycan be read analytically, showing how one might actually come to dis-cover for oneself the conclusions reached. Actual discovery involvesfalse steps as well as true, bad arguments considered and rejected aswell as good arguments that ultimately lead to enlightenment. This, Iclaim, is an important aspect of the expository strategy of the Medita-tions. In the course of reestablishing the epistemic foundations ofknowledge and showing the inadequacy of common sense, Descartesallows the commonsense bias for the senses to have its turn at trying toestablish the way the world is. As a consequence, woven through thetexture of positive arguments in the Meditations is a genuine dialoguebetween the claims of common sense and the claims of reason, betweenthe prejudices of youth and the wisdom of Cartesian maturity, a dia-logue all too easily missed by the reader who focuses too closely on thevalidation of clear and distinct perception and the refutation of skep-ticism. While the dialogue pervades much of the text, I would like toemphasize two important exchanges, the discussion of the wax examplein Meditation II, and the aborted proof for the existence of body fromour adventitious ideas of sensation in Meditation III, before showing insome detail how the claims of sensation and reason are finally resolvedin the discussion of the existence of body that Descartes presents inMeditation VI and in the discussion of the teachings of nature thatimmediately follows.

The path to knowledge begins, of course, in Meditation II, with thecogito and the sum res cogitans, arguments that establish the existence ofthe knowing subject as a thinking thing, “the first and most certain ofall that occurs to one who philosophizes in an orderly way,” as Descartesputs it in the Principles (Pr I 7). But as soon as that first step in the argu-ment is taken, there is an objection from common sense: Certainlybodies, things that we are acquainted with by way of the corporeal fac-ulties of sensation and imagination, are better known to us than themind, which can be conceived through neither of those faculties. Asthe meditator puts it:

But nevertheless it still seems to me and I cannot keep myself from believingthat corporeal things, images [imagines] of which are formed by thought, and which the senses themselves examine, are much more distinctly knownthan that something I know not what of myself which does not fall under theimagination. (AT VII 29)

Descartes’ response to this objection, which he considers natural in thefullest sense, is to let the mind wander and consider what it is that it


really knows and how it knows it. Through the consideration of the cel-ebrated piece of wax, Descartes tries to show us that, contrary to theprejudices of youth as embodied in common sense, neither sensationnor imagination gives us access to the nature of bodies, and that what-ever we are able to learn about the existence of body through the sensesand imagination, our knowledge of the existence of mind is prior tothat of body in a well-defined sense, despite the inaccessibility of mindto the faculties of sensation and imagination.

Descartes begins the response to common sense by pointing out thatsensation cannot give us the distinct comprehension of the piece of waxunder consideration, since all of its sensible properties can change,while the wax itself remains the same; put it by the fire, melt it, and“what remains of the taste evaporates, the odor vanishes, its colorchanges, its shape is lost, its size increases,” and so on (AT VII 30). Thus,the nature of the wax, what makes it the thing it is, what persists throughchange, is inaccessible to the senses. But it is also inaccessible to theimagination. “Rejecting everything which does not belong to the wax,”Descartes suggests that the wax itself, what persists through the sensi-ble changes, is just “something extended, flexible, movable” (AT VII31). But if so, then I conceive the wax as something able to take on aninfinite number of shapes, round, square, triangular, and everything inbetween, an idea that goes beyond my imagination, the capacity I havefor forming mental pictures (cf. AT VII 31). Although the full consid-eration of the nature of body will have to await Meditation V, Descartesat this stage believes that he is entitled to conclude that it is an inspec-tion of the mind alone (solius mentis inspectio) that reveals the nature ofthe wax (AT VII 31), that it is the mind itself, working apart from thebody-connected faculties of sensation and imagination that allows us to“distinguish the wax from its external forms, and consider it as if naked,having removed its clothing” (AT VII 32). “It is only the prejudices ofyouth, and later habits derived from those prejudices, that could con-vince me otherwise” (AT VII 31–32).39

It is thus no challenge to the priority of the cogito and sum res cogi-tans argument to say that the object under consideration, the mind, isnot known through the senses. Neither is the body. And at this point itis easy for Descartes to establish his second claim, that the knowledge

SEMEL IN VITA 245

39 Descartes claims here that we are deceived by our ordinary ways of speaking into think-ing that we literally see the wax, just as we are, strictly speaking, mistaken when we say thatwe see people in the street below the window; if we can be said to see anything at all, itis coats and hats.

of the existence of the mind is prior to the knowledge of the existenceof body. For whether or not the wax I see or imagine really exists, thesensations or imaginations themselves entail that mind exists; andhowever sensory experience may serve to establish the existence ofbody, any such sensory experience demonstrates, with certainty andimmediacy, the existence of a mind (cf. AT VII 33). It is in this precisesense that mind is “better known than the body,” as Descartes puts it inthe title of Meditation II.

The wax example tames the unruly prejudices of childhood, but onlytemporarily. Although the meditator seems to accept the priority ofknowledge of mind over knowledge of body, he keeps on pressing theinsistent claims of commonsense knowledge of body. Meditation IIIbegins with a kind of introduction, where Descartes reflects on the conclusions reached so far and lays out the strategy of the argument to come. Descartes tells us that having established the existence of theknowing subject, we must establish the existence and nature of itscreator before anything can be known for certain (AT VII 34–36). Butafter this introduction, Descartes returns to the argument proper: “andnow good order seems to demand that I should first classify all mythought into certain types and consider in which of these types thereis, properly, truth or falsity” (AT VII 36–37). That is, having establishedthe existence of mind as a thinking thing, we must see what can bedrawn from an examination of the thoughts themselves. And at thispoint, almost as soon as the order of argument is resumed, the claimsof common sense assert themselves again. The meditator againattempts to show that the senses lead us directly to a knowledge of theexternal world of bodies.

The meditator begins by distinguishing ideas, properly speaking,thoughts that are like images of things (tanquam rerum imagines) insofaras they are representative, unlike volitions or emotions (AT VII 37).40

The ideas are then broken down into three categories, the innate ideasthat seem inborn, the adventitious ideas that seem to come fromwithout, and the factitious ideas that seem to have been created by me(cf. AT VII 37–38). Of particular interest to Descartes’ meditator are


40 This passage naturally enough misled Hobbes into thinking that all ideas are mentalimages for Descartes, i.e., that the only cognitive faculties are sensation and imagination.In his reply to Hobbes, Descartes is clear that this is not the intention in this passage. (SeeAT VII 179–181.) Ideas are tanquam rerum imagines only insofar as ideas and images arerepresentative of things other than themselves.

the adventitious ideas: “If I now hear some noise, if I see the sun, if I feel heat, I have hitherto judged that these sensations proceeded from some things which exist outside of myself . . . and resemble thoseobjects” (AT VII 38). The reasons given for these commonsense judg-ments are three: (1) Nature seems to teach me so; (2) the sensations Ihave are independent of my will; and (3) “nothing is more plausible[obvium]” than that the external thing imposes its own likeness (simi-litudo) on me rather than anything else (AT VII 38; cf. AT VII 75–76).The claim is a familiar one; it is, in essence, a reprise of the second prin-ciple of evidence from Meditation I, the claim that our senses give usaccess to the familiar world of middle-sized bodies around us.

And, once again, common sense is rejected. Descartes first of all dis-tinguishes the teachings of nature from the light of nature, the facultyto which he will appeal in his own argument for the existence of God,later in that same meditation (AT VII 40). The teachings of nature,what causes me to judge that my sensory ideas derive from body, is amere inclination to believe (quodam impetu . . . ad hoc credendum) ratherthan an irresistible impulse to belief, a faculty, like the light of nature,which is indubitable in the sense that “there can be no other facultywhich could teach me that what this light of nature shows me as trueis not so, and in which I could trust as much as in the light of natureitself” (AT VII 38–39).41 And, Descartes notes, since natural inclinationshave in the past led me astray in distinguishing good from bad, I shouldnot trust them uncritically in this case either (cf. AT VII 39). As for thefact that sensations are involuntary, this too is insufficient reason forthinking that they derive from something external to us. Recalling thedream argument of Meditation I, Descartes suggests that, for all I know,there might be some faculty in me independent of my will that I donot, at this point, know of, which is responsible for my present sensoryideas, without the need for an external cause, just as some such facultymay cause the dream experiences I have in sleep (AT VII 39).42 And

SEMEL IN VITA 247

41 It is interesting here that Descartes says of the light of nature at this stage, before it hasbeen validated in Meditation IV, that it “shows me that which is true.” The teachings ofnature seem to be the customary judgments connected with the senses from early youth,which we have mistaken for direct deliverances of the senses. (See AT VII 436–439 andthe discussion above in Part 1.)

42 This objection relates closely to the dream argument of Meditation I as interpreted inWilson’s Descartes (chapter 1, sec. 6). According to Wilson, in Meditation I Descartes isnot worried about our supposed inability to tell whether or not we are awake. Rather, sheclaims, the question is why, since we are not tempted to think that our dream experiencesrepresent an external reality, do we think our waking experiences are any different on

finally, Descartes challenges the claim that our ideas of sense resembleexternal things, even if it is conceded that they are caused by thingsexternal. The claim is that the idea we have of an object from our sensesis different from the idea we have of the same object through reason(“certain innate ideas”), and that “reason persuades me that the ideathat seems to come directly from the thing is that which least resem-bles it” (AT VII 39). Although the example he uses here is different(the sun as regarded by the senses and by the astronomers), the pointhere is largely the same as the one he made earlier in connection withthe wax example.

At this point in the argument, Descartes puts aside the ideas of senseand the question of external bodies, and initiates a train of argumentthat leads in a fairly direct way to the existence of God in MeditationIII and to the validation of reason in Meditation IV. While the preju-dices of youth are addressed on a number of occasions in the courseof these arguments,43 it isn’t until Meditation V that the question of our knowledge of the external world is addressed again and, in fact,becomes the focus of Descartes’ attention:

having noticed what must be avoided or done in order to arrive at the knowl-edge of the truth, my principal task now is to attempt to escape from thedoubts into which I have fallen in these last few days, i.e., in the previous[Meditations] and to see if we can know anything certain about materialthings. (AT VII 63)


this score? The objection to the commonsense reason for believing in an external worldis similar. The fact that our sensations are involuntary is no indication that they proceedfrom something external, since our dream experience, which is also involuntary, may notrequire an external cause.

43 In the course of the causal argument for the existence of God in Meditation III Descartesexamines the ideas of sense, attempting to persuade common sense that these ideas areobscure and confused. (See AT VII 43–44 and Wilson, Descartes, chapter 3, sec. 2.) Laterin that same Meditation, after concluding the first causal argument for the existence ofGod, Descartes raises and answers three objections (AT VII 45–47). At least two of thoseobjections depend on the belief that any ideas we have of infinity and perfection mustderive from ideas we have of the finite and imperfect, a belief quite natural to the sen-sualist, for whom every idea derives from sensory experience of finite and imperfectthings. And in Meditation IV, the question of the grounds of our belief in corporeal thingscomes up in the course of an analysis of error (AT VII 58–59). Descartes there contraststhe spontaneous and irresistible urge to believe in the existence of the mind with the lessstrong inclination to believe in body, arguing that only the former is a proper use of thefaculty of judgment, in a passage that recalls the distinction between the light of natureand the teachings of nature in Meditation III.

Descartes begins the project in Meditation V, with an account “of theessence of material things” (AT VII 63), an explicit statement of someideas first introduced in the wax example of Meditation II. The “some-thing extended, flexible, movable” that is the nature of wax, inaccessi-ble to sensation or imagination (AT VII 31), is identified in MeditationV with the nature of body itself: what I “imagine distinctly” in bodiesand what must, therefore, constitute their nature is “extension inlength, width, and depth,” together with various modes that pertain toindividual extended things, “sizes, shapes, positions, and motions”together with duration (AT VII 63).44 This settles part of the debatebetween reason and common sense; the argument of Meditation V isintended to give us a definitive refutation of the commonsense claim,prominent in earlier discussions, that bodies resemble our sensory ideasof them. And once this question is settled, Descartes turns to the ques-tion of the existence of material things in Meditation VI. Here, though,the treatment of common sense is more subtle. Like common sense,Descartes believes in the existence of bodies external to the mind, eventhough his geometrical conception of body is quite distant from thesensuous world of common sense. And, like common sense, Descartesbelieves that sensation and imagination and the teachings of naturehave roles to play in our coming to believe in a world of bodies, eventhough the roles they play are importantly different from the rolesassigned them by common sense. The final move in the dialoguebetween truth and prejudice is not so much a refutation of commonsense as it is a reinterpretation, an attempt to find what is right incommon sense and show how at least some of our youthful convictionscan find their place in the Cartesian system. It is in this way that thedialogue is finally concluded.

The final reconciliation begins in Meditation VI with a considerationof imagination, and a discussion of the extent to which imagination canestablish the existence of an external world of bodies. Imagination,Descartes points out, the faculty we have for forming mental pictures, issomething quite different from pure intellection, the faculty we have for

SEMEL IN VITA 249

44 Duration seems new here. It is also interesting that the starting place for Descartes’ analy-sis of the idea of body is an idea of imagination, a mental picture of body. He begins theanalysis with the claim that “ I imagine quantity distinctly” (AT VII 63). When a shortwhile later an example is considered (AT VII 64), it is the idea of a triangle in imagina-tion (“And when, for example, I imagine a triangle . . .”).

grasping the concepts of, say, geometrical objects in a nonsensuous way.This faculty, Descartes claims, “is in no way necessary to my essence” and thus “depends on something other than myself” (AT VII 73). And, Descartes continues, appealing to a conviction, presumably fromcommon sense, a conviction suggested earlier in Meditation II, “I readilyconceive that if some body exists with which my mind is so joined that itcan consider it whenever it wishes, it could be by this means that it imag-ines corporeal things” (AT VII 73; cf. AT VII 28).45 The passage con-cludes with an appeal to an argument from the best explanation: “I easilyconceive, I say, that the imagination can work in this fashion, if indeedbodies exist, and because I cannot find any other way in which this canbe explained equally well, I therefore conjecture that bodies probably[probabiliter] exist” (AT VII 73). But, Descartes notes, “this is only probable, and although I carefully consider all aspects of the question,I nevertheless do not see that from this distinct idea of the nature of body which I find in my imagination, I can derive any argument whichnecessarily proves the existence of any body” (AT VII 73).46 The facultyof imagination can, indeed, lead us to a belief in body, Descartes seemsto concede. But, the claim is, not in the way that common sense mightoriginally have thought. The argument Descartes offers is very differentfrom the kinds of arguments suggested in the third and fourth princi-ples of evidence in Meditation I, where the meditator suggested that aconsideration of our dream experience, a variety of Cartesian imagina-tion, might give us access to some of the general features of an externalreality. The argument Descartes offers to common sense in its place isan argument from the very faculty of imagination, and Descartes sug-gests, somewhat dogmatically, this is the only argument we are to get fromimagination. Furthermore, the consideration of imagination will give usonly probability, only a belief in the plausibility of an external world, andnot the real conviction that we thought we had.


45 The passage from Meditation II reads: “to imagine is nothing but to contemplate the shapeof a body or its image.” It is not clear here whether Descartes is claiming that imagina-tion is the contemplation of a mental picture, or the contemplation of a physical picturein the brain (the pineal gland), which is the cause of the mental picture. (Cf. the accountof sensation and imagination in the Treatise of Man [AT XI 174–177], in Thomas SteeleHall, trans., René Descartes: Treatise on Man [Cambridge: Harvard University Press, 1972],84–87. In Descartes’ account there, sensation and imagination are literally the contem-plation of a shape in the pineal gland, a shape that is isomorphic to an external body inthe case of sensation.)

46 The conclusion here is probable, plausible rather than certain, because of the hypo-thetical form of argument, presumably.

For real certainty we must turn to the senses, Descartes thinks. Thefinal argument for the existence of external bodies, the argument thatDescartes finally and unambiguously endorses is an argument thatappeals crucially to sensation and the teachings of nature, considera-tions prominent in the abortive Meditation III argument, but now usedin a way that is not open to the objections raised earlier. The argumentgoes as follows.47 I find in myself “a certain passive faculty of sensing,that is, of receiving and recognizing the ideas of sensible objects.” Fromthis it follows that there must be a certain “active faculty for producingor forming these ideas, either in me, or in something else.” This activefaculty cannot be in me, Descartes argues. While this step of the argu-ment is obscure, it seems to depend on two doctrines assumed or estab-lished earlier. One such doctrine is the claim that the mind containsonly two faculties, a cognitive faculty for the apprehension of ideas, andvolition. While this doctrine is not argued for explicitly in the Medita-tions it seems to underlie the analysis of error that leads to the epistemicprinciple of clear and distinct perception in Meditation IV (cf. AT VII56f.; Pr I 32; Passions of the Soul, 17). The second assumption necessaryfor this step is the claim that all cognitive faculties, like imagination andsensation, are modes of pure intellection. This is suggested in the argu-ment from imagination, earlier in Meditation VI, where Descartes distinguishes imagination from pure intellection, and argues that imag-ination (presumably unlike intellection) is not essential to mind. In apassage immediately preceding the argument we are now considering,Descartes extends this conclusion to sensation, and clarifies the statusof both. Sensation and imagination are, he claims, distinct from pureintellection in the same way that shapes are distinct from extension, asmodes from that of which they are modes: “thus in the notion that wehave of these faculties . . . they contain some sort of intellection, fromwhich I conceive that they are distinct from me as figure, motion, andother modes or accidents of body are from the bodies which sustainthose modes” (AT IXA 62).48 The implication here is that all cognitivefaculties must be modes of pure intellection.49 From these two doctrines

SEMEL IN VITA 251

47 All quotations from Descartes’ statement of the argument are from AT VII 79–80.48 I quote here from the French edition, which in this instance is much clearer than the

Latin. The significance of the difference between the two texts suggests Descartes’ ownhand in this passage of the French edition.

49 On this, see Wilson, Descartes, chapter 4, sec. 2. In this important discussion, Wilson arguesfor the primacy of the pure intellect in Descartes’ conception of the mind, as against theview that all mental events, sensations, imaginations, and pure intellections are on the

Descartes can establish that the active faculty that causes sensations inme is not, itself, in me. That active faculty “plainly presupposes no intel-lect,” Descartes claims, from which it follows that it cannot, like sensa-tion and imagination, be a mode of the cognitive faculty of intellection.And since the sensations I have “are produced without my cooperationand often against my will [invitae],” they cannot derive from my will.50

Thus they must derive from something outside of me. But, Descartesargues, God would be a deceiver if they derived from Himself or fromanything other than from bodies themselves:

For since He plainly gave me no faculty to know that [i.e., that ideas of sensa-tion come from something other than body] but on the contrary a very greatpropensity [propensitas] to believe that they come from corporeal things, I donot see how God could for any reason fail to be a deceiver if they [i.e., the ideasof sensation] come from anything but corporeal things. (AT VII 79–80)

From which, Descartes concludes, corporeal things, bodies, exist.51

Descartes’ argument is not altogether unproblematic or convincing,as later philosophers have been quick to remark. But rather thandwelling on the infirmities of the argument, I would like to clarify theway it fits into the debate between commonsense sensualism andreason, which we have been tracing. This argument for the existenceof body makes prominent use of two of the commonsense beliefs thatformed the basis of the abortive Meditation III argument. As he did inthe Meditation III argument, Descartes here appeals to the involun-tariness of sensation. But in contrast to the argument in Meditation III,this commonsense fact about sensation is not, by itself, taken as grounds


same footing. I think that Wilson goes too far, though, when she claims that “Descartesregarded his mind as essentially only intellect” (p. 181). While the intellect may be the onlypassive faculty that pertains essentially to mind, Descartes also recognizes an active faculty,volition, which is distinct from the intellect.

50 It is because the imagination is, at least sometimes, under our voluntary control, that theargument must proceed from sensation rather than from imagination.

51 It is interesting to compare this argument with the parallel argument for the existence ofbody in Pr II, 1. That version lacks the twist at the end, that if my inclination to believein bodies were mistaken, the veracious God would have given me a faculty to correct it.In Pr II. 1 the claim is that we “seem to see clearly that the idea [i.e., of a material thing]comes from something outside of us.” The claim is that God would be a deceiver if thisclear idea were mistaken. This seems to be a direct application of the validation of clearand distinct perceptions to our inclination to believe in bodies, as opposed to the Medi-tations version of the argument, where the inclination to believe is an ingredient in a morecomplex reasoning, and where Descartes never makes the claim that we clearly perceivethe external existence of bodies.

for believing in bodies. In the Meditation VI argument the involun-tariness of sensation gives us only a piece of the argument, the claimthat the active faculty causing sensation is external to mind. And eventhis weaker conclusion is endorsed only in the context of certain claimsabout mind and its faculties, claims that undermine the hidden-facultyobjection Descartes raised against the appeal to the involuntariness ofsensation in Meditation III. The final argument also makes use of theteachings of nature, the strong inclination we have to believe that oursensations derive from bodies, which was presented as one of the prin-cipal supports of our belief in the external world in Meditation III. InMeditation III the inclination to believe is taken as, itself, grounds forbelief, grounds that are rejected because of the known untrustworthi-ness of inclinations in other circumstances. But in Meditation VI, theinclination to believe in bodies as the cause of my sensations is usedonly in the context of a careful examination of when such inclinationsare reliable and when they are not. The claim is that if, in any particu-lar case, the teachings of nature were untrustworthy, then the veraciousGod would have given us the means to correct it. Because, in the spe-cific case at hand, he didn’t, and only because he didn’t give us any-thing to correct the belief our inclination leads us to, we can, in thisspecific instance, trust the teachings of nature and believe that our sen-sations proceed from bodies, in spite of the fact that our inclinationsare not always trustworthy. But when another faculty, reason, of course,gives us the means to correct the teachings of nature, then they mustbe rejected. Such is the case with the inclination we have to believe thatobjects resemble the sensations we have of them, an inclination that isexplicitly noted in the abortive Meditation III argument, and is closelyconnected to the wax example of Meditation II. In the end, whileDescartes uses sensation to establish the existence of bodies, he is verycareful to claim that sensation, by itself, does not establish the nature ofbodies. Immediately after concluding that bodies exist, he wrote:

Nevertheless, they [i.e., bodies] are not perhaps entirely as we comprehendthem through sense, since there are many ways in which the comprehensionof the senses is very obscure and confused. But at least everything that I clearlyand distinctly understand is in them, that is, everything, generally speaking,which is included among the objects of pure mathematics. (AT VII 80)

That is, the bodies whose existence the argument from sensation hasproved, are not the objects of sensation, colored, warm or cold, salty orsweet, but the extended things of Cartesian science.

SEMEL IN VITA 253

It should be clear that implicit in the argument for the existence ofbodies, which Descartes finally endorses, is a general principle of evidence that pertains to all the claims of common sense, a principleof evidence that can guide the use of sensation, imagination, and thenatural and habitual inclinations to belief which Descartes calls theteachings of nature. And just as Descartes draws the principle of clearand distinct perception from the example of the cogito argument in thebeginning of Meditation III, he draws his new principle of evidence forthe teachings of nature from the example of the argument for the existence of bodies in Meditation VI (cf. AT VII 35).52 In the paragraphfollowing the proof, Descartes writes:

As for the rest, there are other beliefs which are very doubtful and uncertain,as that the sun is of such a size or shape, etc., or less clearly understood, aslight, sound, pain, and the like. But however dubious and uncertain they are,from the fact that God is not a deceiver, and that consequently He has notpermitted any falsity in my opinions, without my having some faculty tocorrect them, I have a certain hope of learning the truth about these thingsas well. (AT VII 80)

Common sense is not always wrong, Descartes claims. But before we can trust it, we must examine it carefully using reason, the faculty which, by the argument of Meditation IV, is always trustworthy if usedproperly, the only faculty that God (and Descartes) have given us tocorrect common sense. If reason concurs or is silent, then we can trustcommon sense, otherwise not. The convictions of youth, unceremoni-ously shuffled out in Meditation I, now return, properly tamed byreason.53

And in this we have the resolution of the debate between commonsense and reason that we have been tracing throughout the Meditations.


