SOURCE: “Mind Versus Flesh,” in The Battle of the Gods and Giants: The Legacies of Descartes and Gassendi, 1655-1715, Princeton University Press, 1993, pp. 106-37.

[In this excerpt, Lennon considers in depth Gassendi's Objections to René Descartes' Meditations. Focusing on the problem of representation, Lennon defends Gassendi from the charge, put forth by both Descartes and later critics, that he simply did not understand the nature of Descartes' method. Nevertheless, as Lennon argues throughout his book, the materialism that provided the foundation for Gassendi's critiques eventually could not compete with the dominance of Cartesian philosophy.]

MIND VERSUS FLESH

Early on Descartes had taken Gassendi to be, if not an authority, then at least someone to be regarded seriously in optics, astronomy, and other matters.¹ With Gassendi's Objections, however, Descartes's attitude changes dramatically. On June 23, 1641, he returned to Mersenne Gassendi's objections along with the advice to have them printed without showing them to Gassendi, who when he saw how bad his objections were would want them “suppressed.” Descartes meanwhile was loathe to see his time in replying wasted or the possibility realized that some would think that it was he who, unable to answer the objections, had them suppressed. He concludes the letter: “you will see that I have done all I could to treat Gassendi honorably and gently; but he gave me so many occasions to despise him and to show he has no common sense and can in no way reason, that I would have done too much less than my duty had I said less, and I assure you I could have said much more.”²

A month later Descartes again wrote to Mersenne, saying that Gassendi had no grounds for complaint at his treatment, for he gave only equal in kind despite what he had always heard, namely, that the first blow is worth two and that thus to be really equal the reply should have been doubled. “But perhaps he was affected by my replies because he recognized their truth, while I was not for an entirely different reason; if so, it is not my fault.”³ Two years later Descartes could still muster respect for Gassendi's empirical astronomy,⁴ but by then he could tell just from the index of Gassendi's letters that they contained nothing he needed to read.⁵ The literature has tended to fault Gassendi for failing really to engage Descartes's views; the converse seems no less true. For the most part Descartes treated his would-be adversary as beyond contempt. Discussing why criticism of his work is no burden Descartes wrote that Gassendi's Instantia did not displease him as much as he was pleased by Mesland's judgment that there was nothing in the work not easily answered.⁶

As for Gassendi's attitudes, there are two versions of the story. One has it that he was no less than livid with Descartes and no less than he in his acrimony and petulance. But this version is based on the testimony of Jean-Baptiste Morin, an eccentric to say the least, who on other grounds was concerned to besmirch the reputation of Gassendi. Thus, for example, Descartes's calling Gassendi “Flesh” in his Replies would be understood only as justified retaliation for having been referred to as “Mind” in the Objections. More credible is the account of Bernier, who portrays a rather more detached, long-suffering response from his teacher. In any case, the personal differences between the great antagonists were finally repaired in 1648 following a dinner arranged for them by the Abbé d'Estrées. As it happened, Gassendi was unable to attend because of illness; but after the dinner, the assembled party visited Gassendi, who was embraced by Descartes.⁷

Gassendi's Objections may well be regarded as unique.⁸ Among other objectors, Arnauld, who really asks only for clarifications, and the Scholastic theologians share important metaphysical presuppositions with Descartes. Hobbes does not engage Descartes so much as merely juxtapose his own views to those of Descartes, with Descartes replying in kind. In the case of Gassendi, however, we find an elaborated and systematic metaphysical confrontation. The length of his exchanges with Descartes thus reflects their relative importance. The Fifth Objections (1641) are more than twice as long as any other set, and if Descartes thought them “not the most important”⁹ he nonetheless replied to them at greatest length. Within a year Gassendi had responded with his Rebuttals (Instantiae: literally, follows-up) which with the Fifth Objections and Replies were published in 1644 under the general title of Disquisitio metaphysica, totaling some 150 pages in folio of the Opera omnia.

At the core of Gassendi's critique of Descartes is the notion of representation. There are arguments from Gassendi against Descartes, as there were later to be from Locke and Foucher against Malebranche,¹⁰ to the effect that an idea to represent a square would have to be square. At one point, for example, Gassendi poses the following dilemma, designed to argue that the mind of which Descartes claims to have an idea is essentially extended. Only if an idea has extension can it represent extension; only if it has shape and location can it represent what has shape and location, for an idea must be like what it represents. But only if an idea has no extension, shape, or location can it be joined to the essentially unextended, unshaped, unlocated thing that the mind is alleged to be, for the essence of the mind is supposed to be thought, which excludes properties of extension.¹¹

The dilemma may be released, as Gassendi urges, by denying that the self is an essentially unextended, unshaped, unlocated thing. But this does not explain how Gassendi, or any imagist, deals with the first horn of the dilemma. There seem to me three possible answers to the problem of imagist ideas representing extension. Historically they overlap. One is the skeptical answer, to which Gassendi is certainly inclined, namely, that the idea of a square, for example, just does not represent at all; we do not know the real. But this response by itself is insufficient. Even if the idea does not represent a real square, it is a different idea from the idea of a circle, and it is not clear that the problem of distinguishing them is any more tractable than the problem of how an unextended idea can represent an extended thing. In fact, it might be argued, these are versions of the same problem. It is not surprising, then, that the skeptical answer is to be found combined with others among the philosophically more astute.