52 I don’t mean to claim that in the Meditations the principle of clear and distinct percep-tion is derived from the example of the cogito alone. At the beginning of Meditation IIIthe cogito suggests the principle to the meditator. But in Meditation IV it is given a carefulderivation from an analysis of the proper use of the faculty of judgment and the veracityof God. (See AT VII 56–60.) In that derivation, the earlier statement of the principle playsno role whatsoever.

53 It is interesting that this principle of evidence for sensation seems missing from the Prin-ciples, a fact closely related to the version of the argument for the existence of body in PrII, 1. See the discussion of this argument in note 51. It is not clear to me whether thisrepresents a change in Descartes’ position, or whether it is a consequence of the fact thatDescartes intends only a simplified presentation of the contents of the Meditations in thePrinciples.

Common sense, sensation, imagination are not eliminated. Theyremain part of Cartesian epistemology, but under the watchful eye anddomination of reason. Thus Descartes writes in Meditation VI:

But I do not see that it [nature] teaches me that I should conclude anythingfrom these sense perceptions concerning things outside of ourselves unlessthe intellect has previously examined them. For it seems to me that it is thebusiness of the mind alone, and not of the being composed of mind andbody [from which derives sensation and imagination] to decide the truth concerning such matters. (AT VII 82–83; cf. AT VII 438–439)

With this, the new epistemology is in place. All that remains is to workout the details of the new world that reason will show us with the assistance of the senses.54

In the Meditations, Descartes is thus interested in more than the refu-tation of skepticism. This is not to deny that the refutation of skepti-cism is important; until the skeptical challenges to knowledge aresettled, we can have no genuine knowledge. But Descartes is interestedin more than the possibility of knowledge. He is interested in the actualpursuit of knowledge, in formulating the true account of the way theworld is. The Meditations is intended both to establish the possibility ofknowledge, against the skeptics, and to set knowledge on its proper epis-temic foundations. By delineating the proper path to knowledge, thepriority of the intellect and its clear and distinct perceptions over thedeliverances of the senses, Descartes is intending to lay the epistemicgroundwork for his revolution in physics, and for the arguments thatestablish the world of mechanism and allow us to set aside the com-monsense and sense-bound world that the Aristotelians have mistaken

SEMEL IN VITA 255

54 The fact that sensory knowledge is admitted, under appropriate circumstances, is crucialto reconciling Descartes’ demand for certainty in science with his frequent claims to beingan experimental scientist. For a discussion of this question, see my essay, “Science andCertainty in Descartes,” in Hooker, Descartes, 114–151. The breakdown in certainty comes,I claim, not with experiment, which can, if used properly, under the control of reason,lead to certain knowledge, but with the use of something like hypothetico-deductivemethod, which can never lead to certainty. (For more recent reflections on these ques-tions, see essays 5 and 6 in this volume.) For another recent attempt to deal with thesequestions, see Desmond M. Clarke, Descartes’ Philosophy of Science (University Park: ThePennsylvania State University Press, 1982). Clarke, arguing from Descartes’ scientific writ-ings rather than from his philosophical writings about science (see p. 2), also emphasizesthe proper use of experience in Descartes’ science, under the control of reason (see chapters 2 and 3). But Clarke argues that Descartes’ actual method in science is largelyhypothetico-deductive (chapters 5 and 6).

for the world we live in. It is this project, the dethroning of the sensesthat, from our earliest years, ruled the mind, and the elevation ofreason, the rightful sovereign of the intellect, which must be under-taken, once in life, lest we remain trapped in the false world we havefrom our earliest years imagined ourselves to inhabit.55


55 I would like to thank Amélie Rorty and E. M. Curley for helpful comments on earlierdrafts.

12

FORMS AND QUALITIES IN THE SIXTH REPLIES

The Sixth Objections, like the Second Objections, were collected by FatherMarin Mersenne, and purport to represent the views of the group ofphilosophers and theologians who belong to the so-called Mersennecircle.1 The very first objection that Mersenne and his friends make tothe Meditations in the Second Objections concerns the real distinctionbetween mind and body; Mersenne and his friends simply do not under-stand how Descartes’ arguments exclude the possibility that thought isnot a kind of motion, and why a body cannot think (AT VII 123).Descartes, of course, attempts to answer this question in the SecondReplies (as well as in the Third and Fifth Replies), but evidently not toMersenne’s satisfaction. For in the Sixth Objections, the very same ques-tion is raised yet again (AT VII 413). Mersenne goes on to suggest thateven the Church Fathers believed that thought “could occur by meansof corporeal motions” (AT VII 413). The Sixth Objections ends with anappendix and a letter “from some philosophers and geometricians toM. Descartes” in which these very same doubts are voiced again:

However much we ponder on the question of whether the idea of our mind(or a human mind), i.e., our knowledge and perception of it, contains any-thing corporeal, we cannot go so far as to assert that what we call thought

257

1 The Sixth Replies were composed some time between 23 June 1641 (at which time Descartescomplains that he has not received all the sheets Mersenne sent him), and 22 July 1641;see Descartes’ letters to Mersenne on these dates, AT III 385 and 415. Descartes seems tohave received the objections from Mersenne in bits and pieces, and arranged them himself,probably adding a few sentences here and there to Mersenne’s texts; see Descartes toMersenne, 22 July 1641, AT III 415.

cannot in any way belong to a body subject to some sort of motion. . . . Wehave read what you have written seven times, and have lifted up our minds,as best we could, to the level of the angels, but we are still not convinced.(AT VII 420–21)2

That in the end is the problem: Mersenne and his friends simply donot find Descartes’ arguments convincing. Though in the end they haveno specific objections to bring, they are simply not convinced.

Descartes takes these last worries seriously. But rather than respond-ing with yet another version of the argument to distinguish mind andbody, in the Sixth Replies Descartes offers something very different: anintellectual autobiography of sorts, an account of how he came to dis-cover that the mind and the body are distinct, that thought and exten-sion are different, and how he overcame his natural propensity toconfuse the two. We shall return to the details of this account later. Butfor the moment I would like to point out an interesting feature of theaccount. After discussing how in his youth he had confused mind andbody, Descartes makes the following remarks:

But later on I made the observations which led me to make a careful distinction between the idea of the mind and the ideas of body and corporeal motion; and I found that all those other ideas of “real qualities” or “substantial forms” which I had previously held were ones which I had put together or constructed from those basic ideas. And thus I very easilyfreed myself from all the doubts that my critics here put forward. (AT VII442–43)

This is very curious. In the course of a discussion of the distinctionbetween mind and body, Descartes introduces a question, and a veryimportant question that the objectors appear never to have mentioned:the problem of substantial forms and real qualities. Why? And whatdoes this problem have to do with the distinction between mind andbody? I shall begin with a discussion of the history of these problemsin Descartes’ thought before the Meditations. Then we shall turn to theSixth Replies, first to the question of mind and body, and then to thequestion of forms and qualities.


2 The appendix appears to summarize objections made by R. P. de la Barde; the letter whichfollows appears to be from a different source. Both documents seem to be distinct fromthe text that Mersenne gathered as the Sixth Objections, and seem to have been added tothat text by Descartes himself. See Descartes to Mersenne, 23 June 1641, AT III 385, andDescartes to the Abbé de Launay (?), 22 July 1641 (?), AT III 420.

Forms and Qualities before the Meditations

The Scholastic account of body in terms of matter and form is, not sur-prisingly, a doctrine of great complexity, with a variety of differentschools holding a variety of different positions on the central issues.But, by and large, Descartes was not interested in the subtleties ofScholastic thought. Writing to Mersenne on 11 November 1640, whilepreparing to receive the objections to his Meditations, Descartes noteswith great confidence:

I do not think that the diversity of the opinions of the Scholastics makes theirphilosophy difficult to refute. It is easy to overturn the foundations on whichthey all agree, and once that has been done, all their disagreements overdetail will seem foolish. (AT III 231–32)

What interested Descartes were the foundations of Scholastic thought,and what lay at the foundations was the doctrine of body in terms of(primary) matter and (substantial) form.3

Matter is, quite generally, the subject of properties of a thing, thatwhich remains constant as a thing changes from one sort of thing toanother. The notion of a substantial form is somewhat more complex.Most simply, the form is that which, added to primary matter, results ina complete substance. But more substantively, substantial form is thatfrom which the characteristic behavior of the various sorts of substancesderives, and thus that in terms of which their behavior is to beexplained. And so, Descartes notes, writing to Regius in January 1642,helping him to formulate his attack on Voëtius, “they [i.e., forms] wereintroduced by philosophers for no other reason but to explain theproper actions of natural things, of which actions this form is to be theprinciple and the source” (AT III 506). And so, for example, heaviness,the tendency some bodies have to fall toward the center of the earth(universe) is taken to be a quality (what Descartes often calls a realquality) they have by virtue of having the substantial form they do. Andso Descartes characterized the scholastic account of heaviness:

Most take it [i.e., heaviness] to be a virtue or an internal quality in every bodythat one calls heavy which makes it tend toward the center of the earth; andthey think that this quality depends on the form of each body, so that the

forms and qualities 259

3 For a fuller account of the issues discussed in this section, see my Descartes’ MetaphysicalPhysics (Chicago: University of Chicago Press, 1992), chapter 4, from which this section islargely drawn.

same matter which is heavy, having the form of water, loses this quality ofheaviness and becomes light when it happens that it takes on the form of air.(AT II 223)

Closely connected with substantial forms in Descartes’ discussionsare real qualities. For the most part, Descartes is concerned with qual-ities like heaviness (gravitas) that follow directly out of the substantialforms of heavy bodies like earth and water. But they do come up in asomewhat different context earlier on in the Fourth and Sixth Replies.Descartes terms “real accidents” those qualities that remain in the hostin transubstantiation, the color, taste, smell, etc. (see AT VII 248–49,434–35). They are qualities that can exist independently of the sub-stance in which they are found at one time, and can attach themselvesto another substance, as when in transubstantiation, the bread blessedbecomes the body of Christ; though the substance changes, it has thesame sensible qualities that the bread before it had.

Descartes’ metaphysics and physics are fundamentally opposed tothe forms and qualities of the Schools. For Descartes, as is well known,all there is in body is extension, and thus, he argues, everything in bodyis to be explained in terms of the modes of extension, the size, shape,and motion of the smaller bodies that compose a larger one. There are,thus, no forms in Descartes’ world.4 And since the only qualities he recognizes are those that are modes of extension, there is no apparentplace in Descartes’ world for qualities that can detach themselves fromone substance and attach themselves to another.5

Representative of Descartes’ attitudes in his early writings is the following passage from the World:

When flame burns wood or some other similar material, we can see with thenaked eye that it sets the minute parts of the wood in motion and separatesthem from one another, thus transforming the finer parts into fire, air, andsmoke, and leaving the coarser parts as ashes. Others may, if they wish,imagine the form of fire, the quality of heat, and the process of burning tobe completely different things in the wood. For my part, I am afraid of mis-takenly supposing there is anything more in the wood than what I see must


4 At least there are no such forms in the world of inanimate bodies. Descartes does explic-itly hold that the human soul is the form of the body; see, e.g., Descartes to Regius, January1642, AT III 503, 505.

5 In his responses to Arnauld’s Fourth Objections, Descartes does suggest that there may bereal accidents that God can separate from a substance, without our being able to under-stand how that could happen; see AT VII 249. But that would seem to contradict what hesays elsewhere, for example in the Sixth Replies, AT VII 434–35.

necessarily be in it, and so I am content to limit my conception to the motionof its parts. (AT XI 7)

Here Descartes makes it rather explicit that forms are to be rejectedsince they are not needed for explanation; everything that goes on inthe burning wood can be explained simply in terms of extended matterin motion, and there is no need to bring in anything else. This is particularly true since, as Descartes notes a bit later in the World, theforms and qualities the Schoolmen use are themselves quite obscure.Descartes writes:

If you find it strange that in explaining these elements I do not use the qual-ities called “heat,” “cold,” “moisture,” and “dryness” – as the philosophers do – I shall say to you that these qualities themselves seem to me to needexplanation. Indeed, unless I am mistaken, not only these four qualities butall the others as well, including even the forms of inanimate bodies, can beexplained without the need to suppose anything in their matter other thanthe motion, size, shape, and arrangement of its parts. (AT XI 25–26)

But even though Descartes boldly attacks his teachers in the World,it is extremely important to point out that this attack remained unpub-lished in Descartes’ lifetime. In general Descartes is very careful to avoidexplicitly contradicting the Scholastic view in public, at least after thecondemnation of Galileo. Typical is Descartes’ treatment of the ques-tion in his Discourse and Essays. The only mention of forms and quali-ties in the entire book occurs in the following passage from the Meteors,which Descartes himself cites a number of times as an example of hiscaution in dealing with the issue. Descartes writes:

You should know . . . that, in order not to disrupt the peace with the philoso-phers, I in no way want to deny what they imagine in bodies over and abovewhat I have said, such as their substantial forms, their real qualities, andsimilar things. But it seems to me that my reasons should be all the moreaccepted if I make them depend on fewer assumptions. (AT VI 239)6

There is an implicit argument here against forms and qualities, but onlyan implicit one, that they are not needed for explaining things in theworld. And, at the same time, Descartes can deny (as he does) that herejects the philosophy of the Schools. It is important to remember just


6 These words were chosen with great care, and Descartes cited them elsewhere; see Descartesto Regius, January 1642, AT III 492 and the Fourth Replies, AT VII 248. Furthermore, it isinteresting to note that in the Discourse and Essays, Descartes does not say that the essenceof body is extension, or that the sun is at the center of the planetary system.

how controversial it was to deny the basic Aristotelian metaphysics ofform and matter while Descartes was composing and publishing his firstworks. As late as 1624, a group of maverick philosophers was officiallycondemned and exiled by the Parlement of Paris for publicly contra-dicting the philosophy of Aristotle, and their denial of the doctrine ofmatter and form was at the heart of the official displeasure.7 In late1641, after the Meditations were out, the Dutch theologian GisbertusVoëtius was to attack Descartes’ philosophy, in good part because of its implications for the Aristotelian metaphysics of matter and form.Descartes knew well that any challenge to the accepted doctrines of theSchools could cause serious difficulties, and he was in general veryeager to avoid such problems.

But in the Sixth Replies one finds, for the first time, an explicit andpublic attack on forms and qualities, and an argument quite differentfrom any found even implicitly in Descartes’ earlier writings.

Mind and Body in the Sixth Replies

As I pointed out earlier, it is worries about Descartes’ doctrine of mindand body in the Sixth Objections that elicit the attack on form and qualityin Descartes’ response. But to understand Descartes’ remarks aboutform and quality there, we must first understand these more direct con-cerns, and Descartes’ more direct response to them. Here it is impor-tant to remember that unlike most of the other objections Descarteshad to answer, the Sixth Objections were the work of many hands, anddo not necessarily represent a single coherent point of view. Behind theobjections expressed by the authors of the Sixth Objections, I think thatone can find at least two different sorts of alternatives to Descartes’dualism.

In the second section of the Sixth Objections, the objectors remark:

When you say you are thinking and that you exist, someone might maintainthat you are mistaken, and are not thinking but are merely in motion, andthat you are nothing else but corporeal motion. For no one has yet been ableto grasp that demonstration of yours by which you think you have proved thatwhat you call thought cannot be a kind of corporeal motion. . . . Can you


7 See Jean-Baptiste Morin, Réfutation des thèses erronées . . . (Paris: 1624), reprinted in part inLe Mercure françois, t. X (1625), pp. 503–12. It is interesting to note that Mersenne was avery visible critic of this group as well; see Mersenne, La vérité des sciences (Paris: 1625), pp.78–84, 96–113.

therefore show us . . . that it is self-contradictory that our thoughts should bereducible to these corporeal motions? (AT VII 413)8

In the third section of the Objections, the objectors note that severalChurch Fathers held that both angels and the rational soul are corpo-real, but yet that they think:

[The Church Fathers] appear to have believed that [thought] could occurby means of corporeal motions, or even that angels were themselves corpo-real motions; at any rate they drew no distinction between thought and suchmotions. (AT VII 413–14)

The objectors go on to note that they, with Descartes, believe thatanimals are purely corporeal, yet believe that it is necessary to appealto the notion of sensation, presumably a variety of thought, to explaintheir behavior, suggesting, again, that thought is a variety of motion (ATVII 414).9

The alternative to Descartes’ theory that the objectors are present-ing in these passages seems to be some variety of materialism,10 the viewthat a thought is in some sense identical to a certain motion in the body,or, more generally, to a certain purely physical state of the body. Thisis clearly how Descartes interprets these passages. After citing his objec-tors’ request that he demonstrate the impossibility “that our thoughtsshould be reducible to these corporeal motions,” Descartes notes thatthis can only be understood as asserting that “our thought and corpo-real motion are one and the same” (AT VII 424–25).11


8 I am following the Latin text in this last sentence. The French translation is somewhatweaker: “Can you therefore show us . . . that it is self-contradictory that our thoughts arespread out [répandues] in these corporeal motions?” A very similar remark can also befound in the Second Objections that Mersenne also collected; see AT VII 122–23.

9 This section of the Sixth Objections was constructed by Descartes from (at least) two dif-ferent objections that Mersenne had sent him; see Descartes to Mersenne, 22 July 1641,AT III 415. It is possible that the point about animal thought was originally a separatequestion, which Descartes is responsible for connecting to the question of the identifica-tion of thought and motion in the Church Fathers. It is also possible that it was Mersennehimself who linked the two, and that the point Descartes added was a third and differentpoint discussed in this section, that “there are plenty of people who will say that manhimself lacks sensation and intellect, and can do everything by means of mechanical struc-tures” (AT VII 414). One wonders who exactly they had in mind.

10 Using the term in this way is, I acknowledge, somewhat anachronistic; cf. Olivier Bloch,Le matérialisme (Paris: Presses Universitaires de France, 1985) for a history of the term.But I don’t think that this seriously distorts the historical issues under discussion.

11 There seem to be two distinct varieties of materialism at issue here. In the first quotation,the identification of thought and motion seems to entail that properly speaking, “you arenot thinking at all” (AT VII 413). Presumably, the objectors mean to suggest that if

But there may be another kind of alternative suggested in the textof the Sixth Objections. In the appendix to the Sixth Objections, the objec-tor asks: “How do I know for certain that this idea [i.e., the idea of thesoul] contains nothing of a corporeal nature?” (AT VII 420). In theletter that follows and ends the objection, the objectors remark that “wecannot go so far as to assert that what we call thought cannot in anyway belong to a body subject to some sort of motion” (AT VII 420).12

Similarly, they ask Descartes: “how can you possibly have known thatGod has not implanted in certain bodies a power or property enablingthem to doubt, think, etc.?” (AT VII 421).

While it is not entirely clear what they are suggesting, it is possiblethat these objectors have in mind something quite different from thesort of radical materialism that we saw earlier. On the standard Scholas-tic view, body, properly speaking, is made up of matter and form; matteris properly just a constituent of body, and form is as well.13 And thus,insofar as the human body has a rational soul, which is its form, it isthe human body that can properly be said to think, reason, will, doubt,etc., and not just the soul. It is possible that these last objections aremeant to present this kind of Scholastic alternative to Descartes’ radicaldualism of extended body and thinking soul.

I think that it is fair to presume that Descartes’ answer at the veryend of the Sixth Replies is intended to answer not only the specificpassage that elicits the response, the appendix, and supplementaryletter to the Sixth Objections, but the general worries behind all of thedifficulties that the sixth objectors had in being convinced by the argu-


thought and motion are identical, then there is no thought, strictly speaking. (This sug-gests a variety of what has been called eliminative materialism in recent analytic philoso-phy.) But the rejection of exactly such a view seems to be the point of a later passage fromthe Sixth Replies, cited above, in which the objectors argue that although animals are purelycorporeal, yet we must appeal to thought to explain their behavior. The point here seemsto be that although thought is a kind of motion, it is still proper to attribute thought sounderstood to animals. (This suggests a variety of what has been called the identity theoryin recent analytic philosophy.) Given that the Sixth Objections is a compilation, we shouldnot be too surprised to discover certain internal contradictions.

12 I follow here the Latin; the French translation makes reference to “a body agitated bysecret motions.”

13 For an interesting polemical use of this view, see J.-B. Morin’s Réfutation des thèses erronées.. . . Morin appeals to this conception of body to argue against a group of anti-Aristotelianswho had denied that there are forms. Morin claims that since body is matter and form,if they deny forms, they must also deny the existence of bodies in the world. This, ofcourse, conflicts with the Bible, since at the Last Supper Christ held up the bread anddeclared: “this is my body,” something false if there were no bodies (Morin, pp. 36–49).

ment for the distinction between mind and body. Whether or notDescartes recognized distinctions among the different kinds of objec-tions raised in the Sixth Objections, Descartes offers one general responseto those who claim to be unconvinced by his arguments.

Descartes admits that when he first came to the conclusion that themind and body are radically distinct, he, too, was unconvinced, likeastronomers, “who have established by argument that the sun is severaltimes larger than the earth, and yet still cannot prevent themselvesjudging that it is smaller, when they actually look at it” (AT VII 440).But after he reflected a bit on the question, he came to see that hisresistance to the arguments came not from their weakness, but fromhis own prejudice. When young, Descartes notes, “the mind employedthe bodily organs less correctly than it now does, and was more firmlyattached to them; hence it had no thoughts apart from them and perceived things only in a confused manner” (AT VII 441). Descartes continues:

Although it [i.e., the soul] was aware of its own nature, . . . it never exercisedits intellect on anything without at the same time picturing something in theimagination. It therefore took thought and extension to be one and the samething, and referred to the body all the notions which it had concerning thingsrelated to the intellect. Now I had never freed myself from these precon-ceived opinions in later life, and hence there was nothing that I knew withsufficient distinctness, and there was nothing I did not suppose to be corpo-real. (AT VII 441)

But, Descartes reports, as a fact about his own particular history, “lateron I made the observations which led me to make a careful distinctionbetween the idea of the mind and the ideas of body and corporealmotion . . . and thus I very easily freed myself from all the doubts thatmy critics here put forward” (AT VII 442–43). Once he realized thathe resisted the arguments for the distinction between mind and bodyonly because of this childhood error, this confusion between the mentaland the material, the doubts he had simply fell away.14 No doubt heexpects his readers to have the same experience that he had; as in theDiscourse on the Method, the first-person narrative, the “histoire ou fable”(AT VII 4) constitutes a kind of argument to persuade his readers, anexample for them to follow to lead them to the kind of enlightenmentthat Descartes, himself, has achieved.


14 The comparison with Freudian therapy here is obvious.

Descartes sees his objectors’ resistance to his dualism as arising outof a characteristic confusion of things corporeal with things spiritual;the little autobiography that he presents in the Sixth Replies is an attemptto get his readers to see the error of their ways, not by direct arguments(which he gives elsewhere), but by a kind of persuasion. This, and thepsychological analysis on which it rests is hardly new in this passage.The very same analysis is found in Descartes’ response to the authorsof the Second Objections, who raise the very same problem (AT VII130–31), and can be found as early as the Discourse on the Method of1637, four years earlier (AT VII 37; there are other texts too). But thereis something new here, something worth our attention: For the firsttime I know of, Descartes links the issue of the distinction between mindand body with that of forms and qualities. And for the first time,Descartes discusses that delicate issue in public.

Forms and Qualities in the Sixth Replies: The Argument

The confusion of mind and body that dates from our earliest youthmakes it difficult for us to appreciate the force of Descartes’ argumentsfor the distinction between mind and body.15 But, Descartes suggests inthe Sixth Replies, the same confusion also leads us to posit the forms andqualities of the schools. Consider the following paraphrase of theargument presented at the end of the Sixth Replies that Descartes prob-ably sent to the Abbé de Launay on the very day that he sent the textto Mersenne:

The earliest judgments which we made in our childhood, and later on theinfluence of traditional philosophy, have accustomed us to attribute to thebody many things which belong only to the soul, and to attribute to the soulmany things which belong only to the body. So people commonly mingle thetwo ideas of body and soul when they construct the ideas of real qualities andsubstantial forms, which I think should be altogether rejected. (Descartes tode Launay (?), 22 July 1641 (?), AT III 420)

In this passage, and in the passage from the Sixth Replies that it sum-marizes, Descartes offers an implicit argument against the forms andqualities of the Schoolmen, something very different from what he (oranyone else) had offered earlier. But the argument is complex, andrequires some careful unpacking.


15 See, again, my Descartes’ Metaphysical Physics, chapter 4, for a fuller account of these issues.

As I understand it, the argument has three distinct stages. (1)Descartes begins by interpreting the notions of form and quality interms of his own ontology of mind and body. In short, in the Sixth RepliesDescartes Cartesianizes the Scholastic ontology. (2) Once understoodin his own terms, Descartes has an explanation for how we come tobelieve in the Scholastic ontology of form and matter, a sort of psy-chological account of how we come to hold the Scholastic view of theworld. According to Descartes, the Scholastic view is a consequence ofthe confusions we have in our youth; just as we project colors and painsout onto the world, we project other qualities and tendencies, and inthat way, come to believe in the forms and qualities of the Schoolmen.And (3), this leads Descartes to a new argument against the School-men, or, at least, a sort of therapy. Once we realize the errors involvedin our belief in form and matter, we will be cured of the temptation tobelieve in them any more. Let me explain in more detail.

First, the understanding of form and quality. In the Sixth RepliesDescartes gives an account of how he used to think of the notion ofheaviness (pesanteur), as an illustration of how we confuse mental andcorporeal things in our youth. This example is not lightly chosen. It isimportant to remember here that heaviness is one of the basic quali-ties that distinguishes the Aristotelian elements from one another; bytheir nature, the elements earth and water are heavy, while the elementsair and fire are light. Descartes writes:

For example, I conceived of gravity as if it were some sort of real quality, whichinhered in solid bodies; and although I called it a “quality,” thereby referringit to the bodies in which it inhered, by adding that it was “real” I was in factthinking that it was a substance. In the same way clothing, regarded in itself,is a substance, even though when referred to the man who wears it, it is aquality. Or again, the mind, even though it is in fact a substance, can nonethe-less be said to be a quality of the body to which it is joined. And although Iimagined gravity to be scattered throughout the whole body that is heavy, Istill did not attribute to it the extension which constitutes the nature of abody. For the true extension of a body is such as to exclude any interpene-tration of the parts, whereas I thought that there was the same amount ofgravity in a ten foot piece of wood as in one foot lump of gold or other metal– indeed I thought that the whole of the gravity could be contracted to amathematical point. Moreover, I saw that the gravity, while remaining co-extensive with the heavy body, could exercise all its force in any one part ofthe body; for if the body were hung from a rope attached to any part of it, itwould still pull the rope down with all its force, just as if all the gravity existed


in the part actually touching the rope instead of being scattered through theremaining parts. This is exactly the way in which I now understand the mindto be coextensive with the body – the whole mind in the whole body and thewhole mind in any one of its parts. But what makes it especially clear that myidea of gravity was taken largely from the idea I had of the mind is the factthat I thought that gravity carried bodies towards the center of the earth asif it had some knowledge of the center within itself. For this surely could nothappen without knowledge, and there can be no knowledge except in amind. (AT VII 441–42)

Heaviness, as conceived by the Scholastics, is thus mentalistic in anumber of ways. It is imagined to be diffused throughout a body, yetcapable of acting on a single point, just like the Cartesian soul, whichis somehow thought to be diffused throughout the human body while,at the same time, it is especially connected to the pineal gland.16 Likethe human soul, it is extended, not as bodies are extended, but by virtueof being able to act on body, what he calls an extension of power (exten-sio potentiae) a few years later in a letter to Henry More, in contrast tothe extension of substance (extensio substantiae).17 And finally, the notionof heaviness is mentalistic insofar as it appears to attribute to the heavybody a kind of volition or intention, the intention to bear the bodytoward a particular place, the center of the earth, something that couldonly happen if the real quality of heaviness had some knowledge of thecenter of the earth. This last observation, an observation that Descarteshimself considers most important, cuts right to the heart of the Scholas-tic doctrine. If substantial forms and the real qualities that are supposedto follow from them are supposed to explain the characteristic behav-ior of bodies of various sorts, then we must be thinking of them as inten-tional entities, agents of a rudimentary sort, things capable of formingintentions and exercising volition, little souls joined to matter. Indeed,Descartes thinks, body as understood by the Scholastic philosophers,form and quality joined to matter, is just the image of the Cartesianhuman being, immaterial soul united to extended body, projected outonto the material world.18


16 See, e.g., Descartes, Passions of the Soul, §§ 30–31.17 See Descartes to More, 15 April 1649, AT V 342. See also Descartes to Elisabeth, 28 June

1643, AT III 694.18 This is a theme that comes up a number of times in the 1640s. When he is pressed to

explain how the human mind and its body are joined to one another and how they caninteract, as many correspondents sought to understand in the years following the expo-sition of that doctrine in the Meditations, Descartes often compares his account with the

Once we understand forms and qualities in these mentalistic terms,we have a clear understanding of how we came to posit them, a psy-chological account of how we came to view the world in that way; thisis what I called the second stage in the argument. As I discussed earlier,Descartes holds that in our youth, we attribute to bodies properties thatbelong to mind, and to minds, properties that belong to bodies. In par-ticular, we attribute redness to apples, pain to fingers. And we attributethe tendency to be cold to water, and the tendency to be dry to earth,the tendency to fall to rocks, and the tendency to rise to fire. It is inattributing these mental qualities to bodies themselves that we createforms and qualities. Once forms and qualities are assimilated to mentalsubstances, it is possible to see their attribution to bodies as the samemistake that we make when we attribute pain to the finger that hurtsor red to the apple.

And with this, we arrive at the third stage of the argument, the rejec-tion of forms and qualities. The implication here is that once we cometo maturity (at least Descartes’ conception of what maturity is), we willrecognize the distinction between the mental and the corporeal, seeour mistake, and reject the forms and qualities that we mistakenlyattribute to bodies, just as we reject the attribution of color and pain tobodies.

This is the argument that Descartes presents at the end of the SixthReplies. But does he really make a compelling case for rejecting formsand qualities?

Descartes’ argument shows that if forms and qualities exist, then theymust be regarded as substances of a sort, mental substances. TheSchoolman might be brought to agree with that. But is that sufficientto show the Scholastic philosopher that he is wrong? The problem withthe argument becomes evident when we reflect on what it entails forhuman beings and their bodies. If extension is the nature of body, thenhuman bodies cannot think, strictly speaking, as Descartes insisted. (See,e.g., AT VII 444.) But he does not conclude from that that people don’t


Scholastic account of form, quality, and matter. See, e.g., Descartes to Hyperaspistes,August 1641, AT III 424; Descartes to Elisabeth, 21 May 1643, AT III 667–68; Descartesto Clerselier, 12 January 1646, AT IXA 213; Descartes to Arnauld, 29 July 1648, AT V222–23. It is not clear that this view of form and matter corresponds to any Scholasticaccount in particular. But this, in a way, is not the right question to ask. Descartes’ char-acterization of the Scholastic view here should be regarded as polemical rather thanexegetical, a kind of rational reconstruction (and probably something of a caricature) thatprecedes a rational rejection of the foundations of Scholastic physics.

think, or that everything in the human body is explicable in purelymechanical terms. Rather, he concludes that human beings have minds,immaterial souls distinct from their bodies, which think and, underappropriate circumstances, guide the behavior of the unthinking body.But, we might ask, why can’t the Scholastic argue in a parallel way tohis position? Descartes’ argument shows that thought is not in bodiesbut in the soul. This shows that a body, strictly speaking, an extendedthing, cannot contain knowledge of the center of the earth, nor can it will itself to move in that direction. But why can’t we infer from thatthat heavy bodies must have tiny souls, souls distinct from their bodies, in order to think about the place they would rather be and willthe bodies to which they are attached in the appropriate direction? Andso, a Scholastic might respond to Descartes’ argument, the claim thatthe essence of body is extension no more establishes the mechanicalexplicability of the behavior of a falling stone than it establishes themechanical explicability of the behavior of the human being whodropped it.

To put it another way, from the point of view of the Scholastic oppo-nent, Descartes can show, perhaps, that if hylemorphism is true, itinvolves attributing tiny immaterial souls to extended bodies. But if theargument is to refute the doctrine of hylemorphism, Descartes mustshow why there are not or cannot be such tiny souls in nature, whyhuman bodies are to be treated so differently from their inanimatecousins, why outside of humans there is no thought, in body or in mind.Descartes might well have an answer to that; in essence, this is theproblem he confronts when he is attempting to argue that there are nosouls in animals, a question into which I don’t want to enter right now.But it is important here to see that to refute the Scholastic account ofbody, it is necessary to draw a real distinction not only between mindand body, but between human bodies, which have minds, and otherbodies in nature, which don’t. And this is something that goes farbeyond the account Descartes gives in the Sixth Replies.

Forms and Qualities in the Sixth Replies: Further Considerations

In the Sixth Replies Descartes presents his first public attack on theScholastic ontology of form and matter, and for the first time, he pre-sents it in terms of his own ontology of mind and body. Why there? Why


does Descartes choose to answer a question he was not even asked, andturn the discussion of mind and body toward the highly controversialquestion of forms and qualities?

Part of the answer may be that the Scholastic account of body is, ina way, implicated in the very questions that he is asked. As I pointedout earlier, one plausible reading of the objection that elicits thisresponse involves an implicit appeal to a Scholastic view of body. If myinterpretation is correct, one objection to Descartes’ account of mindand body is the counterclaim that body is, by definition, matter andform, and that when the body in question is a living human body, andthe form its soul, then the body can, indeed, think. On that reading,the challenge to Descartes to show that a body can’t think is precisely achallenge to show that this Scholastic analysis of body is incorrect. Ifthat is the question, then Descartes’ answer is that the soul is a sub-stance separate from the body, and when we think of the mind or soulas a constituent of body, part of what makes our body the body it is,then we are confusing mental with material. If that is how Descartesunderstood the objection, then introducing the notions of form andquality into his answer would be quite natural, only making explicitwhat was already implicit in the question. Understood in this way, it isnot Descartes who introduces form and quality into the discussion, buthis objectors.

However, it is not clear that Descartes understood these objectionsin this way. And even if he did, he no doubt also had the more mate-rialistic objections of the earlier passages of the Sixth Objections in mind.These other objectors suggest a more materialistic point of view, thatthought is simply a kind of motion. This, too, represents a confusionof the mental and the material insofar as thought, which pertains exclu-sively to the mental, can in no way be identified with motion, whichbelongs exclusively to the material. Understood in this way, the intro-duction of the question of form and quality into Descartes’ discussionmay have a sort of rhetorical or polemical function, perhaps. Read inthis way, Descartes is claiming that the materialist identification ofthought and motion rests on exactly the same confusion that gives riseto a view of the human being generally thought to be diametricallyopposed to it, the view taught in the Schools. One can, perhaps, readDescartes’ appeal to forms and qualities here as making the point that the two accounts of human beings, so apparently different, are atroot instances of the same mistake. (Here we might see Descartes as


addressing the Epicureanism of Pierre Gassendi as well, trying to showthat despite his anti-Aristotelian pretensions, Gassendi’s views are nobetter than those of the Schools.)19

I think that there is some truth in both of these explanations. But Iwould also like to call attention to something else at work behind thescenes in the Sixth Replies, and offer a more historical explanation forDescartes’ concern with forms and qualities at this time.

It is important to remember that at the very time that Descartes is com-posing the Replies to the Objections, he is also beginning work on his Prin-ciples of Philosophy, which was to appear a few years later, in 1644. Theobjections to the Meditations first start coming in December 1640, andcontinue through the early part of 1641. But the project of the Princi-ples first takes shape in the autumn of 1640 as well. In September 1640Descartes wrote to Mersenne asking him to suggest some Scholastic text-books he might peruse, presumably to prepare himself for the remarkson the Meditations he expected to receive from the Schoolmen, remarksthat he intended to answer in preparation for the publication of the workthe following year. By November 11 he had received Mersenne’s sug-gestions, and reports having purchased the Summa philosophica of FatherEustachius a Sancto Paulo, which he judged to be “the best book whichhas ever been done on that subject” (AT III 232). It is in this same letterthat he first announces the publication of his system:

I would willingly answer your question about the flame of a candle and similarthings but I see that I can never really satisfy you on this until you have seen all the principles of my philosophy. So I must tell you that I have resolvedto write them before leaving this country, and to publish them perhaps within a year. My plan is to write a series of theses which will constitute a com-plete textbook of my philosophy. I will not waste any words, but simply putdown all my conclusions with the true premises from which I derive them. Ithink I could do this without many words. In the same volume I plan to haveprinted a textbook of traditional philosophy, perhaps Father Eustachius’s,with notes by me at the end of each proposition. In the notes I will add thedifferent opinions of others, and what one should think of them all, andperhaps at the end I will make a comparison between the two philosophies.(AT III 232–33; see also AT III 259–60 and a probable reference at AT III 270)


19 One might include Hobbes here, too. But in 1641 Hobbes’ materialism would not havebeen nearly so well known to Descartes as Gassendi’s Epicureanism. At that point,Gassendi, friend of Mersenne and a well known figure, would have been the obvious materialist opponent for Descartes to attack.

Given Descartes’ past practices, this passage is quite remarkable. Afteryears of caution, Descartes announces that he will discuss, explicitly andin public, his differences with the philosophy of the Schools.

Descartes soon gave up the plan to write a commentary onEustachius; Eustachius died the very next month, in December 1640,and Descartes thought it inappropriate to attack him after his death.But by the end of December 1640, Descartes was at work on Part I ofhis new book at the very same time he was answering the objectionsmade to the Meditations. And even though he had abandoned the ideaof an explicit response to Eustachius, there is every reason to believethat Descartes continued to think of his Principles as an answer to thephilosophy of the Schools for some time thereafter, at least until theend of 1641.20 It is in this context that I think Descartes first came upwith the idea that the Scholastic theory of form and quality rests on theconfusion of mind and body, and the projection of the Cartesian soulonto nature as a whole. And so when in early 1641 Mersenne and hisfriends queried him about the distinction between mind and body, it is not surprising that his thoughts would turn to substantial forms and real qualities; though forms and qualities may not have been onMersenne’s mind, they were very much on Descartes’ at that moment,and very much linked to the question of mind and body.

In the end, following his own practice in the Discourse and Essays,Descartes seems to have decided not to attack his Scholastic opponentsso directly. Perhaps Descartes’ instinctive caution returned to him, orperhaps his celebrated problems with Voëtius reminded him of thedangers of attacking the “vulgar philosophy,” as he called it from timeto time. Although Part I of the Principles as it comes down to us con-tains much of the psychology of Aristotelianism that we find in the SixthReplies (see particularly Principles I 71ff.) it contains nothing of the dis-cussion of forms and qualities; indeed, nowhere in the book Descartespublished in 1644 is there the direct attack on forms and qualities thathe seems to have envisioned when he first planned the Principles inDecember 1640. (See, though, Principles IV 200ff.) But in the SixthReplies we have, perhaps, the first draft of the doctrine of the Principia,directed explicitly against its intended target, the forms and qualitiesof the Schoolmen.


20 It is not until December of 1641 that there is any indication that Descartes has given upthe idea of explicitly attacking Scholastic philosophy. See Descartes to Mersenne, 22December 1641, AT III 470.

PART IV

LARGER VISIONS

13

DESCARTES, OR THE CULTIVATION OF THE INTELLECT

René Descartes (1596–1650) aimed to sweep away the past, and startphilosophy anew. Much of what made Descartes important for his con-temporaries, and for us as well, concerns the contents of his philoso-phy. Descartes’ philosophy was directed squarely against theAristotelian philosophy taught in the Schools of his day. For the Aris-totelians, all cognition begins in sensation: Everything in the intellectcomes first through the senses. Descartes’ philosophy, on the otherhand, emphasizes the priority of reason over the senses. Furthermore,Descartes substitutes a purely mechanical world of geometric bodiesgoverned by laws of motion for an almost animistic world of Aristoteliansubstances with innate tendencies to different kinds of behavior. Theseoriginal doctrines, together with his work in metaphysics, optics, math-ematics, the theory of the passions, among other areas, made Descartesa central figure in his age.1

But in this essay I would like to concentrate on something different.Descartes opposed himself not only to the content of the philosophy ofthe Schools, but to their very conception of what knowledge is and howit is to be transmitted. Connected with the new Cartesian philosophy isa genuine philosophy of education, a conception of the aims and goalsof education very different from the one that dominated the Schoolwhere Descartes himself had been educated as a youth. My project inthis essay is to tease out some aspects of this philosophy.

277

1 This is not the place to present a full picture of Descartes’ philosophical and scientificaccomplishments. For a recent overview of Descartes’ thought, see John Cottingham, ed.,The Cambridge Companion to Descartes (Cambridge, Cambridge University Press, 1992).

Rejecting Authority

Let us begin with one of Descartes’ most important texts, the Discourseon the Method of Rightly Conducting One’s Reason and Seeking the Truth inthe Sciences, published in 1637 as the introduction to three scientifictexts, the Geometry, the Dioptrics, and the Meteors. The Discourse is pre-sented as the autobiography of the author, outlining the path he tookto the discoveries that he outlines later in the Discourse (Parts IV andV), and selections of which he gives in the three treatises with which itappeared. But though presented as an autobiography, the Discourse is akind of moral tale, “a history or, if you prefer, a fable” as Descartes putsit [AT VI 4 (CSM I 112)].2 Let us leave aside the question of historicalveracity and simply call the protagonist of the Discourse “RD.”

Part I of the Discourse is largely concerned with RD’s adventures inschool; it gives an interesting account of what school might have beenlike for the young Descartes. (Descartes attended the Jesuit college ofLa Flèche.) The account begins:

From my childhood I have been nourished upon letters, and because I waspersuaded that by their means one could acquire a clear and certain knowl-edge of all that is useful in life, I was extremely eager to learn them. [AT VI4 (CSM I 112–13)]

The young RD was thus eager for learning, eager for school. The schoolhe was sent to was “one of the most famous schools in Europe, whereI thought there must be learned men if they existed anywhere on earth”[AT VI 5 (CSM I 113)]. Furthermore, he thought himself among thebest of the students there, and did not doubt that “the age in which welive [is] as flourishing, and as rich in good minds, as any before it”(Ibid.). But yet, all he found was disappointment:

But as soon as I had completed the course of study at the end of which oneis normally admitted to the ranks of the learned, I completely changed myopinion. For I found myself beset by so many doubts and errors that I cameto think I had gained nothing from my attempts to become educated butincreasing recognition of my ignorance. [AT VI 4 (CSM I 113)]

Because of his dissatisfactions with the learning of the Schools, RDdecided to leave it all behind and travel the world:

278 larger visions

2 References to Descartes’ writings will generally be given in the text of the essay, with theoriginal language edition followed by the translation, in parentheses.

That is why, as soon as I was old enough to emerge from the control of myteachers, I entirely abandoned the study of letters. Resolving to seek noknowledge other than that which could be found in myself or else in the greatbook of the world, I spent the rest of my youth traveling. [AT VI 9(CSM I 115)]

Travel for RD ultimately led to contemplation; having put to one sidewhat he learned in school, RD made the following resolution:

But after I had spent some years pursuing these studies in the book of theworld and trying to gain some experience, I resolved one day to undertakestudies within myself too and to use all the powers of my mind in choosingthe paths I should follow. [AT VI 10 (CSM I 116)]

This is the project that Descartes then represents in the Meditations,published four years later in 1641. Descartes begins the first of the Meditations with the following observation:

Some years ago I was struck by the large number of falsehoods that I hadaccepted as true in my childhood, and by the highly doubtful nature of thewhole edifice that I had subsequently based on them. I realized that it wasnecessary, once in the course of my life, to demolish everything completelyand start again right from the foundations if I wanted to establish anythingat all in the sciences that was stable and likely to last. [AT VII 17(CSM II 12)]

As with the Discourse, the Meditations begins in rejection. The Meditator,to give the protagonist of the Meditations a name, begins by rejecting allformer beliefs, doubting everything that can be called into doubt, fromthe most obvious deliverances of the senses to the simplest truths ofarithmetic and geometry. The First Meditation ends with the hypothe-sis of the evil genius, “a malicious demon of the utmost power andcunning [who] has employed all his energies in order to deceive me”[AT VII 22 (CSM II 15)]. By reflecting on this hypothesis, supposing itto be true, I can keep all my former beliefs and prejudices at bay, andmaintain myself in this state of epistemic detachment. Unlike RD of theDiscourse, who simply sets his former beliefs to one side, the Meditatoruses the strongest arguments possible, arguments derived from theskeptical tradition, to cleanse the mind of all former belief.

In this way, the rejection of the past, of tradition, of the authority ofteachers seems central to the Cartesian philosophy. There is rejectionat a number of different levels. First, there is the rejection of the senses.Descartes begins Meditation I with an explicit discussion of the senses,

the cultivation of the intellect 279

how they can and do deceive us, both while awake and while we areasleep and dreaming. Every schoolboy in Descartes’ day was drilled inthe Aristotelian dictum that “everything in the intellect comes firstthrough the senses.” While the senses return in Meditation VI, theynever regain the full authority that they appear to have had beforebeginning the process of meditation; Descartes’ final judgment is thatthey cannot tell us the way things really are, nor are they even com-pletely reliable in the practical situations for which they were given tous. The point of the opening of the Meditations, then, is at least in partto lead the mind away from its dependence on the senses.3

But the rejection is more profound than that. The skeptical argumentsin Meditation I attack not only the senses, but, more generally, every-thing which the Meditator had learned in the past. At the conclusion ofthe series of arguments, the Meditator admits that “there is not one ofmy former beliefs about which a doubt may not properly be raised” [ATVII 21 (CSM I 14–15)]. Here the arguments of Meditation I join thecomments Descartes made on his own education in the Discourse on theMethod. Descartes lived in a learned intellectual culture, one that empha-sized the importance of tradition and authority. St. Thomas’s Summa the-ologica, for example, still authoritative in the Jesuit college of La Flèchewhere Descartes studied from about 1606 to 1615 or so, is full of rea-soned arguments. But it is also grounded in the authority of Aristotle andthe Church Fathers, whose opinions are constantly cited and discussed.Many of the other books to which Descartes would have been exposedat school were commentaries on Aristotle’s texts, which regularly quotedand discussed the philosophers of the past, both ancient and medieval,contrasting their opinions, weighing their authority.4 In the Renaissance,there were various reactions against the intellectual tradition of theSchools, a diverse movement that went under the general name ofHumanism. Descartes would have been exposed to Humanist trends inthe Jesuit academy, along with the more orthodox Scholasticism that wasat the core of the curriculum there. But Humanism, too, was a learnedtradition, grounded in new scholarship concerning the texts of Greek

280 larger visions

3 For an excellent treatment of Meditation I that emphasizes the rejection of the senses, seeHarry Frankfurt, Demons, Dreamers, and Madmen (Indianapolis, IN, Bobbs-Merrill, 1970),esp. chapters 1–9.

4 On the place of Aristotle and Aristotelianism in the School curriculum in this period, seeL. W. B. Brockliss, French Higher Education in the Seventeenth and Eighteenth Centuries: A Cul-tural History (Oxford, Oxford University Press, 1987) and Charles B. Schmitt, Aristotle andthe Renaissance (Cambridge, MA, Harvard University Press, 1983).

and Roman antiquity, seeking to introduce into the canon new texts, lit-erary and philosophical. Like Scholasticism, Humanism was groundedin a respect for the past. To be educated, then, in the early seventeenthcentury, was to know the wisdom of the past, to understand the differ-ent intellectual traditions.5

It is in this context that we must read Part I of the Discourse and theopening of the Meditations. Descartes seems to be rejecting an entireintellectual tradition, Scholasticism and Humanism, the idea that wemust begin with the wisdom of the past, as well as the authority of thosewho teach the tradition. What Descartes seems to be telling his con-temporaries (and us as well) is that the tradition and those who teachit are not relevant to real knowledge. It is significant here that an admir-ing disciple reports that Descartes gave all his books away when he leftLa Flèche.6 While this is probably not true, it says something about theway in which some of Descartes’ contemporaries read him. If there is aphilosophy of education in Descartes this would seem to be it: Trueeducation must be done by the individual alone, outside of history,outside of tradition, outside of school.

But all this is rather negative; it tells us something about what Carte-sian pedagogy is not, but it tells us little if anything about what it is, whatDescartes thinks the schools and their students should be doing. It isthat to which we must turn.

Intuition, Deduction, and Knowledge

Descartes’ philosophy begins with a rejection of the past. But the firstpositive step is the affirmation of the self. In the Discourse, after reject-ing the learning of the Schools, RD takes off, alone, to experience theworld, ultimately to reject that too, and to turn to himself. In the Med-itations, after the skeptical arguments of Meditation I, the Meditatorbegins the reconstruction of the world with the famous cogito argument,“I think therefore I exist,” building the world out of the self. ForDescartes, the rejection of tradition and authority goes hand in handwith the view that knowledge, properly so-called, must be grounded inthe individual and in the individual alone.

This view can be traced back to one of Descartes’ earliest surviving


5 For a recent survey of the Humanist tradition, see Jill Kraye, ed., The Cambridge Companionto Renaissance Humanism (Cambridge, Cambridge University Press, 1996).

6 The report is contained in the notes of Frans van Schooten the elder, given in AT X 646.

writings, the Rules for the Direction of the Mind (Regulae ad directionemingenii). This work is a treatise on the method of finding truth, whichDescartes probably wrote between 1620 and 1628, abandoning itincomplete approximately ten years before he published his Discourse,though it is summarized in Part II of that work. The main focus of the book is the development of a procedure for investigation which,Descartes claimed, will lead us to genuine knowledge. As a preliminaryto this investigation, Descartes begins with an account of the nature ofknowledge, the goal of this inquiry. Rule 3 reads:

Concerning objects proposed for study, we ought to investigate what we canclearly and evidently intuit or deduce with certainty, and not what otherpeople have thought or what we ourselves conjecture. For knowledge can beattained in no other way. [AT X 366 (CSM I 13)]

Intuition is defined as follows:

By “intuition” I do not mean the fluctuating testimony of the senses or thedeceptive judgment of the imagination as it botches things together, but theconception of a clear and attentive mind, which is so easy and distinct thatthere can be no room for doubt about what we are understanding. Alterna-tively, and this comes to the same thing, intuition is the indubitable concep-tion of a clear and attentive mind which proceeds solely from the light ofreason. [AT X 368 (CSM I 14)]

Intuition is, for Descartes, supplemented by deduction. By deductionDescartes means “the inference of something as following necessarilyfrom some other propositions which are known with certainty” [AT XI369 (CSM I 15)]. Strictly speaking, deduction is not entirely separatefrom intuition. As Descartes writes, “the self-evidence and certainty ofintuition is required not only for apprehending single propositions, butalso for any train of reasoning whatever” [AT XI 369 (CSM I 14–15)].In this way, a deduction is just a train of intuitions. Were our memorybetter, we could dispense with deductive reasoning altogether, andknow by intuition alone. In this way it is fair to say that for Descartes,knowledge, strictly speaking, is grounded in intuition, the immediateoperation of this faculty.

Descartes’ view is that we are all blessed with an innate ability to seecertain truths. The opening passage of the Discourse reads:

Good sense is the best distributed thing in the world: for everyone thinkshimself so well endowed with it that even those who are the hardest to pleasein everything else do not usually desire more of it than they possess. In this

282 larger visions

it is unlikely that everyone is mistaken . . . [T]he power of judging well andof distinguishing the true from the false – which is what we properly call“good sense” or “reason” – is naturally equal in all men. [AT VI 1–2(CSM I 111)]

There is a touch of sarcasm in this, to be sure. But, at the same time,it is a good summary of one of Descartes’ basic commitments: we allhave reason, a faculty given to us by God for distinguishing true fromfalse. This is what he means when he talks of intuition in the Rules.

As Descartes conceives them, intuition and deduction are groundedin the experiences individuals have. Knowledge, for Descartes, does notreside in books or in authorities; for an individual to have genuineknowledge, he or she must actually have the experience that counts asan intuitive grasp of the truth of a proposition or the validity of an infer-ence from one proposition to another. In this way, learning cannot bea spectator sport, a passive absorption of what the teacher has to tell.The student who does not have the actual experience itself has noknowledge, properly speaking. Descartes wrote in the Rules:

And even though we have read all the arguments of Plato and Aristotle, weshall never become philosophers if we are unable to make a sound judgmenton matters which come up for discussion; in this case what we would seemto have learnt would not be science but history. [AT X 367 (CSM I 113)]

True knowledge thus can come neither from teacher nor from tradi-tion. This has obvious consequences for Descartes’ conception of education. True education, then, must involve not the transfer of infor-mation, doctrine, or dogma, but simply the cultivation of the intellect.

The Cultivation of the Intellect

The Rules for the Direction of the Mind is, in a very general sense, meantas a pedagogical work intended to teach us a way to use our native intel-ligence (the literal translation of ingenium in the Latin title of the work)as well as we can. As such it includes mental exercises to help preparethe reader to use the method for finding truth that Descartes there out-lines. The idea of the cultivation of the intellect is basic to this regimen.In Rule 9, for example, Descartes suggests the following exercise:

We must concentrate our mind’s eye totally upon the most insignificant andeasiest of matters, and dwell on them long enough to acquire the habit ofintuiting the truth distinctly and clearly. [AT X 400 (CSM I 33)]


He goes on to say later in the body of the Rule:

Everyone ought therefore to acquire the habit of encompassing in histhought at one time facts which are very simple and very few in number – somuch so that he never thinks he knows something unless he intuits it just asdistinctly as any of the things he knows most distinctly of all. Some people ofcourse are born with a much greater aptitude for this sort of insight thanothers; but our minds can become much better equipped for it throughmethod and practice. [AT X 401–2 (CSM I 34)]

This idea, that we need to practice having intuitions and making deduc-tions before beginning the process of following Descartes’ method andseeking knowledge in earnest, appears again in the Discourse, in onlyslightly different form. There the cultivation of the intellect is not apreparation for using the method. Rather, Descartes recommends in thevoice of RD that we accustom our minds to having intuitions andmaking deductions by practicing the method itself in the domain of mathematics, where intuitions and deductions seem easier to come by.He writes:

Reflecting, too, that of all those who have hitherto sought after truth in thesciences, mathematicians alone have been able to find any demonstrations –that is to say, certain and evident reasonings – I had no doubt that I shouldbegin with the very things that they studied. From this, however, the onlyadvantage I hoped to gain was to accustom my mind to nourish itself ontruths and not to be satisfied with bad reasoning. [AT VI 19 (CSM I 120)]

It is, thus, by practicing this method that RD trained his intellect tograsp truth through intuition and deduction. He writes, again in theDiscourse:

But what pleased me most about this method was that by following it I wassure in every case to use my reason, if not perfectly, at least as well as was inmy power. Moreover, as I practised the method I felt my mind graduallybecome accustomed to conceiving its objects more clearly and distinctly. [ATVI 21 (CSM I 121)]7

284 larger visions

7 When Descartes here talks about mathematics as an appropriate subject for cultivating theintellect, he doesn’t mean Euclidean geometry, the kind of mathematics taught in theSchools. RD was no happier with the mathematics taught in School than he was with anyother subjects. See AT VI 17–18 (CSM 119–20). The kind of mathematics Descartes has inmind here is his own analytic geometry. For a discussion of Descartes’ mathematics, seeStephen Gaukroger, “The nature of abstract reasoning: Philosophical aspects of Descartes’work in algebra,” in Cottingham, ed., Cambridge Companion to Descartes.

A view very similar to that of the Discourse and the Rules is also found inone of Descartes’ latest and most self-consciously pedagogical texts. Inthe 1640s, after having published the Discourse and the Meditations,Descartes began to ruminate about how to get his own philosophy intocirculation in the Schools, and how to get it to replace Aristotle as thenew master. It is with this in mind that he undertook to write a book, thePrinciples of Philosophy, published in Latin in 1644 and in French in 1647.8

While it is not exactly like any textbook in philosophy then in use, it is amore systematic presentation of his philosophy than is found elsewherein the corpus. For the French translation, Descartes composed a prefacethat addresses explicitly the question of how one ought to learn philos-ophy. The idea of the method as a kind of mental exercise for trainingthe intellect is very prominent there as well. After providing for ourselvesa code of behavior to govern our actions while we are rebuilding ourbeliefs, Descartes recommends that we study logic:

I do not mean the logic of the Schools, for this is strictly speaking nothingbut a dialectic which teaches ways of expounding to others what one alreadyknows or even of holding forth without judgment about things that one doesnot know. Such logic corrupts good sense rather than increasing it. I meaninstead the kind of logic which teaches us to direct our reason with a view todiscovering the truths of which we are ignorant. Since this depends to a greatextent on practice, it is good for the student to work for a long time at prac-ticing the rules on very easy and simple questions like those of mathematics.Then, when he has acquired some skill in finding the truth one these questions, he should begin to tackle true philosophy in earnest. [AT IXB 13(CSM I 186)]

The true logic, he tells us later in the preface, is just the doctrine ofmethod as taught in Part II of his Discourse, itself a summary of themethod as taught in the Rules [see AT IXB 15 (CSM I 186)]. In thispreface to the Principles, as in the Rules and the Discourse, Descartes suggests that we begin by cultivating reason, practicing finding truth.

In recommending the cultivation of the intellect through practice inintuition and deduction, Descartes set himself squarely against two features of the Scholastic educational regimen that were intended toexercise the intellect: the study of formal logic and the practice of dis-putation.


8 On the conception of the Principles as a textbook for the classroom, see, e.g., Descartes toMersenne, 31 December 1640, AT III 276 (CSMK 167).

A course on logic, based on Aristotle’s Organon, digested into sim-plified form and rules of thumb by many generations of pedagogues,was a central part of the arts curriculum (i.e., the course of studies pre-liminary to advanced work in law, medicine, or theology) in everyschool in Europe in the early seventeenth century.9 As taught in theSchools, Aristotelian logic was very formal and abstract. Learning logicwas a matter of memorizing numerous rules to enable the student torecognize valid and invalid syllogisms.

While on occasion Descartes felt that he had to mute his public rejec-tion of formal logic, just as he had to tone down his opposition to otheraspects of Scholastic doctrine and practice,10 it is clear that Descartesthought little of formal logic as a part of the education of the young.First of all, Descartes argues, the kind of logic taught in the Schools isof extremely limited utility. Unlike his method, which Descartes some-times refers to as logic, the Aristotelian logic of the schools cannot helpus find new truths, but only to arrange truths that we have already dis-covered by some other means. He writes in the Rules:

On the basis of their method, dialecticians are unable to formulate a syllo-gism with a true conclusion unless they are already in possession of the sub-stance of the conclusion, i.e., unless they have previous knowledge of the verytruth deduced in the syllogism. It is obvious therefore that they themselvescan learn nothing new from such forms of reasoning, and hence that ordi-nary dialectic is of no use whatever to those who wish to investigate the truthof things. Its sole advantage is that it sometimes enables us to explain toothers arguments which are already known. [AT X 406 (CSM I 36–37); cf.AT VI 17 (CSM I 119); AT IXB 13 (CSM I 186)]

Furthermore, Descartes notes, the rules taught in logic are confusingand may lead the student astray. And so, he remarks in the Discourse:

And although logic does contain many excellent and true precepts, these aremixed up with so many others which are harmful or superfluous that it is

286 larger visions

9 See Brockliss, French Higher Education, pp. 194–205 for the teaching of logic in France.10 See, e.g., Descartes’ answers to the Jesuit Father Bourdin’s Seventh Objections to the Medi-

tations, AT VII 522, 544 (CSM II 355, 371). Descartes at that moment was particularlykeen to get the Jesuits, his old teachers, on his side, and bent over backwards not to offendthem. This was not an isolated incident. In writing to his then disciple Henricus Regiusin January 1642, Descartes explained his general policy of tempering his views in delicatesituations so as not to cause unnecessary hostility; see AT III 491–92 (CSMK 205). Regiushad recently gotten in some trouble at the Protestant University of Utrecht for present-ing his Cartesian views with too much boldness, and Descartes was trying to tell him howto avoid future troubles of this kind.

almost as difficult to distinguish them as it is to carve a Diana or a Minervafrom an unhewn block of marble. [AT VI 17 (CSM I 119)]

And so, Descartes suggests in his dialogue, The Search after Truth, weshould set formal logic aside, and cultivate the light of reason directly:

When this light operates on its own, it is less liable to go wrong than when itanxiously strives to follow the numerous different rules, the inventions ofhuman ingenuity and idleness, which serve more to corrupt it than renderit more perfect. [AT X 521 (CSM II 415); cf. AT X 439–40 (CSM I 57)]

The practice of disputation was also a central element of Scholastic edu-cation in the early seventeenth century. According to the Jesuit Ratiostudiorum, or “Order of study,” an overarching curriculum that gov-erned Jesuit education at the time that Descartes attended La Flèche,students were to participate regularly in these exercises, in which theywere expected to argue extemporaneously for and against theses thatwere posed.11

Again, as with his criticism of Scholastic logic, Descartes sometimesmutes his criticisms of the practice. In the Rules, for example, he writes:

Yet I do not wish on that account to condemn that method of philosophiz-ing which others have hitherto devised, nor those weapons of the School-men, probable syllogisms, which are just made for controversies. For theseexercise the minds of the young, stimulating them with a certain rivalry; andit is much better that their minds should be informed with opinions of thatsort – even though they are evidently uncertain, being controversial amongthe learned – than that they should be left entirely to their own devices. [AT X 363–64 (CSM I 11)]

But even in his apparent praise, there are criticisms of the practice. Firstof all, insofar as the aim of the disputation is to convince the listenerof the truth of one side of the disagreement, the emphasis is generallynot on certainty, but on the probable syllogisms used in rhetoric, syllo-gisms whose premises are not necessarily certain, but only plausible tothe intended audience. This, Descartes argues in the Discourse, if any-thing only undermines the student’s ability to discern the truth, unlikethe kind of cultivation of the intellect that he proposes in its place. He writes:


11 For the rules concerning disputations in the Ratio studiorum of 1599, which governed LaFlèche while Descartes was studying there, see Edward A. Fitzpatrick, ed., St Ignatius andthe Ratio Studiorum (New York, McGraw-Hill, 1933), pp. 144ff.

Nor have I ever observed that any previously unknown truth has been dis-covered by means of the disputations practiced in the Schools. For so longas each side strives for victory, more effort is put into establishing plausibil-ity than in weighing reasons for and against; and those who have long beengood advocates do not necessarily go on to make better judges. [AT VI 69(CSM I 146)]

Furthermore, the student who considers argument a competitive sportwill actually resist the light of reason if it appears to oppose a positionhe is obligated to defend. Descartes writes in the Second Replies:

This is why I wrote “Meditations” rather than “Disputations.” . . . In so doingI wanted to make it clear that I would have nothing to do with anyone whowas not willing to join me in meditating and giving the subject attentive con-sideration. For the very fact that someone braces himself to attack the truthmakes him less suited to perceive it, since he will be withdrawing his consid-eration from the convincing arguments which support the truth in order tofind counterarguments against it. [AT VII 157 (CSM II 112)]

Finally, in contrast to his regimen, which leads to certainty, and thus toagreement, the practice of disputation leads only to conflict. Writing inthe preface to the French Principles, Descartes notes:

The truths contained in these principles, because they are very clear and verycertain, will eliminate all ground for dispute, and so will dispose people’sminds to gentleness and harmony. This is the opposite result to that pro-duced by the debates in the Schools, which – slowly and without their notic-ing it – make the participants more argumentative and opinionated, andhence are perhaps the major cause of the heresies and disagreements whichnow plague the world. [AT IXB 18 (CSM I 188)]

In an era very much aware of the religious wars that plagued France inthe late sixteenth century and still plagued Europe during Descartes’lifetime, this was a powerful consideration. Descartes’ hope was that ina world in which every student was taught to cultivate reason and seekonly certainty, disagreement would end and harmony would reign.

The Order of Reasons: Starting on a Firm Foundation

Cartesian pedagogy begins with the cultivation of the intellect, exercisesdesigned to practice finding truth, accustoming the mind to settling fornothing less than the certainty of intuition and deduction. But afterthese exercises, one must “begin to tackle true philosophy in earnest,”as Descartes advises the reader in the preface to the French Principles

288 larger visions

[AT IXB 14 (CSM I 186)]. At this point the Cartesian pedagogy followsthe Cartesian philosophy. We begin with the self, with the cogito argu-ment establishing the existence of a thinking thing (Meditation II).From the self flows everything else. From the idea of God found in theself Descartes proves the existence of God external to the mind (Med-itation III). From the existence of God, His benevolence, and certainfeatures of the mind, Descartes is then able to argue that everythingthat he clearly and distinctly perceives is true (Meditation IV). On thishe is able to ground his arguments for the real distinction betweenmind and body, and his proof of the existence of a real world of bodies,conceived without color or taste, heat or cold, a world of geometricalobjects made real (Meditations V and VI). And with this his physics isoff and running, leading him from the general laws of nature, througha cosmology to a biology, and eventually, he hoped, to an account ofthe human being sufficient to ground both medicine and an accountof the passions, which, in turn, is to ground a truly scientific moraltheory. In this way he offers a systematic mechanist alternative to thephilosophy of Aristotle and the Schools.

The full details of this story go far beyond the bounds of this essay,and constitute an account of the Cartesian philosophy itself. But thereis one aspect of this story that is very important to note in this con-nection. Descartes’ philosophy is clearly organized in a hierarchicalmanner. As he writes in the preface to the French Principles,

The whole of philosophy is like a tree. The roots are metaphysics, the trunkis physics, and the branches emerging from the trunk are all the other sciences, which may be reduced to three principal ones, namely medicine,mechanics, and morals. [AT IXB 14 (CSM I 186)]

This hierarchy has an obvious epistemological consequence forDescartes. If we are to have real knowledge, then we must study thesesciences in the proper order, beginning with the metaphysical founda-tions in the self and God, progressing then to body and physics, beforeending with the practical sciences. In an important letter to his friendMarin Mersenne, Descartes wrote the following explanation of his pro-cedure in the Meditations:

It should be noted that throughout the work the order I follow is not theorder of the subject matter, but the order of the reasoning. This means thatI do not attempt to say in a single place everything relevant to a given subject,because it would be impossible for me to provide proper proofs, since mysupporting reasons would have to be drawn in some cases from considerably


more distant sources than in others. Instead, I reason in an orderly way fromwhat is easier to what is harder, making what deductions I can, now on onesubject, now on another. This is the right way, in my opinion, to find andexplain the truth. The order of the subject matter is good only for thosewhose reasoning is disjointed, and who can say as much about one difficultyas about another. [AT III 266–7 (CSMK 163), cf. AT VII 155 (CSM II 110)]

This attitude is central to Descartes’ assessment of the science of Galileo(1564–1642), a rough contemporary who, like him, opposed the philosophy of Aristotle and the Schools. He writes, again in a letter to Mersenne:

Generally speaking, I find he philosophizes much more ably than is usual, inthat, so far as he can, he abandons the errors of the Schools and tries to usemathematical methods in the investigation of physical questions. On thatscore, I am completely at one with him, for I hold that there is no other wayto discover the truth. But he continually disgresses, and he does not take timeto explain matters fully. This, in my view, is a mistake: it shows that he hasnot investigated matters in an orderly way, and has merely sought explana-tions for some particular effects, without going into the primary causes innature; hence his building lacks a foundation. [AT II 380 (CSMK 124)]

Approaching the study of nature in this way, Descartes thinks, takes usto new heights. In the preface to the French Principles, Descartes notesfive levels of wisdom. The first includes self-evident truths “so clear inthemselves that they can be acquired without meditation”; the secondincludes what we learn from the senses; the third involves what we learnby talking with others, and the fourth, what we learn by reading books“written by people who are capable of instructing us well.” To theseDescartes adds a fifth degree, which, he modestly claims, he is the firstto attain:

This consists in the search for the first causes and the true principles whichenable us to deduce the reasons for everything we are capable of knowing.. . . I am not sure, however, that there has been anyone up till now who hassucceeded in this project. [AT IXB 5 (CSM I 181)]

In emphasizing the proper order of instruction, Descartes perhapsthought of himself as departing from his teachers. One suspects that inthe practical world of the classroom, where one is told to do ethics oneyear, physics another, teachers were not nearly so careful about follow-ing the order of reasoning as Descartes would have liked to have been.But in a deeper sense, his concern with order and the interconnected-

290 larger visions

ness of knowledge is connected with deeper strands in Scholasticthought. Scholastic pedagogues worried considerably over the questionof the order of the curriculum, what should be taught before what, andwhy. Furthermore, the more general point, that true knowledge isgrounded in knowledge of first causes, is something that teachers wouldhave acknowledged, though they would have disagreed about the start-ing place of knowledge. In this way, perhaps, the insistence on startingwith the most basic, and proceeding in order down from there wouldnot have been such a radical idea.12

Indeed, on this score, the true radicals may have been those likeGalileo and, later, Newton, who relaxed the Cartesian (and Aristotelian)insistence on starting with first principles and ultimate causes, andworked in the other direction, from phenomena observed, backtowards the first causes. Galileo, for example, started with observationsof balls falling down inclined planes and bobbing at the end of pendula,and arrived at mathematical accounts of the motion of heavy bodies.Newton (1642–1727), working later in the century, famously claimedto “deduce causes from effects,” and in this way claimed to discover thetheory of universal gravitation. Neither worried about the basic princi-ples and ultimate causes with which Descartes insisted on beginning.Ironically enough, the freedom that came from this move may haveallowed natural philosophers to come closer to penetrating the real firstcauses than Descartes himself did.

A Final Question: A Place for Books and Teachers

In the beginning of this essay, I discussed the way in which Descartesbegins in rejection, rejection of authority of all kinds, including thekind of authority represented by books and teachers. But this wouldseem to raise a special problem for Descartes; he would seem to beposing as an authority who is telling us to reject authority, an authorwho writes books telling us not to read books! To write a book whosemessage is not to read books – this book included – would seem to beself-contradictory; to stand up in front of a class as a teacher, teachingthe lesson that one cannot learn from teachers, would seem to be self-defeating. How can Descartes set himself up as a teacher, even if it is asa teacher who is teaching us this radical truth?


12 On the relations between Descartes and the Schoolmen on the question of starting withfirst causes and the order of knowledge and instruction, see Daniel Garber, Descartes’ Metaphysical Physics (Chicago, University of Chicago Press, 1992), pp. 58–62.

Descartes was very much aware of this paradox. His answer lies in thepersonae that he adopts to present his philosophy. In Descartes’ day, itwas common for the teacher to stand in front of the class, his lecturescarefully written out, and dictate them to the students, who would copythem word for word into their copybooks, to be carefully studied. Insuch a classroom, it was clear who was the master, and who was thestudent, who had the knowledge and wisdom, and who was receivingit.13 The Principles is a textbook, written for the classroom in the hopeof being used in teaching children, and it shares the didactic qualitiesof other textbooks of the era. But Descartes’ personae in others of hiswritings, in the Search after Truth, the Discourse, and the Meditations,Descartes’ stand-in Eudoxus, as well as RD and the Meditator, as I havecalled them, are not teachers of this sort.

In the Search after Truth, Descartes begins in his introduction to thedialogue with a discourse about how we should not judge opinions onthe grounds of who it is that holds them. He writes:

I hope, too, that the truths I set forth will not be any less well received fortheir not being derived from Aristotle or Plato, and that they will have cur-rency in the world in the same way as money, whose value is no less when itcomes from the purse of a peasant than when it comes from a bank. More-over I have done my best to make these truths equally useful to everybody. Icould find no style better suited to this end than that of a conversation inwhich several friends, frankly and without ceremony, disclose the best of theirthoughts to each other. [AT X 498 (CSM II 401)]

The dialogue form is an ideal way of presenting philosophical ideas ina non-dogmatic way. Though it is clear from the beginning which posi-tion Descartes himself endorses (unfortunately, in his use of the form,Descartes is not the equal of Plato or Hume, or even Berkeley), it isthrough debate and the interchange of arguments that the reader isled to see the wisdom of the Cartesian point of view, and not throughbeing told what to think.

While the dialogue is a very traditional form of philosophical instruc-tion, in others of his works, Descartes experiments with different

292 larger visions

13 Many such copybooks survive, which provide a window into the early seventeenth-centuryclassroom. A number of such books of notes are listed as “courses” in the bibliography ofmanuscripts in Brockliss, French Higher Education, pp. 486ff. For some excerpts from philosophy courses that particularly concern seventeenth-century Scholastic reactions to Descartes’ philosophy, see Étienne Gilson, Études sur le rôle de la pensée médiévale dans laformation du système cartésien (Paris, Vrin, 1975), pp. 316–33.

literary devices for presenting his thought in non-dogmatic ways. In theDiscourse, Descartes’ protagonist RD emphasizes that he does not haveany special talent or wisdom that sets him above others: “For my part,I have never presumed my mind to be in any way more perfect thanthat of the ordinary man” [AT VI 2 (CSM I 111)]. Rather, he claims, itwas luck that led him to his discoveries, the method that he will outlinein Part II of the Discourse and the scientific discoveries that he willpresent in the three “essays” that the Discourse introduces:

[T]he diversity of our opinions does not arise because some of us are morereasonable than others but solely because we direct our thoughts along dif-ferent paths and do not attend to the same things. For it is not enough tohave a good mind; the main thing is to apply it well. . . . I consider myselfvery fortunate to have happened upon certain paths in my youth which ledme to considerations and maxims from which I formed a method whereby,it seems to me, I can increase my knowledge gradually and raise it little bylittle to the highest point allowed by the mediocrity of my mind and the shortduration of my life. [AT VI 2, 3 (CSM I 111, 112)]

Even this is not presented dogmatically; RD freely admits that he maybe deceived here: “perhaps what I take for gold and diamonds isnothing but a bit of copper and glass” [AT VI 3 (CSM I 112)]. And so,he tells the reader:

My present aim, then, is not to teach the method which everyone must followin order to direct his reason correctly, but only to reveal how I have tried todirect my own. One who presumes to give precepts must think himself moreskillful than those to whom he gives them; and if he makes the slightestmistake, he may be blamed. But I am presenting this work only as a story [his-toire] or, if you prefer, a fable in which, among certain examples worthy ofimitation, you will perhaps also find many others that it would be right notto follow; and so I hope it will be useful for some without being harmful to any, and that everyone will be grateful to me for my frankness. [AT VI 4(CSM I 112)]

In another image that he uses, RD tells the reader that he is present-ing his intellectual history “as if in a painting, so that everyone mayjudge it for himself” (ibid.). This may be something of a pose; I stronglysuspect that the historical Descartes did think himself to be more intel-ligent than the common person, and that he had enormous confidencein his method and in the discoveries that he made with its help. But nomatter. What is important is that he did not represent himself in thatway: his persona RD does not see himself as transmitting truth to the


reader, but as telling a story, providing an example, some aspects ofwhich the reader may find worthy of following. If RD is a teacher, he isteaching by the example of his own life; he is not telling you to rejectteachers, but showing you how he did, and hoping that you will agreethat the results are worthy of imitation.

Descartes’ Meditator, his persona in the Meditations, is somewhat dif-ferent from RD. Though one can read the opening of the Meditationsas a kind of continuation of the Discourse, RD sitting down to actuallypursue the intellectual program that he prepares in Parts I and II ofthe Discourse (and, in a preliminary version, outlines in Part IV), therhetorical strategy is not the same in the two works. Whereas RD is thecompanion at the tavern, telling you his life’s story, the Meditator is akind of guide. “Guide” is, perhaps, not quite the right word here.Descartes writes in the preface to the reader “I would not urge anyoneto read this book except those who are able and willing to meditate seriously with me” [AT VII 9 (CSM II 8)]. The point is that the Medi-tations is not simply a book to be read as an account of what its pro-tagonist (the Meditator) happened to think on a particular occasion,as the Discourse represents itself. Nor is it a book whose conclusions we are supposed to believe simply because we are told that they are true by the author. Rather, we are supposed to enter into the argu-ments, and meditate with the protagonist. When we read the skepticalarguments of Mediation I, we are supposed to feel their force, and wemust reject everything we formerly believed. When in Meditation II the Meditator discovers his own existence as a thinking thing throughthe cogito argument, we are each supposed to discover our own existence.For the Meditations to work then, we must actually identify with the Meditator, and have, for ourselves, the experiences that lead towardintellectual enlightenment.14 Again, Descartes is not telling you what to believe, but, in a way different from that in the Discourse, showing you how you can come to the knowledge that he thinks he hasobtained.

In these ways, then, Descartes can play the teacher without violatinghis own philosophy of education, and lead us to reject authority, turnto ourselves, and discover, for ourselves, the truths that Descartes would have us learn. But a touch of irony still remains. Descartes, the

294 larger visions

14 On the background to Descartes’ use of the meditation as a literary form for his philosophy, see the essays by A. Rorty and G. Hatfield in A. Rorty, ed., Essays on Descartes’Meditations (Berkeley, CA, University of California Press, 1986).

fresh, new voice in the 1630s and 1640s, when he burst upon the scene,the philosopher who sought to liberate philosophy from the past, hasover the years become one of the classics himself, one of the ancientauthors from which we must liberate ourselves, if we are to follow hisown advice.


14

EXPERIMENT, COMMUNITY, AND THECONSTITUTION OF NATURE IN THE

SEVENTEENTH CENTURY

Introduction

In his important and influential book, How Experiments End,1 PeterGalison discusses how it is that scientists decide when a given experi-ment is finished and when the supposed fact that it purports to estab-lish can be accepted as fact and not a mistaken reading of the apparatus,not a result of a malfunctioning piece of equipment, not a misinter-pretation of a given observation, and so on. This epistemological ques-tion – the transition between individual observations, individual runsof a complex experiment, and the experimental fact that they are sup-posed to establish – is a matter of some discussion in the recent litera-ture in the history, philosophy, and sociology of science.2 It is thisquestion that Galison (and others) have called attention to that I wouldlike to explore in this article.

What strikes me as interesting here is that the very question underscrutiny has a history; while, in a sense, the question has been with usas long as people turned to experience to try to figure out how theworld is, people were not always interested in or aware of the question,and when they were, the answers that they suggested were not always

296

1 Chicago: University of Chicago Press, 1987.2 In addition to Galison, see, for example, Andrew Pickering, Constructing Quarks (Chicago:

University of Chicago Press, 1984); Bruno Latour, Science in Action (Cambridge, MA:Harvard University Press, 1987); Bruno Latour and Steve Woolgar, Laboratory Life: the Con-struction of Scientific Facts, 2nd ed. (Princeton: Princeton University Press, 1986); StevenShapin and Simon Schaffer, Leviathan and the Air-Pump: Hobbes, Boyle, and the ExperimentalLife (Princeton: Princeton University Press, 1985), etc.

the ones that we find most comfortable now. That is what will interestme here, the history of the notion of an experimental fact, if you will,or, as Lorraine Daston has dramatically dubbed it, the “prehistory ofobjectivity.” In Robert Boyle and his generation in the Royal Society, asSteven Shapin and Simon Schaffer have emphasized, we have much ofwhat we take for granted in experimental life, experiments performedon complex and temperamental equipment that often goes wrong, thecentrality of the idea of reproducibility, the idea of a community of sci-entists, and so on. But a generation before Boyle, much of this famil-iar landscape was missing. What I would like to do is give a preliminarysketch of the way all that came to be. What I would like to do is sketchhow the experimental life, as we now think of it, began. In particular,I am interested in the way in which the establishment of experimentalfacts became social. Recent writers have emphasized the role played bythe community of investigators in deciding what counts as an experi-mental fact and what does not. This is a very prominent feature of theaccount of experimental facthood in the Royal Society. However, I shallargue, this is a very recent development.

One cannot tell the whole story in these few pages, though, and Iwill have to be selective. I will begin with a brief discussion of experi-mental facthood in late Renaissance thought before turning to Baconand Descartes and showing the extent to which their conception ofexperimental facthood is radically individualistic. I will then discuss theself-consciously social conception of experimental facthood found in the writings of the early Royal Society. After a digression about somerecent issues concerning the rhetoric of scientific experiments in theperiod, I will end with some speculations about why the transitionoccurred when it did. The transformation in the philosophical viewabout the role of community in the establishment of experimental facts,I suggest, is closely connected with the emergence of a community enti-tled to make the judgments necessary to establish such facts.

Before beginning my story, I should comment briefly on the notionsof experiment and observation. It is important to many discussions todistinguish between observation and experiment, between informationwe get about the world from observing it as it follows its own naturalcourse and information we get from torturing nature, as Bacon put it,setting up situations not normally found in nature and observing whathappens. Important as this distinction is in the seventeenth century, it will not be relevant for my story. And so I will speak indifferently ofobservation, experiment, and experience.

experiment, community, and nature 297

Some Common Sense

People have turned to their senses for information about the world onwhich to ground their natural philosophy, their medicine, and so on aslong as there have been such disciplines. And as long as there have beensuch disciplines, halfway reflective people must have worried at least tosome degree about how it is that one can establish empirical facts aboutthe world at the very lowest level – how you can be sure that the indi-vidual and particular observations you make on a given occasion arenot misleading in some way, the product of chance or happenstance,malfunctioning equipment, a distracted observer, a nonrepresentativespecimen, and so on. And to this apparently simple question, we findin much early literature a relatively simple answer: When in doubt abouta given observation or experiment, do it again.3

Peter Dear has recently found a very nice instance of this way ofthinking about experiment in an obscure and generally unremarkableJesuit textbook on optics published in 1613 by one Franciscus Aguilo-nius. Aguilonius writes, “A single sensory act does not greatly aid in theestablishment of sciences and the settlement of common notions, since error can exist which lies hidden for a single act. But if the act isrepeated time and again, it strengthens the judgment of truth until[that judgment] finally passes into common assent; whence afterwardsthey [i.e., the “common notions”] are put together, through reasoning,as with the first principles of a science.”4 As Dear emphasizes, there isnothing particularly original in Aguilonius’ statement here. He pointsout that this statement is, in a way, just a paraphrase of Aristotle himself.Writing in the Posterior Analytics, Aristotle notes, “So from perceptionthere comes memory, as we call it, and from memory (when it occursoften in connection with the same thing), experience; for memories

298 larger visions

3 This, of course, will not work for astronomy, where the events observed are radically unique,the observation of a particular heavenly body in a particular position in the sky at a giventime. Different strategies evolved for dealing with the fallibility of astronomical observa-tions, generally involving numerous observations made over long periods of time. See, e.g.,the discussions of the determination of mean motions of heavenly bodies in N. Swerdlowand O. Neugebauer, Mathematical Astronomy in Copernicus’s De revolutionibus (2 vols.) (NewYork: Springer-Verlag, 1984), passim.

4 Franciscus Anguilonius, Opticorum libri sex (Antwerp, 1613), pp. 215–16, quoted and trans-lated in Peter Dear, “Narratives, Anecdotes, and Experiments: Turning Experience intoScience in the Seventeenth Century,” in Peter Dear, ed., The Literary Structure of ScientificArgument: Historical Studies (Philadelphia: University of Pennsylvania Press, 1991), pp.135–63, quoted on p. 139.

that are many in number form a single experience.”5 Aristotle’smeaning here is by no means clear.6 But it is not too implausible to seeAristotle as standing behind Aguilonius’ statement. The point is thatAristotelian science is grounded not on individual events of sensoryexperience, particular observations made on particular occasions, buton the general course of experience, on common assent. It is not suf-ficient for an Aristotelian science that we have a particular observationthat it snowed on the morning of January 23, 1979, in Chicago, Illi-nois, or that on a September 26, 1664, a particular apple was observedto fall from a tree and hit one Isaac Newton on the head. What is nec-essary for Aristotelian science is that it be generally accepted that itsnows in northern climes in the winter months or that heavy bodies fall;this is what constitutes experience, properly speaking, as opposed tomere perception. And to go from perception, the individual deliver-ance of the senses on a particular occasion, to what Aristotle and Aguilo-nius call experience, what we might call an experiential fact, requiresthe repetition of these individual perceptions. Should these individualperceptions speak with sufficient unanimity, then memory will trans-form them into experiential facts, facts that can be acknowledged bycommon consensus and used as the foundation of a genuine body ofknowledge. In this way an experimental fact can be regarded as a kindof low-level general statement established by repetition.

Now, these perceptions can be repeated by many different observers,of course. But (and this is something I want to emphasize) it is suffi-cient for them to be repeated by one observer alone; one observer,repeating the observation a sufficiently large number of times, iscapable of constituting an experiential fact, on this conception. Thisconception of facthood is reflected in quite a number of figures in earlymodern science and represents what might well be considered the commonsense view on the question at hand, the question as to howexperimental facts are to be constituted. Consider, for example, WilliamGilbert, one of the most obviously experimental of the very earlymoderns. Gilbert writes in the preface to his De Magnete of 1600, “Letwhosoever would make the same experiments, handle the bodies carefully, skillfully, and deftly, not heedlessly and bunglingly; when an


5 Aristotle, Posterior Analytics II.19, 100a 5–7, trans. in Jonathan Barnes, Aristotle’s PosteriorAnalytics (Oxford: Oxford University Press, 1975), p. 81. See also Metaphysics I.1, 980b28–30.

6 See Barnes, Posterior Analytics, p. 253 for some indications of the complexities.

experiment fails, let him not in his ignorance condemn our discover-ies, for there is naught in these Books that has not been investigatedand again and again done and repeated under our eyes.”7 Gilbert isvery aware that the complexity of the experiments he has performedand the temperamental nature of the equipment he used may make itdifficult for others to get the same outcomes that he did on his trials.Indeed, he begins the book proper by reporting on the mistaken resultsthat others have gotten from antiquity to the present, mistakes that arecorrected by his own, more careful experiments. Gilbert is completelyconvinced that his own results are correct, that he has captured genuineexperimental facts by virtue of the fact that he repeated his trials overand over again: “There is naught in these Books that has not been inves-tigated and again and again done and repeated under our eyes,” Gilbertwrites, and these repetitions give him the authority to present his obser-vations as fact.

Gilbert is hardly unusual here. Dear reports finding the same threadgoing throughout a number of other writers of the period, includingGalileo and Marin Mersenne:8 “I did the trial a hundred times, and itcame out the same on every occasion” is a phrase that for these natural philosophers (and for many others, too, I strongly suspect) con-stitutes the ultimate justification for their confidence in a given exper-imental fact.

This may look a great deal like modern notions of the repeatabilityof an experiment as a criterion for accepting the experimental fact that it purports to establish. It is. But it is important to emphasize here that what is at issue is not repeatability in general but repeatabil-ity by the individual experimenter; to constitute a genuine fact, it mustbe possible for an experiment or observation to be reproducible, butto establish reproducibility, it is sufficient for the individual investiga-tor to be able to reproduce the result a sufficiently large number oftimes. And so the individual investigator speaks with complete author-ity. If you the reader are not convinced, you can, of course, try theexperiment yourself. But the benefits of this repetition accrue to youand you alone; as far as the investigator is concerned, the numerousrepetitions that he did suffice to establish the result of his experimentas fact.

300 larger visions

7 William Gilbert, De Magnete, trans. P. Fleury Mottelay (New York: Dover Books, 1958), p. xlix.

8 See Dear, “Narratives.”

So much for common sense. Although natural philosophy and med-icine had depended on observation of nature and experiment for manyyears before the new philosophers of the seventeenth century, with the new science, the increasing dependence on experience, and theincreasingly sophisticated forms that the appeal to experience took,there came a new attention to the notion of experiment and experi-ence. I would like to turn now to a number of such accounts. I shallbegin with some reflections on the premier theorist of experimentalscience in the seventeenth century, Francis Bacon, and then turn toRené Descartes before confronting the thought of the early membersof the Royal Society.

Bacon and Experimental Facts

No seventeenth-century figure is more closely identified with the newexperimental spirit in science than Bacon. His program for science, hisInstauratio magna, a plan for the revival and restoration of the sciences,has at its center the Novum organum of 1620, a new logical instrumentthat is supposed to tell us how to build a new science more adequatethan the Aristotelian science that still very much dominated the intellectual world in which Bacon grew up. And at the center of the new method outlined in the Novum organum are observation and exper-iment, the collection of facts and their arrangement into natural histories.

In rough terms, the procedure goes like this. The first step is simplythe collection of experiments and observations. Bacon writes, “For firstof all we must prepare a natural and experimental history, sufficientand good; and this is the foundation of all, for we are not to imagineor suppose, but to discover, what nature does or may be made to do.”9

But a natural history, a random collection of facts, is too unwieldy towork with directly. And so, Bacon suggests, “We must therefore formtables and arrangements of instances, in such a method and order thatthe understanding may be able to deal with them.”10 Take, for example,the investigation of the nature of heat, the example that Bacon devel-


9 Novum organum II.10. In the Novum organum itself, Bacon gives little guidance as to howwe might plan a series of experiments. On this see the discussion in the De augmentis(1623) V.2, in Bacon, The Works of Francis Bacon, ed., J. Spedding, R. L. Ellis, and D. D.Heath (Cambridge, MA: Riverside Press, 1863), vol. IX, pp. 71ff.

10 Novum organum II.10.

ops in most detail in the Novum organum. Bacon begins with what hecalls the table of “Instances Agreeing in the Nature of Heat,” or “Tableof Essence and Presence,” in which are listed a variety of circumstancesin which heat may be found, including fiery meteors, quicklime sprin-kled with water, iron dissolved in acid, and fresh horse dung.11 Thesecond table is what Bacon calls “Instances in Proximity Where theNature of Heat Is Absent.” In this table, Bacon examines one by onethe entries in the table of essence and presence and tries to find similarcircumstances in which heat is absent. So, for example, connected withthe observation that iron in acid produces heat, Bacon notes that softermetals such as gold and lead do not give off heat when dissolved inacid. The third table is what Bacon calls the “Table of Degrees.” Here,Bacon makes observations about things that contain the nature of heat,for example, in greater or lesser degree. And so he observes that whileold dung is colder than fresh dung, it has what Bacon calls a potentialfor heat insofar as it will produce heat when enclosed or buried, heclaims. Similarly, Bacon observes that different substances burn withdifferent degrees of heat.12

Once we have compiled the natural history and arranged it into theproper tables, we are ready for the inductive step, at least the first induc-tive step, what Bacon calls the first vintage. At this point, Bacon says,“The problem is, upon a review of the instances, all and each, to findsuch a nature as is always present or absent with the given nature, andalways increases and decreases with it.”13 That is, in the case of heat, wewant to find that which is always present when heat is present and alwaysabsent when heat is absent. This proceeds in two stages. First, Baconuses his tables to exclude possible natures. And so, for example,although Bacon thinks that heavenly bodies are hot, being a heavenlybody cannot be part of the nature of heat, since there are terrestrialbodies that are hot as well.14 Once we have excluded candidates for thenature of heat in this way, we can then examine what is left and say whatit is that all hot things have in common. What Bacon suggests in thecase at hand is that heat is a particular kind of motion: “Heat is amotion, expansive, restrained, and acting in its strife upon the smallerparticles of bodies, . . . not sluggish, but hurried and with violence.”15

This, Bacon claims, is what all instances of heat found in the tables ofour natural history have in common. After this, the Novum organum is

302 larger visions

11 Novum organum II.11. 12 Novum organum II.13. 13 Novum organum II.15.14 Novum organum II.18. 15 Novum organum II.20.

not altogether clear where we go. Presumably, the first vintage is followed by successive vintages in which we press more knowledge ofnature from our initial observations. Furthermore, Bacon suggests, theknowledge we have derived from experiment will in some way suggestto us new experiments to perform, although he does not indicate howexactly this might work.16

The method of the Novum organum is exemplified in the organiza-tion of the House of Salomon, the perfect scientific society that Baconenvisions in his science fiction story, New Atlantis, published posthu-mously in 1627. At the bottom of the organization are those who formthe tables of natural history, a total of twenty-four investigators. Twelve“Merchants of Light” “sail into foreign countries under the names ofother nations . . . [and] bring us the books and abstracts, and patternsof experiments of all other parts.”17 Three “Depredators” collect experiments from books three “Mystery-men” collect experiments frommechanical arts and liberal sciences, and three “Pioneers or Miners” trynew experiments of their own devising. They are joined by three “Com-pilers,” who arrange these observations and experiments into propertables.18 Twelve workers are employed at the next stage of the enter-prise. Three “dowry-men or Benefactors” examine the initial tablescompiled by the Compilers and draw out both technological applica-tions and the first theoretical conclusions that can be drawn from thetables, presumably what Bacon calls the first vintage in the Novumorganum. Three “Lamps,” as he calls them, then draw new experimentsout of the work of the Compilers and Benefactors, which experimentsare then performed by three “Inoculators.” And finally, “we have threethat raise the former discoveries by experiments into greater observa-tions, axioms, and aphorisms. These we call Interpreters of Nature.”19

In this way, Bacon’s method for investigating nature is quite readilyadapted to science as a social and cooperative enterprise; it is nowonder that organizations such as the Royal Society looked back toBacon for inspiration.


16 See Novum organum II.10, where Bacon suggests that the interpretation of nature involvesboth deriving axioms from experience and deducing and deriving “new experiments fromaxioms.”

17 Bacon, The Advancement of Learning and New Atlantis (Oxford: Oxford University Press,1906), p. 273.

18 Ibid.19 Ibid., p. 274. These thirty-six investigators are, of course, assisted by helpers and servants

of various kinds.

A great deal of attention has been given to the inductive stage inBacon’s method, what it is and why it does not really work. But I wouldlike to focus instead on the first and apparently less problematic stage,the collection and construction of natural histories, in particular, onthe way Bacon thinks that the empirical facts contained in a naturalhistory are to be established and checked.

In Advancement of Learning (1605), and later in the expanded andLatinized version of that work, De augmentis (1623), Bacon offers a cat-egorization of all human learning based on his conception of the mind:“The best division of human learning is that derived from the three fac-ulties of the rational soul, which is the seat of learning. History has ref-erence to the Memory, poesy to the Imagination, and philosophy toReason.”20 Of most interest to us here is Bacon’s conception of the category of history. Bacon recognizes a number of different kinds ofhistory; in addition to natural history, Bacon recognizes civil, ecclesias-tical, and literary history. Unlike philosophy proper, which deals withabstractions and generalities, history deals with particulars, on Bacon’sconception, particular events in nature that happened at particulartimes. But, Bacon suggests in De augmentis, matters are somewhat com-plicated here. He writes, “History is properly concerned with individu-als, which are circumscribed by place and time. For though NaturalHistory may seem to deal with species, yet this is only because of thegeneral resemblance which in most cases natural objects of the samespecies bear to one another; so that when you know one, you know all.And if individuals are found, which are either unique in their species,like the sun and moon; or notable deviations from their species, likemonsters; the description of these has as fit a place in Natural Historyas that of remarkable men has in Civil History.”21 History is the domainof atomic facts, as it were. But Bacon recognizes that some of these factsare more general than others. When we are dealing with knowledgeabout specific individuals, the sun, the moon, Julius Caesar, and so on,then history deals with statements keyed to particular places and times:The sun or moon was observed to be at such and such a position in thesky from such and such a place at such and such a time; Julius Caesarwas observed to have uttered such and such words at a particular placeat a particular time. But when dealing with natural historical matters,

304 larger visions

20 De augmentis II.1 (Bacon, The Works, vol. VIII, p. 407); cf. Advancement of Learning II.I.1(Bacon, Advancement, pp. 75–76). See also the account in the Descriptio globi intellectualis(Bacon, The Works, vol. X, p. 404).

21 Ibid. See also The Works, vol. X, p. 407.

a certain kind of generality can creep in. One can drop a certain pieceof gold in a particular vat of aqua regia at a given time and note thatit dissolves. This, of course, might happen because of the particulari-ties of the situation, the particular characteristics of the samples of gold or aqua regia, or, indeed, the observer may be mistaken in think-ing that the gold dissolved on that occasion. But, given the general similarity of samples of gold and aqua regia and the general reliabilityof observers, at least with respect to events such as this, “it would be asuperfluous and endless labor to speak of [each individual case] severally,” as Bacon put it elsewhere.22 And so, in compiling natural histories it is permitted to speak generally and include as a fact that gold dissolves when put in aqua regia. In general, this is exactly the sort of entry one finds in Bacon’s own natural histories.23 Althoughthe facts are based on observation and experiment, Bacon includes not the reports of the particular observations and experiments he (orothers) might have made at some particular place and time but thereport of the general fact that came out of the particular events ofobservation or experiment. In this I suspect that Bacon exemplifieswhat I called the commonsense conception of how experimental facts are to be established. For Bacon, as for the commonsense view,experimental facts seem to be just the unproblematic generalization of repeated experience, similar instances repeated, that constitute ageneral experience.

But there is a further complexity in Bacon’s account worth noting.Bacon’s natural histories are compiled from a number of sources, fromhis own observations and experiments, from those others have madeand either published or related to him, from accounts travelers havebrought back, from books, encyclopedias, ancient accounts, and evenfrom common sayings and proverbs.24 Bacon suggests that we should


22 Descriptio globi intellectualis II (Bacon, The Works, vol. X, p. 406).23 In the Historia ventorum, for example, an account of the winds, Bacon writes that “the west

wind is the attendant of the afternoon, for it blows more frequently than the east windwhen the sun is declining. The south wind is the attendant of the night, for it rises oftenerin the night, and blows stronger” (Bacon, The Works, vol. IX, p. 399). It is clear here thatthis represents not one particular observation but the general run of our experience.

24 For example, on one page of the Historia ventorum, Bacon takes some observations aboutthe winds from Herodotus and Pliny and another from what is simply identified as “thenarrative of a Spanish pilot” (Bacon, The Works, vol. IX, p. 395). Other sources cited passimin this work include Acosta, Columbus, Aristotle, Knolles’s History of the Turks, Gilbert, andVirgil. At one point he writes, “In Britain, the east wind is considered injurious, insomuchthat there is a proverb, When the wind is in the east, ‘Tis neither good for man nor beast”(Bacon, The Works, vol. IX, p. 402).

be quite liberal in what we are to include in our natural histories; theonly things that he categorically excludes are “superstitious stories . . .and experiments of ceremonial magic,” which he dismisses as “oldwives’ tales.”25 But, among the sorts of things that Bacon does allow inhis natural history, he recognizes that there will be differences in thedegree of certainty. In the Parasceve, the portion of the Instauratio magnain which Bacon discusses the preparation of the natural history, henotes that possible entries in the natural history will be of three sorts:certainly true, certainly false, and doubtful.26 As for the first two cate-gories, there is no particular problem; facts that are certainly truebelong in, and commonly accepted “facts” that are generally acceptedbut false should be exposed and rejected as such. But Bacon’s treat-ment of the third category is the most interesting. Bacon suggests thatwe add them to the natural history, but with appropriate indication oftheir status. He continues, “Nor is it of much consequence to the busi-ness in hand because . . . mistakes in experimenting, unless theyabound everywhere, will be presently detected and corrected by thetruth of axioms.”27 Bacon’s idea seems to be something like this.Inevitably, we will find in the reports of others, or even in our ownexperimental work, that false statements are accepted as experimentalfacts. When too many of the entries in our natural history have thatcharacter, we are obviously in trouble. But if our natural history is gen-erally reliable, then we have a way of weeding out these nonfacts. ForBacon suggests, we can use the general statements derived by inductionfrom our natural history to correct that natural history. That is, oncewe have derived general statements from our experience using thecareful method that Bacon outlines in the Novum organum, we are enti-tled to reject anything that does not conform to those general state-ments, “axioms” as he calls them. In this way, Bacon writes in theInstauratio magna, “the senses deceive, but then at the same time theysupply the means of discovering their own errors” (Plan of the “Instau-ratio magna” ).28

In this way, Bacon seems to go beyond the commonsense conceptionof facthood that I outlined earlier in this article. For common sense, a

306 larger visions

25 Parasceve III (Bacon, The Works, vol. VIII, p. 360). See also the Plan of the Instauratio magna(Bacon, The Works, vol. VIII. p. 49).

26 Parasceve VIII (Bacon, The Works, vol. VIII, pp. 366–68).27 Parasceve VIII (Bacon, The Works, vol. VIII, pp. 366–68); see also Novum organum I.118.28 Bacon, The Works, vol. VIII, p. 43.

fact is established as a fact through repetition alone; one begins withthe individual occurrence, the particular observation, the single run ofan experiment and repeats the event until it is certain that there is nomistake of any sort. But to this Bacon adds another criterion, at leastwhen we are dealing with doubtful results. Facts, as embodied in anatural history, determine theory. But, Bacon holds, theory determinesfact as well; for a purported experimental or observational fact to enterthe body of knowledge, it must conform to theory.

As interesting to me as the account Bacon hits upon is the one thathe misses. One presumes that in at least many of the doubtful cases thatBacon has in mind, at least one investigator has done the experimentin question numerous times and has established to his own satisfactionthat he has identified a genuine experimental fact. It would be a naturalsuggestion that the doubtful results could be checked by having otherinvestigators try the experiment as well. But Bacon does not suggestthis. It is quite striking to me that in Bacon’s elaborately organizedHouse of Salomon, among the thirty-six investigators employed full-time in exploring nature, not one is ever asked to redo an experimentoriginally done by another investigator. As we shall see, matters arequite different when the House of Salomon is actually organized a gen-eration later as the Royal Society. But, before turning to the questionof experiment in the Royal Society, I would like to turn to anotherimportant theorist of method in early seventeenth-century natural philosophy, Descartes.

Descartes and Experimental Facts

For us, Bacon and Descartes are opposites, the experimentalist versusthe rationalist. There is, of course, a good deal of truth in this. But, atthe same time, it is important to recognize that the two are not sodistant from one another as we might think. Both are moderns fromthe point of view of the early seventeenth century, opponents of thesterile Aristotelian science of the Schools, and both saw a new methodof investigation as central in the attack against the old and in the estab-lishment of a new science more adequate than the old. Descartes makesa number of complimentary references to Bacon and his program inhis correspondence, and in his Discours de la méthode of 1637 one cansee the echo of the very rhetoric of the Instauratio magna, publishedwhile the young Descartes was working out his own ideas about scientific procedure in the unfinished and unpublished Regulae of the


1620s.29 Descartes is certainly more circumspect about experiment thanBacon is and trusts to reason more than Bacon does. But it is impor-tant to recognize that while experiment may play a somewhat morerestricted role in Descartes’ enterprise than it does in Bacon’s,Descartes considered experiment crucial to the advance of his ownprogram as well. Experiment appears prominently in his celebratedaccount of the rainbow in Discourse 8 of the Météores, a discussion thathe points to as a paradigm of the method of the Discours; thereDescartes appeals to experiments done with prisms and flasks of waterto support his conclusions about the cause of the rainbow.30 In responseto a criticism of his Principia philosophiae of 1644, transmitted byHuygens, that his views are insufficiently confirmed by experience or experiment, Descartes claims that there are “almost as many experiments as there are lines in my writings.”31 And, finally, in Part 6of the Discours, Descartes’ most prominent complaint is that the com-pletion of his work is hindered by the lack of sufficient observationsand experiments.32

But, before turning to Descartes’ conception of experiment and theconstitution of experimental facts, it would be helpful to take a brief andsketchy excursion into Descartes’ conception of method, particularly asit is set out in the early Regulae.33 What Descartes ultimately wants to con-struct is a deductive science. At the bottom is what Descartes calls intu-ition, the ability we have to immediately grasp certain truths and to graspthem with complete certainty. Descartes thinks that we can also see intu-

308 larger visions

29 References to Bacon can be found in AT I 109, 195–96, 251. On the relation betweenBacon’s and Descartes’ writings, see A. Lalande, “Sur quelques textes de Bacon etDescartes,” Revue de métaphysique et de morale 19 (1911), 296–311. Unless otherwise noted,the translations are from Descartes, The Philosophical Writings of Descartes, ed. and trans.,John Cottingham, Robert Stoothoff, Dugald Murdoch, and Anthony Kenny (3 vols.)(Cambridge: Cambridge University Press, 1984–91), where they are keyed to the pagenumbers of AT.

30 The account of the rainbow can be found in AT VI 325–44; it is identified as a productof the method on p. 325, line 7, the only reference to the method of the Discours in anyof the three Essais that accompany it. Furthermore, it is identified as “a brief sample ofthe method,” the only example so identified, in a letter, Descartes to Vatier, 22 February1638, AT I 559.

31 Descartes to Huygens(?), June 1645(?), AT IV 224.32 See, for example, AT VI 63, 65, 73.33 For a fuller account of the method of the Regulae, see Garber, Descartes’ Metaphysical Physics

(Chicago: University of Chicago Press, 1992), chapter 2, or Garber, “Descartes and Experiment in the Discourse and Essays,” essay 5 in this volume. The discussion hereborrows liberally from the discussions in those two places.

itive connections between some propositions known and others; this iswhat he calls deduction. All knowledge properly speaking, scientia, mustcome from intuition and deduction; completed science will have thestructure of conclusions deduced from initially intuited premises. Hismethod is a procedure for constructing such a science.34

The precise method Descartes has in mind is nicely illustrated by anexample he gives of methodical investigation in Rule 8 of the Regulae.As illustrated in that example, Descartes’ method has two parts: a reduc-tive step, leading us from a question posed to an intuition, and a con-structive step, in which a deduction of the answer to the question ispresented. The problem Descartes poses for himself in Rule 8 is thatof finding the anaclastic line, that is, the shape of a surface “in whichparallel rays are refracted in such a way that they all intersect in a singlepoint after refraction.”35 Now, Descartes notices – and this seems to bethe first step in the reduction – that “the determination of this [ana-clastic] line depends on the relation between the angle of incidenceand the angle of refraction.”36 But, Descartes notes, this question is still“composite and relative,” that is, not sufficiently simple, and we mustproceed further in the reduction. Rejecting an empirical investigationof the relation in question, Descartes suggests that we must next askhow the relation between the angles of incidence and refraction iscaused by the difference between two media – for example, air and glass– which in turn raises the question as to “how the ray penetrates thewhole transparent thing, and the knowledge of this penetration pre-supposes that the nature of the illumination is also known.”37 But,Descartes claims, in order to understand what illumination is we mustknow what a natural power (potentia naturalis) is. This is where what wemight call the reductive step of Descartes’ method ends. At this point,Descartes seems to think that we can “clearly see through an intuitionof the mind” what a natural power is.38 Other passages suggest that thisintuition is intimately connected with motion.39 Once we have such anintuition, we can begin the constructive step and follow, in order,through the questions raised until we have answered the original ques-tion, that of the shape of the anaclastic line. This would involve under-standing the nature of illumination from the nature of a natural power,the ways rays penetrate transparent bodies from the nature of illumination, and the relation between angle of incidence and angle


34 See Rules 1–3, AT X 359–70. 35 AT X 394. 36 AT X 394.37 AT X 394–95. 38 AT X 395. 39 AT X 402.

of refraction from all that precedes. Finally, once we know how angleof incidence and angle of refraction are related, we can solve theproblem of the anaclastic line.

If we take the anaclastic line example as our guide, then methodicalinvestigation begins with a question, a question which, in turn, isreduced to questions whose answers are presupposed for the resolutionof the original question posed (i.e., q1 is reduced to q2 if and only if wemust answer q2 before we can answer q1). And so, in a sense, the reduc-tion leads us to more basic and fundamental questions, from the ana-clastic line to the law of refraction and eventually back to the nature of a natural power and to the motion of bodies. Ultimately, Descartesthinks, when we follow out this series of questions, from the one thatfirst interests us to the “simpler” and more basic questions on which itdepends, we will eventually reach an intuition. When the reductivestage is taken to this point, we can begin the constructive stage, turnthe procedure on its head, and begin answering the questions that wehave successively raised in an order the reverse of the order in whichwe have raised them. What this should involve is starting with the intu-ition that we have attained through the reductive step and deducingdown from there until we have answered the question originally raised.Should everything work out as Descartes hopes it will (which it will not,but that is another story), when we are finished we will have the certainknowledge he wants; an answer arrived at in this way will constitute aconclusion deduced ultimately from an initial intuition.

All of this is impressive, in a way. But where does experiment comein? How could Descartes have thought that experiment fits into his con-ception of scientific practice? There is not the space in this article toenter into this question in the full detail that it deserves.40 But, in brief,I think that the answer goes something like this. Experiment does notenter into the method proper. Rather, Descartes conceives of experi-ment as a kind of auxiliary in the reductive stage of the method, onethat allows us to pass from one question to the next.

Consider the anaclastic line example, for instance. At one point inthe argument, Descartes says that the investigator must notice that therelation between the angles of incidence and refraction itself dependson the changes in these angles due to the differences in the mediathrough which the ray is passing (e.g., from air into glass or water intoair) and that these changes, in turn, depend on the way in which the

310 larger visions

40 For a fuller account, see Garber, “Descartes and Experiment.”

ray penetrates the transparent body.41 While it may not require sophis-ticated optical experiments, it seems that this step requires at very leastsome minimal experience with light rays and lenses or other actualinstances of refraction in order to see that light is typically bent bypassing from one medium into another and to come to the realizationthat in order to discover the law that refraction obeys, we must firstunderstand how light passes through media of different sorts. Experi-ment thus helps to perform the reduction and to determine what ques-tion we should take up next in our investigation.

Descartes uses such appeals to experience more explicitly in his dis-cussion of the rainbow. In that case, he is interested in discovering howit is that colors arise in the rainbow. On the basis of experiments witha spherical flask of water, Descartes claims that the rainbow has two dis-tinct bands of color, a primary and a secondary bow, and that the twobows of the rainbow derive from two combinations of reflection andrefraction in a droplet of water. From this, one might conjecture thatthe color might arise from the reflection, the refraction, or the fact thatthe droplets are spherical. But experiments with a prism show that colorcan arise from refraction alone. Reflection and the spherical surface ofthe droplet are thus judged irrelevant to the phenomenon, and in thenext step of the reduction, Descartes focuses on the question as to howrefraction might produce colors in white light. Once again, experimenthelps us to determine what question we should next ask ourselves.

Descartes’ use of experiment is quite different from Bacon’s.Science, although experimental in a sense, remains deductive forDescartes; Baconian induction has no apparent role to play. But exper-iment seems to play a role in preparing the deduction. Insofar as ithelps perform the reductive part of the method, the sequence of stepsthat leads from a question to an intuition, it helps determine the deduc-tion, the same steps followed in reverse order that lead from intuitionto the answer to the question posed. The deductive chain that the Carte-sian scientist seeks in reason, the chain that goes from more basic toless, is exemplified in the connections one finds in nature itself. Insofaras these latter connections are open to experimental determination, wecan use experiment to sketch out the chain of connections in natureand find out what depends on what, and thus we can use the connec-tions we find in nature as a guide to the connections we seek in reason.It may not be obvious to us at first just how we can go deductively from


41 AT X 394.

the nature of light to the rainbow, but poking about with water droplets,flasks, and prisms may suggest a path our deduction might follow.

Descartes’ science is not grounded in natural history in the directway that Bacon’s is, yet the sorts of tables that Bacon recommends arenot altogether irrelevant to Descartes’ procedure. Writing to MersenneMay 10, 1632, Descartes notes that “it would be very useful if some . . .person were to write the history of celestial phenomena in accordancewith the Baconian method . . . without any arguments or hypotheses.”42

Such tables of phenomena and their correlations with one another,independent of any theory, are precisely what Descartes needs to determine the relations of dependence of one phenomenon onanother necessary to perform the reductive step of the method. Butwhat status do the experimental facts that go into a natural history havefor Descartes?

Descartes, of course, is well known for his distrust of the senses. Anddistrust them he did; he warns us that things are not at all as our sensestell us they are, that they are not red and green, sweet or salty, that ournaive belief that all our knowledge derives ultimately from our sensesis a prejudice of sense- and body-bound youth, a prejudice that mustbe rejected before we will be able to penetrate to the true nature ofthings. But it is important to recognize that he did not reject experi-ence altogether.

The fullest account of the senses is in Meditation 6. Descartes’account there is complicated, but, in brief, the strategy is as follows.Descartes is here dealing with something that God gave us, just as Hegave us clear and distinct perceptions. As such, Descartes argues, theymust be in some sense true: “It is doubtless true that everything thatnature teaches me [and this includes the senses] has some truth in it.”43

When it is truth about the nature of things that we are interested in, itis the light of reason, clear and distinct perceptions, that we must turnto first. And so, while some of the teachings of nature will turn out tobe true, it is only the intellectual examination of them that will estab-lish this. In this way, Descartes rejects the hyperbolic rejection of the senses that begins the Meditations and, indeed, goes on to rejecteven the dream argument that is so prominent in Meditation 1.44 But,although the teachings of nature – what we learn from our senses – arerestored, they are subordinate to reason; they may be trusted to some

312 larger visions

42 AT I 251. 43 AT VII 80. 44 AT VII 89–90.

extent and in some circumstances, but only after they have been givena clean bill of health by reason.

What this means in more specific terms can be illustrated by anexample from the rainbow case discussed earlier. In the rainbow case,Descartes begins by observing that on his flask, the stand-in for the rain-drop, there are two regions of color at roughly 42 and 52 degrees fromthe ray of sunlight; these two regions correspond to the primary andsecondary bows of the rainbow. This observation is the starting placeof his account, and one can presume that he repeated it often enoughto convince himself that it was trustworthy.45 But, in the end, Descartesactually deduces from his law of refraction that parallel rays of lightfrom the sun will converge at almost exactly those two angles after theappropriate number of reflections and refractions. After giving hisaccount, Descartes notes that an earlier observer, Maurolycus, set theangles incorrectly at 45 and 56 degrees on the basis of faulty observa-tions. Descartes remarks that “this shows how little faith one ought tohave in observations which are not accompanied by the true reason.”46

It is only because we can calculate the angles of the primary and sec-ondary bows from the account we have of the rainbow that we can besure of what they are, despite the fact that the investigation began withan experimental determination of those angles. Although it is an obser-vation that starts the ball rolling, it is only through deduction that anexperimental fact observed can actually enter the body of scientificknowledge, in strict terms. Experience is important, but only under thecontrol of reason, as Descartes took great pains to emphasize in Medi-tation 6. In a way, Descartes uses experiment here in the way one mightin geometry. In geometry, one might use carefully drawn diagrams andmeasurements made from them to suggest possible theorems. But still,one would want to hold, any geometrical facts found in this way aregrounded in the geometrical demonstration and not in the diagramthat may have originally suggested the fact to the investigator. In thisway, Descartes notes, “When [Pierre Petit] promises to refute my [lawsof ] refraction through experience, there is no more reason to listenthan if he wanted to show that the three angles of a triangle aren’t equal


45 See the letter to Mersenne, 29 January 1640, AT III 7, where he suggests that in order tohave complete assurance, a given observation with respect to the declinations of a magnetshould be performed “a thousand times” rather than just three, as another investigator,John Pell, had done. We can presume that this is a standard that he would have adoptedfor his own work, in principle if not in practice.

46 AT VI 340.

to two right angles by way of some faulty square rule.”47 Observationand experiment may play an important role in establishing an experi-mental fact, but it is reason that must confer the ultimate status of fact-hood on an observation. While there are important differences indetail, of course, Descartes’ account here is not unlike Bacon’s; for boththere is an important sense in which theory must constitute experi-mental facts.

There is one further feature of Descartes’ attitude toward experi-ment and experimental facts that I would like to call attention to here.One of Descartes’ basic commitments, indeed, one of his obsessions, isthe rejection of authority and the consequent centrality of the individ-ual over community. In the Regulae Descartes emphasizes that only whatan individual intuits and deduces is real knowledge for him; knowledgeby authority is no knowledge at all (Rule 3). The whole message of theDiscours de la méthode is the rejection of authority and the importanceof the individual’s building a world for himself.48 This is the project thatis actually taken up in the Meditations, where the meditator begins byobliterating the world around him and, starting from scratch, builds aworld from the cogito, the thought of a solitary self. This radical indi-vidualism is also reflected in Descartes’ attitude toward experimentalscience. Part 6 of the Discours de la méthode is concerned with the needfor additional experiments in order to complete Descartes’ scientificprogram. Descartes begins by reporting the attitude he took in hisyouth. Originally, he reports, he believed that he should publish thedetails of his foundations for physics and the full system based on thosein order to stimulate the work of others, to get others to build on thefoundations he had laid and make the new observations necessary tofinish the job. And so, Descartes thought, publishing his thoughtswould convince others to “assist me in seeking those [observations]which remain to be made” (i.e., send money). At that time Descartesalso hoped that others would “communicate to me the observations

314 larger visions

47 Descartes to Mersenne, 9 February 1939, AT II 497.48 In Part 1 of the Discours, e.g., Descartes elaborately goes through what he learned at

School, only to argue that there is little of value in it; instead, he concludes, he must leave School and the traditions that it embodies and find out how things are for himself.In Part 2, he employs city planning and architectural metaphors to put forward the viewthat the best cities and houses are those that arise not from the accidents of history butfrom the careful planning of a single individual. It is no accident that the Discours is writtenin the first person, as a single individual giving an account of the world he builds forhimself.

they have already made.”49 But, motivated at least in part by the con-demnation of Galileo in 1633, Descartes reports that he changed hismind about the wisdom of publishing his full system of physics.50 Andwith that change came others. He came to think, first of all, that otherswere not really in much of a position to advance his program; as thebuilding metaphors of Part 2 of the Discours suggest, that work is bestdone that is done by a single individual.51 But Descartes also changedhis mind about the value of experiments done by others. Descartesadmits that “as regards observation . . . one man could not possiblymake them all.”52 But, he asserts, apart from paid assistants, “he couldnot usefully employ other hands than his own.”53 Descartes continues,“And as for observations that others have already made, even if theywere willing to communicate them to him . . . they are for the most partbound up with so many details or superfluous ingredients that it wouldbe very hard for him to make out the truth in them. Besides, he wouldfind almost all of these observations to be so badly explained or indeedso mistaken . . . that it would simply not be worthwhile for him to spendthe time required to pick out those which he might find useful.”54

Descartes concludes, “So, if there were someone in the world whom weknew for sure to be capable of making discoveries of the greatest pos-sible importance and public utility, and whom other men accordinglywere eager to help in every way to achieve his ends, I do not see howthey could do anything for him except to contribute towards theexpenses of the observations that he would need and, further, preventunwelcome visitors from wasting his free time.”55 The message is clear:send your money, not your observations, to R. Descartes, care of thepublisher. (And do not visit, either.) Experimental science is thus, forDescartes, a solitary activity, one that does not require a community,one that would be in fact hindered by having to take place within acommunity.

While Descartes and Bacon may agree to some extent about the con-stitution of experimental facts, the contrast with Bacon here is dramatic.In response to the failure of the philosophy of the Schools – their argu-ments from authority, their book learning, and their disputations –Bacon turns to a new society and new forms of cooperative enterprise.Bacon’s new society, the House of Salomon, is a society that institution-


49 AT VI 65; see also AT VI 63. 50 AT VI 60, 65.51 AT VI 69; on the building metaphor, see AT VI 11ff.52 AT VI 72. 53 Ibid. 54 AT VI 73. 55 Ibid.

alizes his new experimental philosophy and exists outside the Schools.Descartes’ response to the crisis in the Schools is altogether different.Instead of trying to create a new society, Descartes sees inherent prob-lems in any cooperative conception of the creation of new knowledge.Descartes thus chooses to place the scientist outside society.

The Royal Society and the New Experimental Philosophy

When Thomas Sprat stepped forward in 1667 to defend the new RoyalSociety of London, a self-professed society for the promotion and per-fection of the experimental approach to science, it was Bacon to whomhe turned as a distinguished ancestor. He writes,

I shall onely mention one great Man, who had the true Imagination of thewhole extent of this Enterprize, as it is now set on foot; and that is, the LordBacon. In whose Books there are every where scattered the best arguments,that can be produc’d for the defence of Experimental Philosophy; and thebest directions, that are needful to promote it. All which he has alreadyadorn’d with so much Art; that if my desires could have prevail’d with someexcellent Friends of mine, who engag’d me to this Work: there should havebeen no other Preface to the History of the Royal Society, but some of hisWritings.56

But in what sense was Bacon an inspiration? Certainly Sprat and his colleagues were attracted by his emphasis on experiment and naturalhistory as the basis of all natural philosophy and by his emphasis on thecooperative and communal nature of scientific investigation. Thus,Joseph Glanvill writes in his Plus ultra (1668), a sympathetic, althoughnot-quite-authorized, account of the Society:

The deep and judicious Verulam [i.e., Bacon] . . . proposed . . . to reform andinlarge Knowledge by Observation and Experiment, to examine and recordParticulars, and so to rise by degrees of Induction to general Propositions,and from them to take direction for new Inquiries, and more Discoveries,

316 larger visions

56 Thomas Sprat, The History of the Royal-Society of London for the Improving of Natural Knowl-edge (London: 1667; reprinted Gainesville, FL: Scholars’ Facsimiles and Reprints, 1958),pp. 35–36. The “excellent Friends” Sprat mentions in this passage are the other membersof the Royal Society. Sprat’s History was closely supervised by the Society, and it is fair toread it as a representation of their collective views; on the history of the History, seeMargery Purver, The Royal Society: Concept and Creation (London: Routledge and KeganPaul, 1967), pp. 9–19; P. B. Wood, “Methodology and Apologetics: Thomas Sprat’s Historyof the Royal Society,” British Journal for the History of Science 13 (1980), pp. 1–26.

and other Axioms . . . So that Nature being known, it may be master’d,managed, and used in the Services of humane Life. This was a mighty Design,groundedly laid, wisely exprest, and happily recommended by the GloriousAuthor, who began nobly, and directed with an incomparable conduct of Witand Judgment: But to the carrying it on, It was necessary there should bemany Heads and many Hands, and Those formed into an Assembly, thatmight intercommunicate their Tryals and Observations, that might joyntlywork, and joyntly consider. . . . This the Great Man desired, and form’d aSOCIETY of Experimenters in a Romantick Model; but could do no more:His time was not ripe for such Performances.57

The “Romantick Model” is, of course, the House of Salomon in NewAtlantis; although the time may not have been ripe for the realizationof such a design in the 1620s, when Bacon envisioned it, Glanvill andhis friends thought that the 1660s was just the time to realize Bacon’sambitious vision.

But, although Bacon was the inspiration for the Society and waslauded for his great vision, he was not followed in every particular.Sprat, for example, appears to reject systematic rules of experimentalmethod: “The true Experimenting has this one thing inseparable fromit, never to be a fix’d and settled Art, and never to be limited by con-stant Rules.”58 Although it is not entirely clear exactly what Sprat meansto reject, and although Bacon is not mentioned by name here, it is notimplausible to see in this a criticism of Bacon’s fixed (although notrigidly so) methodology for experimental procedure. But most inter-esting for my purposes is another criticism Sprat directs at Bacon:

His Rules were admirable: yet his History not so faithful, as might have beenwish’d in many places, he seems rather to take all that comes, then to choose;and to heap, rather, then to register. But I hope this accusation of mine canbe no great injury to his Memory; seeing, at the same time, that I say he hadnot the strength of a thousand men; I do also allow him to have had as muchas twenty.59

Although Bacon saw the importance of observation and experiment forthe advancement of science, his natural histories are defective, Spratargues. Sprat seems to recognize that Bacon does eventually sortthrough and reject some of the purported experimental facts that find


57 Glanvill, Plus ultra: or, the Progress and Advancement of Knowledge since the Days of Aristotle(London: 1668), pp. 87–88. On the status of Glanvill’s Plus ultra, and its relation to theRoyal Society, see Purver, The Royal Society, pp. 13–14.

58 Sprat, The History, p. 89. 59 Sprat, The History, p. 36.

their way into his natural histories; although Bacon initially takes “allthat comes,” he does eventually choose what to base his induction onand does eventually reject observations that conflict with the generalprinciples arrived at by induction. But, Sprat suggests, one must bemore selective in the first place and weed out bad observations beforethey find their way into one’s natural history. Such careful attention tothe establishment of experimental facts is basic to the mission of thenew Royal Society, Sprat argues; indeed, it is built into the very struc-ture of that community.

The Royal Society was interested in gathering experimental factsfrom all who had them to contribute. But, Sprat writes, “I shall lay itdown, as their Fundamental Law, that whenever they could possibly getto handle the subject, the Experiment was still perform’d by some ofthe Members themselves.”60 It is crucial here that the experiment beperformed not by one of the members, but by some of the members.When the Royal Society took it on itself to sponsor an experiment orseries of experiments, it was a matter of policy, Sprat reports, that a number of different members be involved. Experiments were organized, Sprat writes,

either by allotting the same Work to several men, separated one fromanother; or else by joyning them into Committees. . . . By this union of eyes,and hands there do these advantages arise. Thereby there will be a full com-prehension of the object in all its appearances; and so there will be a mutualcommunication of the light of one Science to another: whereas single labourscan be but as a prospect taken upon one side. And also by this fixing of severalmens thoughts upon one thing, there will be an excellent cure for that defect,which is almost unavoidable in great Inventors. It is the custom of suchearnest, and powerful minds, to do wonderful things in the beginning; butshortly after, to be overborn by the multitude, and weight of their ownthoughts; then to yield, and cool by little and little; and at last grow weary,and even to loath that, upon which they were at first the most eager. . . . Forthis the best provision must be, to join many men together.61

The claim that experiments must be done by a number of differenthands is quite explicit and quite carefully thought out; it is only if a

318 larger visions

60 Sprat, The History, p. 83. When reporting this as “their Fundamental Law,” Sprat is report-ing what they agreed to do; what they actually did is quite another question, of course. Inwhat follows I shall limit myself to a discussion of what the Royal Society thought of themselves as doing, their avowed practice, and shall not be concerned with what theyactually did.

61 Sprat, The History, pp. 84–85; cf. 100. See also Glanvill, Plus ultra, pp. 108–9, 114.

number of different people are involved in carrying out experimentsand replicating the experiments that others submit that we can avoidthe errors that inevitably creep in if only one experimenter is involved,even if he repeats his experiment numerous times.

In addition to the claim that experiments must be repeated by avariety of hands, Sprat further reports that facts must be establishedthrough the consensus of the community as a whole. He writes:

[After the performance of an experiment] comes in the second great Workof the Assembly; which is to judg, and resolve upon the matter of Fact. In thispart of their imployment, they us’d to take an exact view of the repetition ofthe whole course of the Experiment . . . ; never giving it over till the wholeCompany has been fully satisfi’d of the certainty and constancy; or, on theotherside, of the absolute impossibility of the effect. This critical, and reiter-ated scrutiny of those things, which are the plain objects of their eyes; mustneeds put out of all reasonable dispute, the reality of those operations, whichthe Society shall positively determine to have succeeded. . . . There is not anyone thing, which is now approv’d and practis’d in the World, that it is con-firm’d by stronger evidence, than this, which the Society requires; exceptonely the Holy Mysteries of our Religion.62

Experimental facts are now established by the community as a whole.In this way, we have a new sense of reproducibility entering into the

conception of an experimental fact. On the commonsense view, a viewcommon to the generation preceding the Royal Society, what wasimportant was simply the repetition of an observation or experiment;it did not matter who or how many did the actual repetition as long asit was done a number of times sufficient to convince the investigator(s)that the result was a genuine experimental fact and not just a fluke ofcircumstances. But, on the view of the Royal Society, this is not suffi-cient. For Sprat and his “excellent Friends,” to establish an experi-mental result as a genuine fact, to enshrine it in one’s natural historyand use it as the basis of induction, the result must be repeatable (andbe repeated) by a number of different persons, or, at least, it must havebeen repeated in their presence.63 Repeatability in this sense is a con-siderably more stringent requirement for facthood.

In this we also have an important transformation in Bacon’s con-ception of the scientific community. Sprat’s appeal to community is a


62 Sprat, The History, pp. 99–100.63 See Shapin and Schaffer, Leviathan and the Air-Pump, pp. 55ff. on the practice of per-

forming experiments in public for this purpose.

repudiation of the Cartesian ideal of the solitary investigator, to be sure,but it is also a repudiation of Bacon’s conception of experimentalscience. Bacon saw experimental philosophy as a cooperative venture.But for Bacon, the main advantage of the numerous investigatorsworking together is that more facts can be collected for one’s naturalhistory and consequences derived more expeditiously. As I pointed outearlier, not one of the thirty-six investigators in the House of Salomon,Bacon’s “Romantick Model” for the Royal Society, is involved in repro-ducing experiments done originally by others; experiments, done byindividuals, working alone (with their servants and assistants acting onlyas extensions of themselves), enter into the natural histories directly.Nor do they ever gather together to discuss the experiments that somemembers of the House are deputed to perform. But things are very dif-ferent when the House of Salomon is built in London. There it is builtinto the very structure of the Society that experiments are to be per-formed by many hands, witnessed by many eyes, and certified as fact by the Society as a whole. For what I have called the commonsense conception of experiment, an experimental fact is established by anindividual through the senses. For Bacon, and in a different way forDescartes, an experimental fact is also established by an individual,although not directly through the senses; although Bacon recognizesthe importance of community to the advance of knowledge, in theestablishment of particular experimental facts, he seems to be as much an individualist as Descartes is. But, on the new conception ofthe Royal Society, an experimental fact can be established through thesenses but not by an individual. Experiments end and experimentalfacts are constituted not when the individual investigator decides thatit is time but after an experiment is repeated by more than one inves-tigator and when the community as a whole is satisfied that a fact hasbeen established.

My account of the view of experimental facthood in the Royal Societyis, in a way, not particularly novel; in essence, it is at the backbone ofShapin and Schaffer’s important study, Leviathan and the Air-Pump.64 But

320 larger visions

64 See Shapin and Schaffer, Leviathan and the Air-Pump, pp. 55ff. I should emphasize that thequestions that most interest me are different from the questions that Shapin and Schaffer attempt to answer in Leviathan and the Airpump. Their questions concern thehistory of experimental philosophy as such, why experimental philosophy as such arose inEngland when it did, and how and why it came to triumph over a different and nonex-perimental conception of science, such as that represented by Hobbes. To answer this,they appeal to the political context of the debates, and the way Hobbes’s and Boyle’s posi-tions fit into that context. The answer they offer is interesting and worth taking seriously;

what I want to emphasize is that this communitarian view of experi-mental facthood was something quite new, a self-conscious innovationintroduced by the Royal Society in the 1660s. One might possibly beable to find precedents for this, although I doubt it. But what is impor-tant is that it is an idea that is not found in the important theorists ofscientific practice in the generations immediately preceding the foun-dation of the Royal Society and was regarded as an innovation by theRoyal Society itself, a new and improved way of thinking about experi-ment. Which is to say that the social conception of experimental facthood isan idea with a history; it arises at a particular time, in particular con-tingent circumstances.

The Rhetoric of Experimental Reports and theConstitution of Experimental Fact

Recent work on mid-seventeenth-century experimental science, partic-ularly that of the Royal Society, has called attention to a very interest-ing feature of the way in which experimental results are reported. Theclaim is that in the mid-seventeenth century, experimental reportsbecome quite radically particular in contrast to what they had been inearlier writers. In earlier writers, it is claimed, experimental reports aregiven in quite general terms: Such and such may be observed in suchand such circumstances. It is striking, though, when one turns to thereports of experimental results in the Royal Society, for example, howparticular they are; what one finds characteristically is the report ofexactly what was observed to happen in a particular place, at a partic-ular time, with particular equipment, and particular people in atten-dance, both the successes and the failures. This, indeed, seems to be amatter of conscious policy. Sprat writes, “Whatever they have resolv’dupon; they have not reported, as unalterable Demonstrations, but aspresent appearances: delivering down to future Ages, with the goodsuccess of the Experiment, the manner of their progress, the Instru-ments, and the several differences of the matter, which they haveapply’d: so that with their mistake, they give them also the means offinding it out.”65


but it isn’t an answer to the questions that interest me most. Shapin and Schaffer seemto take it for granted that the very idea of experimental science carries with it a social criterion of experimental facthood. See, e.g., Shapin and Schaffer, Leviathan and the Air-Pump, pp. 25ff., 77–78, 225–26, 281–82. My interest is in the circumstances underwhich this criterion first arose.

65 Sprat, The History, p. 108.

Dear has been particularly insistent on this point in a series of penetrating articles.66 Dear relates this change to the rejection of an Aristotelian conception of natural philosophy. He writes:

“Experience” as an element of scholastic natural philosophical discourse took the form of generalized statements about how things usually occur; as an element of characteristically seventeenth-century, non-scholasticnatural philosophical discourse it increasingly took the form of statementsdescribing specific events. . . . For the scholastic natural philosopher, writinghis commentaries on Aristotle, the grounding in experience of the physicalfacts debated in his discussions was guaranteed by their generality as experi-ential statements – “heavy bodies fall” is a statement to which all could assent,through common experience embodied in authoritative texts. . . . The new“experience” of the seventeenth century . . . established its legitimacy in his-torical reports of events, often citing witnesses.67

This apparently stylistic difference between the old and the new is actu-ally quite substantive, Dear argues. When experience functions as theillustration of the universal statements that constituted the startingplace of a scientific syllogism, as it does in Aristotelian science, there islittle reason to expect controversy; all will agree that stones fall and firerises. But in the new experimental science, particularly as practiced inthe Royal Society, experiment functions to create novel facts. And herethe situation is quite different. And when we are dealing with novelfacts, there is a possibility for controversy that simply did not exist inearlier, Aristotelian science. Dear writes, “Controversy, however, or thethreat of controversy, demanded more radical measures, and at thesame time placed greater emphasis on discrete events as justificationfor assertions.”68 When experiment makes novel claims, Dear argues,then the reporting of an observation or an experiment has a new func-tion, not that of reminding the reader of something already known butthat of actually convincing the reader that the conclusion reported actu-ally happened. This was done, Dear claims, by particularizing the report

322 larger visions

66 Peter Dear, “Totius in verba: Rhetoric and Authority in the Early Royal Society,” Isis76(1985), pp. 145–61; Dear, “Jesuit Mathematical Science and the Reconstitution of Experience in the Early Seventeenth Century,” Studies in History and Philosophy of Science18(1987), pp. 133–75; Dear, “Narratives,” See also Shapin and Schaffer, Leviathan and theAir-Pump, pp. 60ff.

67 Dear, “Jesuit Mathematical Science,” p. 134. The quotation is offered by Dear as asummary of the main argument of Dear, “Totius in verba.”

68 Dear, “Jesuit Mathematical Science,” p. 169.

and making it the report of a particular witness on a particular occa-sion, a procedure that bears an obvious relation to legal reasoning, asShapin and Schaffer note.69 Shapin and Schaffer go on to argue thatan important point of this new style of presentation is to give the readerfaith in the truth of the outcomes reported by giving him faith in thescientist producing those outcomes. In the case of Boyle, they argue,“It was the burden of Boyle’s literary technology to assure his readersthat he was such a man as should be believed. He therefore had to findthe means to make visible in the text the accepted tokens of a man ofgood faith.”70 And what made Boyle a credible witness was detail upondetail that made the story credible as a report of something that actu-ally happened in the world in a particular place and at a particulartime.71

The phenomenon that Dear, Shapin, and Schaffer are pointing to is certainly quite real; although one can certainly find earlier writers who appear to be presenting direct reports of actual events and laterwriters who present their experimental results in general terms, thereis certainly a general trend in the experimental literature toward moreand more particularity in reporting the results of experiments. AndDear has certainly made the case that in some circumstances, at least,this increased particularity is connected with the problem of convinc-ing an audience to accept novel and unexpected results.72 But, I think,novelty and the rejection of an Aristotelian conception of the functionof experience in natural philosophy are not the only factors at work here.

Let me begin by noting that the use of general statements in report-ing the outcomes of experiments is not necessarily connected eitherwith an Aristotelian conception of the use of experience or with thereporting of non-novel facts. Take the case of Francis Bacon. Bacon


69 Shapin and Schaffer, Leviathan and the Air-Pump, pp. 56–57.70 Shapin and Schaffer, Leviathan and the Air-Pump, p. 65. On the notion of a literary tech-

nology, see p. 25. The style has another function for Shapin and Schaffer, to bring thereader into the community of experimenters, and thus make him a “virtual” witness ofthe experiment, contributing to its success in constituting an experimental fact. See pp. 60, 63.

71 One should not overestimate the degree of detail in Boyle’s experimental reports. In theProëmial Essay to his Certain Physiological Essays and Other Tracts (1661), Boyle notes thathe often leaves out important particulars in reporting his experiments, for a variety ofreasons. See Robert Boyle, The Works of the Honorable Robert Boyle in Six Volumes, ed., ThomasBirch, 2nd ed. (London: 1772), vol. I, pp. 315–16.

72 See, e.g., Dear, “Jesuit Mathematical Science,” p. 169; “Narratives,” p. 163n.

takes great pains to emphasize that his use of experience is quite dif-ferent from Aristotle’s. His is a new organon, a completely different wayof going from experience to a knowledge of the world; for Bacon, asfor his later followers, it is quite clear that his science will be built froma collection of facts, many of which will be quite novel. Yet, as we haveseen, Bacon is quite clear that the statements of fact that make up thebulk of his natural histories will be general rather than particular.Although “history is properly concerned with individuals, which are cir-cumscribed by place and time,” he writes in De Augmentis, “because ofthe general resemblance which in most cases natural objects of the samespecies bear to one another,” natural history most often deals withgeneral statements about species of things: “when you know one, youknow all.”73

Consider also the case of Gilbert. Gilbert takes great pains to empha-size the originality of his exploration of the magnet. He writes, “Thisnatural philosophy [physiologia] is almost a new thing, unheard-ofbefore; a very few writers have simply published some meagre accountsof certain magnetic forces. . . . Our doctrine of the loadstone is con-tradictory of most of the principles and axioms of the Greeks.”74 But,despite the self-conscious novelty of the experiments that Gilbert isreporting, the form of the reports is decidedly general. Often Gilbertsimply reports the properties he has observed (numerous times, pre-sumably) in his lodestones. For example, he writes, “Iron rubbed andexcited by a loadstone is seized at the fitting ends by a loadstone morepowerfully than iron not magnetized.”75 Even when more complicatedand more directly experimental facts are related, they are given in a rel-atively nonparticular way: “A concave hemisphere of thin iron, a finger’swidth in diameter, is applied to the convex polar superficies of a load-stone and properly fastened; or an iron acorn-shaped ball rising fromthe base into an obtuse cone, hollowed out a little and fitted to thesurface of the stone, is made fast to the pole. . . . Fitted with this con-trivance, a loadstone that before lifted only 4 ounces of iron will nowlift 12 ounces.”76 Although the important details are there, Gilbert givesus only the results of the experiment, and those in general terms; inhis report there are no indications of time or place, who performed theexperiment or observed it, how many times the experiment was per-formed, what difficulties there might have been in constructing the

324 larger visions

73 De augmentis II.1, Bacon, The Works, vol. VIII, p. 407; see also ibid., p. 406.74 Gilbert, De magnete, p. 1. 75 Ibid., p. 159. 76 Ibid., p. 137.

apparatus, and so on – all the features that we can find in later RoyalSociety experiments, such as those of Boyle. Yet, what Gilbert was relat-ing was decidedly novel, as he fully recognized.

The new importance of novel facts in science cannot completelyexplain the new forms that experimental reports took; the importanceof novel facts was recognized without necessarily resulting in anychanges in the way in which experimental results were reported. Whatother factors are relevant here? Why did the Royal Society find it nec-essary to couple novelty of results with a new form of presentation forthose results? My suggestion is that we look to the change in the con-ception of experimental facthood that I have been developing in thisarticle.

I have tried to show that with the Royal Society, we have a new conception of experimental facthood. For earlier investigators, it was possible for an individual working entirely alone to establish anexperimental fact, either through simple repetition of a trial or throughreasoning. And so, when a Gilbert or a Bacon or a Descartes reportsthe outcome of an experiment, he can report it as fact; others may chal-lenge what he claims to have established, but the epistemology of exper-imental facthood does not in any way demand the concurrence ofothers to constitute a fact. But, I have argued, matters are entirely dif-ferent with respect to the conception of experimental facthood in theRoyal Society. There it is essential that others perform the experimentand witness the results before a purported experimental fact can enterthe register of attested facts. And so, when an experimenter reports theoutcome of an experiment, or even a series of experiments, he is notreporting anything that could possibly be an experimental fact; factscannot be established in that way. And so, the best that can be reportedis, as Sprat puts it, “present appearances,” the way things looked to anindividual at a given time in a given place. Only by putting this togetherwith the observations of others can we constitute a fact. And so, Isuggest, it is no surprise that new conventions for reporting the outcomes of experiments come at the same time as the Royal Societyis explicitly rethinking how it is that experimental facts are to be established.

Community and Fact

In the earlier parts of this article, I have been tracing the developmentof a social conception of experimental facthood or, better, the explicit


recognition of the social character of experimental facts. This is an interesting claim in the history of the philosophy of science, perhaps.But there is something even more interesting going on here, I think.

It seems clear that at very least, the social criterion of experimentalfacthood that I discussed in connection with the Royal Society pre-supposes certain social structures. To consider just one way in which this is true, consider the strong notion of reproducibility. On thatnotion, as embodied in the communal conception of facthood, to be a candidate for an experimental fact, a given experimental result must be capable of being reproduced by different hands and eyes, and it requires the consensus of the scientific community as a whole. Itis important here that not just anyone can participate in this enter-prise.77 If an experiment is performed by a member of the community,and I, for example, cannot reproduce it, that would not necessarilycount against it in the least; standing outside of the community, I am not competent to cast my vote for or against a purported experi-mental fact. But, on the other hand, if others in the community couldnot reproduce an experiment, that might count against it. And so, thevery standard of strong reproducibility would seem to presuppose somecriterion for membership in the community of peers. In a similar way,it presupposes various kinds of social structures that are relevant todoing experiments and evaluating their outcomes in an appropriatelypublic way.

It is important to recognize that the social structures necessary forone to be able to adopt the Royal Society’s conception of experimen-tal facthood were not always present in society. Indeed, the communitynecessary to support such a conception of science was created only inthe mid-seventeenth century, and then quite explicitly to enable itsmembers to realize such a communal conception of scientific activity.This is not to say that there were not communities before the mid-seventeenth century. To be sure, there were communities, there wereschools and universities, there were even academies and scientific soci-eties of a sort. But (and this seems quite crucial to me) they were notorganized in a way appropriate for the performing and reperformingof experiments or for the communal judging of the outcome of

326 larger visions

77 Sprat, The History, p. 344, claims that virtually anyone, no matter how idle or industrious,how learned or ignorant, can participate in the program of experimental science. But, ofcourse, in practice, this was not so.

experiments. This is not to say that such communities could not havearisen before then. Descartes might perhaps have transformed theJesuit fathers of La Flèche or the Collège de Clermont or the membersof the Mersenne Circle into such a group. But he did not, and no oneelse did either. Such a community might also have come with Bacon aswell. But even though Bacon dreamed of a community of gatherers offacts and gave it many tasks and an elaborate organization, he neverdreamed that they would cooperate with the production of facts, andthe structure he proposed assumed that the many workers in the House of Salomon would work alone. This suggests to me that we must view the rise of the new communal conception of experimentalfacthood, a feature of the way practitioners thought about their naturalphilosophy, as intimately connected with the social transformation ofthe institutional structure in which science (natural philosophy) isdone. I do not know which, if either, came first – the social transfor-mation or the philosophical transformation. But it seems clear that thetwo must go hand in hand. Thus, of course, does not answer the ques-tion as to why the social conception of experimental facthood arosewhen it did. But it does suggest a direction in which we might look foran answer: The rise of the social conception of facthood must go handin hand with the emergence of the institutions appropriate to itssupport.

This leads me to a final moral. It has recently become very fashion-able to press the social factors in experimental facthood and the rolethat the community plays in the establishment of experimental facts.Indeed, the importance of social factors in recent experimental sciencehas led some to the view that the establishment of experimental facts can be explained entirely in sociological terms. On their view,establishing an experimental fact is simply a matter of social negotia-tion among members of the relevant community. With regard to theconcept of experimental facthood in the Royal Society, Shapin andSchaffer write that “the objectivity of the experimental matter of factwas an artifact of certain forms of discourse and certain modes of socialsolidarity.”78 Indeed, they go so far as to claim that matters of fact are “social conventions,” the result of “negotiations between experi-menters.”79 “A fact,” Bruno Latour writes in a similar spirit, “is what iscollectively stabilized from the midst of controversies when the activity


78 Shapin and Schaffer, Leviathan and the Air-Pump, pp. 77–87; cf. p. 25.79 Shapin and Schaffer, Leviathan and the Air-Pump, p. 226; see also pp. 281–82.

of later papers does not consist only of criticism or deformation butalso of confirmation.”80

While I have considerable sympathy for the view, I think that it hassome historical limitations. One might take it for an almost a prioritruth: Belief in experimental facts, as in everything, must simply be afunction of some communal agreement or other, explicit or tacit; belief,one might claim, as with the language in which it is framed, is by itsnature social, and whatever Descartes or Bacon or anyone else mighthave thought about it, they, too, were caught up in the invisible web ofsocial structure. Understood in this way, the thesis would seem to begrounded in very, very general facts about language, belief, and society,largely independent of any particularities about history and circum-stance. Regarded in this way, though, the thesis is a general philo-sophical claim, one largely without any special interest to the historianor philosopher of science. But if the sociological claim is taken to be athesis with real content and relevance for the historian of philosophyand science, then I think that, at best, it can only be an account thatholds for experimental science as practiced in the last 350 years or so,since the appropriate social (and intellectual) structures were simplymissing before then.

But, even when the social constructivist is suitably historicized, I havemy doubts. The thesis that the world of facts established by science issimply a matter of social agreement has an obvious deflationary con-sequence for the whole enterprise of science, turning what was thoughtto be objective fact into the collective illusion of a particular commu-nity. It would be a great irony if the social criterion of experimental fact-hood that, in a sense, marks the beginning of modern experimentalscience also marks the beginning of its demise.

328 larger visions

80 Latour and Woolgar, Laboratory Life, p. 243; Bruno Latour, Science in Action (Cambridge,MA: Harvard University Press, 1987), p. 42. For the more general account, see Latourand Woolgar, Laboratory Life, pp. 174–83, 236–52; and Latour, Science in Action, pp. 41–44.This, of course, is at one extreme of those who call themselves social constructivists. Thereis a wide variety of such views in the literature, many too many to survey in this shortarticle.

SOURCES

I. Historiographical Preliminaries

1. “Does History Have a Future? Some Reflections on Bennett andDoing Philosophy Historically” was originally published in P. Hare, ed.,Doing Philosophy Historically (Buffalo: Pergamon Press, 1989), pp. 27–43.It is reprinted with permission.

II. Method, Order, and Certainty

2. “Descartes and Method in 1637” was originally published in Frenchas: “Descartes et la méthode en 1637,” in J.-L. Marion and N. Grimaldi,eds., Le Discours et sa méthode (Paris: Presses Universitaires de France,1987), pp. 65–87. It was later published in an English version as“Descartes and Method in 1637,” in A. Fine and J. Leplin, eds., PSA1988: Proceedings of the 1988 Biennial Meeting of the Philosophy of ScienceAssociation (East Lansing: Philosophy of Science Association, 1989), vol.2, pp. 225–36. It is reprinted with permission from both the PressesUniversitaires de France and from the Philosophy of Science Association.

3. “A Point of Order: Analysis, Synthesis, and Descartes’ Principles”(with Lesley Cohen) was originally published in Archiv für Geschichte derPhilosophie 64 (1982), pp. 136–47. It is reprinted with permission of theArchiv für Geschichte der Philosophie and Walter de Gruyter and Co.(Berlin).

4. “J.-B. Morin and the Second Objections” was originally published inRoger Ariew and Marjorie Grene, eds., Descartes and His Contemporaries:

329

Meditations, Objections, and Replies (University of Chicago Press, 1995),pp. 63–82. © 1995 by The University of Chicago. All rights reserved. It is reprinted with permission.

5. “Descartes and Experiment in the Discourse and Essays” was originallypublished in Stephen Voss, ed., Essays on the Philosophy and Science of René Descartes (New York and Oxford: Oxford University Press, 1993),pp. 288–310. Copyright © 1993 by Oxford University Press, Inc. Usedby permission of Oxford University Press, Inc.

6. “Descartes on Knowledge and Certainty” was originally published in J.-R. Armogathe and Giulia Belgioioso, eds., Descartes: PrincipiaPhilosophiae (1644–1994) (Naples: Vivarium, 1996), pp. 341–63. It isreprinted with permission.

III. Mind, Body, and the Laws of Nature

7. “Mind, Body, and the Laws of Nature in Descartes and Leibniz” wasoriginally published in Midwest Studies in Philosophy 8 (1983), 105–33.It is reprinted with permission.

8. “Understanding Interaction: What Descartes Should Have Told Elisabeth” was originally published in Southern Journal of Philosophy21 supp. (1983), 15–32. It is reprinted with permission.

9. “How God Causes Motion: Descartes, Divine Sustenance, and Occa-sionalism” was originally published in Journal of Philosophy 10 (1987),pp. 567–80. It is reprinted with permission.

10. “Descartes and Occasionalism” was originally published in StevenNadler, ed., Causation in Early Modern Philosophy (University Park, PA:Pennsylvania State University Press, 1993), pp. 9–26. Copyright 1993by The Pennsylvania State University. Reproduced by permission of thepublisher.

11. “Semel in Vita: the Scientific Background to Descartes’ Meditations”was originally published in Amélie Rorty, ed., Essays in Descartes’ Meditations (Berkeley and Los Angeles: University of California Press,1986), pp. 81–116. It is reprinted with permission.

12. “Forms and Qualities in the Sixth Replies” was originally publishedin French as: “Formes et qualités dans les Sixièmes Réponses,” in J.-M.Beysadde and J.-L Marion, eds., Objecter et répondre (Paris: Presses

330 sources

Universitaires de France, 1994), pp. 449–69. It is reprinted in transla-tion with permission.

IV. Larger Visions

13. “Descartes, or the Cultivation of the Intellect” was originally pub-lished in Amélie Rorty, ed., Philosophers and Education (London and NewYork: Routledge, 1998), pp. 124–38. It is reprinted with permission.

14. “Experiment, Community, and the Constitution of Nature in theSeventeenth Century” was originally published in Perspectives on Science3 (1995), pp. 173–205. © 1995 by The University of Chicago. All rightsreserved. It is reprinted with permission.

sources 331

INDEX

Alquié, F., 56n, 57n, 59n, 164nAmesius, Guilelmus, 195nanalysis, 243–244

digressions in, 45and the Meditations, 52–63and synthesis, 35n, 52–63, 78–84,

87nsynthesis and the Principles, 52–63,

79, 84Anguilonius, Franciscus, 298–299animals, Cartesian conception of, 270Anscombe, E., 222nAquinas, St. Thomas, 16, 195–198,

280Ariew, Roger, 59n, 214nAristotle, 16, 222–223, 229n, 280,

298–299, 305n, 324Armogathe, J.-R., 83n, 95n, 114nArnauld, Antoine, 172nastrology, 66–67astronomy, 298natomism, 66, 69

temporal, 193–194authority, 70n, 314–315

Descartes on, 278–281and the teacher, 291–295

Averroes, 17Ayers, Michael, 5

Bacon, Francis, 102, 301–307,316–318, 319–320, 323–324, 327

333

and Descartes, 102, 307–308, 312,314, 315–316

Baillet, A., 71nBarnes, Jonathan, 299nBattail, J.-F., 203n, 210nBeck, L.J., 35n, 39n, 46nBeeckman, Isaac, 2, 47–48Belgioioso, Giulia, 114nBennett, Jonathan, 6, 13–30Bérule, Cardinal Pierre de, 67Beyssade, J.-M., 56n, 57n, 84, 193–194Bizer, Ernst, 195nBloch, Olivier, 263nBoas, Marie, 224nbody

existence of, 45, 105–106, 214–216,244–255

nature of, 26–27, 112, 179–180, 253wax example, 45, 244–246

Boehm, A., 227nBourdin, Father, 286nBoyer, Carl, 95n, 108nBoyle, Robert, 2, 297, 320n, 323Brahe, Tycho, 66Broad, C.D., 143n, 150nBrockliss, L.W.B., 280n, 286n, 292nBroughton, Janet, 179nBrunschwig, J., 56n, 57n

causefinal, 155–157, 162–163

334 index

cause (cont.)secundum esse vs. secundum fieri,

192–193, 195total, 204–205

certainty, 7, 111–129, 287–288, 306childhood, 231–233, 266Clair, Pierre, 189n, 210nClarke, Desmond, 38n, 91n, 92n,

110n, 114n, 255nClauberg, J., 149n, 203clear and distinct perception, 105–107

validation of, 49–50Clerselier, Claude, 149n–150n, 203,

204n, 218Cohen, Lesley, 7, 35n, 52n, 106n,

243nColumbus, Christopher, 305nCopernicanism, 69, 76Copernicus, Nicholas, 66Cordemoy, G. de, 203, 218–219Costabel, Pierre, 37n, 65n, 89n, 136n,

137nCottingham, John, 54n, 59n, 114n,

199n, 277n, 284nCouturat, Louis, 56nCurley, E.M., 57–58, 59n, 63n, 84,

222n, 223n, 243n

Daston, Lorraine, 297Dear, Peter, 298, 300n, 322–325deduction, see intuition and deductionDijksterhuis, E.J., 224ndisputation, academic, 287–288distinction

mind-body, 257–273Doney, Willis, 134n, 159n, 181n, 212n,

223nDrake, Stillman, 180n

education, Descartes on, 20–23,277–295

Elisabeth, Princess, of Bohemia, 134,172, 176n, 204–205

Eustachius a Sancto Paulo, 63n, 228,272

experiment, 7, 41–43, 85–110,111–129, 296–328

and social structure, 326–327and theory, 109–110

external world, see body, existence of

fact, experimental, 296–328in Bacon, 301–307in Descartes, 108–109, 307–316in the Royal Society, 316–321, 325

Fitzpatrick, Edward, 287nFontialis, Jacobus, 200nform, substantial, 112, 196–200,

207–208, 219–220, 227–231,257–273

Frankfurt, Harry, 222n, 239nFreddoso, Alfred, 205n

Gabbey, Alan, 137n, 149n–150n, 225n

Gale, George, 139nGalilei, Galileo, 2, 180n, 290

condemnation of, 120Galison, Peter, 196Gassendi, Pierre, 2, 67, 70, 172n, 272Gaukroger, S., 34n, 93n, 137n, 164n,

190n, 225n, 240n, 284nGeach, P.T., 222nGeulincx, Arnold, 203Gilbert, William, 299–300, 305n,

324–325Gilson, Étienne, 39n, 147n, 175n,

195–196, 198n, 223n, 227n,228n, 229n, 231n, 242n, 292n

Girbal, F., 210nGlanvill, Joseph, 316–317, 318nGod

arguments for the existence of, 56,58, 71, 74–78

and divine sustenance, 163–164,189–202, 206–208, 209–210

and the laws of nature (motion),155–167, 181–186

and motion, 136, 163–165,181–186, 189–202, 206–210

as total cause, 204–205Gouhier, Henri, 48n, 57n, 169n, 171n,

179n, 203n, 204n, 210n, 222n,234n, 241n, 242n

index 335

Goujet, C.-P., 67ngravity, see heavinessGrene, Marjorie, 197nGrimaldi, N., 95nGueroult, Martial, 7, 26n, 54n–55n,

56–57, 84, 139n, 164n, 190n,214n, 225n, 229n, 241n

Hacking, Ian, 109Hall, Thomas, 250nHamelin, Octave, 144nharmony, pre-established, 135,

141–143, 160–161, 166nHatfield, Gary, 190n, 225n, 294nheaviness, 147–149, 175–177, 187,

197, 227n, 228–229, 259–260,267–268

Heppe, Heinrich, 195nHessing, Siegfried, 57nHintikka, J., 53nhistoriography, 3–6, 13–30Hobbes, Thomas, 2, 29, 143n, 181n,

237, 246n, 272n, 320nHoenen, P.H.J., 181n, 212nHooker, Michael, 53n, 91n, 242n,

255nhumanism, 280hypotheses, 115–116

hypothetical mode of exposition, 44

hypothetical reasoning, 93,115–116, 118–119, 121–129

Hyppolite, J., 57n

ideas, 246–247Iltis, Carolyn, 139nimagination, 245, 249–250impact, see interaction, body-bodyintellect, 237, 245, 249–250, 251–252,

277cultivation of, 283–295

interactionbody-body, 170, 172, 178–188,

205–209, 212body-mind, 213–218mind-body, 8, 133–167, 168–188,

209–213

intuition and deduction, 35, 48,86–87, 91–94, 99–103, 107, 117,119, 120–121, 124n, 281–283,308–309

intuition, 90–91, 305nvalidation of, 49–50

Iwanicki, Joseph, 71n

Jardine, Lisa, 56n

Kant, Immanuel, 16Kemp Smith, Norman, 144n,

149n–150nKenny, Anthony, 17, 169nKnappich, Wilhelm, 67nknowledge

of particulars, 111–129social factors in, 296–328

Knudsen, Ole, 137nKoyré, Alexandre, 149n, 222nKraye, Jill, 281n

La Forge, Louis de, 189–191, 198,201, 203, 204n, 209

Lalande, A., 308nLaporte, Jean, 144n, 168n–169n,

210nLarmore, Charles, 93nLatour, Bruno, 296n, 327–328, 328nlaws of nature (motion), 8, 25–26,

27n, 112, 180–184, 198, 206–208,229n

and God, 155–167and mind-body interaction,

133–167, 212–213Leeuwenhoek, A. van, 28Leibniz, G.W. von, 1, 8, 16, 26, 28,

133–167 passim, 168n, 181n,229n

Lennon, Thomas, 134n, 159n, 219n,231n

Lindberg, David, 227nLocke, John, 2Loeb, Louis, 170nLoemker, Leroy, 229nlogic, 46–47, 91, 285–287Lojacono, Ettore, 114n

336 index

Machamer, Peter, 157n, 164nMackie, John, 17magnet, 122Mahoney, Michael, 224nMaimonides, 17Malebranche, Nicolas, 134n, 159n,

165n, 182n, 189n, 202n, 203,209n, 219, 231n

Malpighi, Marcello, 28Marion, Jean-Luc, 38n, 95nMarlies, Mike, 242nMartinet, Monette, 65nMarx, Karl, 16Mates, Benson, 17mathematics, 284mathesis universalis, 38–39Mattern, Ruth, 179nMaurolycus, Franciscus, 108, 313mechanical philosophy, see philosophy,

mechanicalMersenne, Marin, 2, 64, 70–71, 75, 82method, 6–7, 33–51, 86–91, 117, 285,

308–311anaclastic line example, 36–37,

87–90, 94, 107–108, 117, 309–311Baconian, 102, 301–307and experiment, 91–103in the Meditations, 44–46the “noblest example,” 49–50

Milhaud, G., 38nMillet, Joseph, 54nmind-body distinction, see distinction,

mind-bodymind-body interaction, see interaction,

mind-bodymind-body union, see union, mind-

bodymiracles, 157–159More, Henry, 28–29, 172n, 184–185,

200Morin, J.-B., 7, 64–84, 262n, 264nMorris, Thomas, 205nMouy, Paul, 38n, 224n

Neugebauer, Otto, 298nNewton, Isaac, 26Nicholas, John, 89n

observation, theory-ladenness of,109–110

occasionalism, 8–9, 134, 142, 160,189–202, 203–220

Ohana, J., 137nOlscamp, Paul, 134n, 159n, 219n,

224n, 231nO’Neill, Eileen, 170norder of reasons, 44, 289–291

Pascal, Blaise, 17Patrizi, Francesco, 68Pedersen, Kurt, 137nPell, John, 313nPetit, Pierre, 313philosophy

mechanical, 2, 112–113, 156, 206,222–256 passim, 277

Scholastic, 68–69and science, relations, 1, 9,

24–29tree of, 2, 289

physicsCartesian, 133–167 passimcommonsense, 226, 231, 234epistemological foundations of,

222–256Leibnizian, 138–140mechanist, see philosophy,

mechanicalScholastic, 1, 112, 119–120, 227,

234. See also heaviness.Pickering, Andrew, 296npineal gland, 145–147Popkin, Richard, 222nPrendergast, Thomas, 164n, 229nProst, Joseph, 203n, 210nPurver, Margery, 316n, 317n

quality, real, 147–148, 175–177, 197n,227–231, 257–273

Quine, W.V.O., 1, 29

rainbow, 40–43, 94–104, 108–109,117–119, 311–312, 313–314

Randall, J.H., 53nRatio Studiorum, 287

index 337

reasonvs. common sense, 222–256 passimand experience, 104–110

Regius, Henricus, 222n, 286nRemes, U., 53nRemnant, Peter, 144n, 153–155Richardson, Robert, 169nRichelieu, Cardinal Alphonse-Louis du

Plessis de, 67Roberval, Gilles Personne de, 2Robinet, André, 219nRochot, B., 57n, 71nRodis-Lewis, Geneviève, 35n, 144n,

155n, 171n, 179n, 229nRohault, Jacques, 115–116Rorty, A., 83n, 105n, 214n, 243n,

294nRoyal Society of London, 297, 316–321

Sabra, A.I., 137nSchaffer, Simon, 296n, 297, 319n,

320, 320n–321n, 322n, 323, 327Schmitt, Charles, 280nSchooten, van, Franz the elder, 281nSchouls, Peter, 46nSchuster, John, 34n, 49n, 240nSédillot, M.L. Am., 67nsenses and sensation, 45, 104–110,

127–128, 232–235, 277, 279–280,298–299, 312–313

knowledge from, 222–256 passimand the Meditations, 104–107

Serres, Michel, 28nSerrus, Charles, 35n, 46nShapin, Steven, 296n, 297, 319n, 320,

320n–321n, 322n, 323, 327skepticism, 222–223, 235–242, 255,

279–280

dream argument, 238–239Soprani, Anne, 66nSpecht, Rainer, 203n, 210nSpinoza, Benedict, 13–30 passim,

143n, 168n, 185nSprat, Thomas, 316–321, 325, 326nSwerdlow, Noel, 298nsynthesis, see analysis, and synthesis

Thomson, G.T., 195ntransubstantiation, 61ntree analogy, see philosophy, tree ofTronson, Guillaume, 65nTurnbull, Robert, 157n, 164n

union, mind-body, 153–155, 171

vacuum, 26–27, 227n, 231n, 234Vair, Guillaume du, 65Virgil, 305nVoëtius, Gisbertus, 273Voss, S., 117n

Waard, C. de, 65nWahl, Jean, 190nWallace, William, 227nWatson, Richard, 169n, 170nwax example, see body, wax exampleWeber, J.-P., 34n, 39nWestfall, Richard, 224nWiener, P.P., 181nWilliams, Bernard, 4–5, 17Wilson, Margaret, 145n, 157n, 223n,

238n, 242n, 247n, 248n, 251nwisdom, levels of, 290Wood, P.B., 316nWoolgar, Steve, 296n, 328nWoolhouse, R.S., 157n

garber descartes embodied

Documents