A second possible answer, one emphasized by Locke, is to give an account of resemblance that would allow the unextended to resemble and thus to represent the extended. I do not have in mind here the realist account in terms of the same thing existing in different ways—a mental, nonspatial square being qualitatively identical to, and thus representing, a material, spatial square. Many agree that this is the account advanced by the orthodox Cartesians. … It is also the account advanced by Leibniz, according to whom even ideas of secondary qualities resemble. The idea of pain resembles, although confusedly, a certain motion in the body, for example.¹² There are many reasons, however, why this account is inimical to the tradition whose view of ideas I am discussing here. The one of immediate relevance is that this realist account is incompatible with the skeptical answer above, for it grounds knowledge of the real. Far from cutting us off from the real, this account of ideas gives us immediate access to it. As Arnauld put it, the idea is the thing itself insofar as it exists in the mind.¹³

Instead, the possible answer I have in mind would have it that an unextended idea resembles an extended thing just in case it stands in a certain causal relation to it. Thus we may call it the causal answer. Later there will be a more appropriate occasion to develop Locke's version of this answer. The gist of it is that the idea of a square, and of primary qualities generally, resembles and thus represents its object in that given the corpuscularian hypothesis, the object could produce only that idea. The reality, inseparability, and resemblance to ideas of primary qualities come to the same thing. Given the corpuscularian hypothesis, the world must have certain features in order to appear to us as it does. Secondary qualities, on the other hand, are imputed, separable, and fail to resemble. Even given a fully articulated corpuscularian account, there is no reason why the motion that in fact produces the idea of pain should not produce the idea of redness. That it does not is attributable to the biological utility of the former arrangement. Still, even ideas of secondary qualities are said by Locke to represent and may be said to represent their (partial, remote) causes. However contingent it might be, the fact of the matter is that lemons taste tart and pineapples sweet. And we perceive them in this way at least partly because of their fine structure, between which and our perceptions there is a regular, if mechanistically inexplicable, correlation. Thus, Gassendi's dilemma is generally resolved by denying the crucial premise of the first horn. Representation does not depend on resemblance in the sense that representing idea and represented object both have the same property, but rather on the causal connection between them. The historical irony is that Locke's actual version of the causal answer allows for resemblance in a way that construes ideas as extended and the thing represented as unextended. … The short of it is that (1) ideas are impressions of which we are aware, that is, corporeal states or motions, and are thus extended, and (2) extension is relational and thus only phenomenal, and hence the noumenal object represented by an idea is unextended.

Ideas are also extended according to the third answer, the materialist answer. According to it, the idea of a square represents a square because it actually is square. This is to embrace the first horn, suitably interpreted, as it is for Locke in the causal answer just given. The historical plausibility of my thesis here will be enhanced by noting that, although essentially materialistic, this tack was taken by Berkeley. Consider, for example, his argument with respect to the intense degree of heat as pain, the very first argument of the Three Dialogues. No unperceiving thing is capable of pain; a material substance is unperceiving, hence incapable of pain; but the intense degree of heat is pain, thus material substance is incapable of, that is, cannot have as a property, the intense degree of heat. And, of course, what is true of the intense degree of heat, we later learn, is true of all properties. Given his own analysis of objects as bundles of ideas or sensations, Berkeley is not entitled to the first premise, and he effectively rejects it. More generally, ideas and only ideas have the properties they appear to have, so that when Berkeley takes representation to depend on resemblance in the sense that representing idea and represented object both have the same property, representationalism comes under attack. As Berkeley insists repeatedly, an idea can resemble only another idea. The moral of the story is that the essential element in this approach to ideas is not the invitation to materialism, but the construal of ideas as images. Just how Berkeley distinguishes his view from the materialists is a question beyond the scope of my story. Here it will suffice to show how the imagist conception of ideas ties together the three answers to Gassendi's dilemma.

Taken no further, the dilemma Gassendi poses for Descartes remains at a superficial level. A deeper concern is the whole notion of representation that Descartes employs in extrapolating from ideas to things, or more precisely, from the realm of thought to that of essences. According to one particularly relevant interpretation, the key to Descartes's theory of ideas is his analytic geometry, which provides the model for representation: An idea represents in the way in which the algebraic equation for a curve expresses that curve by giving a rule for the deduction of all its properties.¹⁴ I am not quite convinced that this Leibnizean reading can be wrung from the Cartesian texts. At a minimum, however, it is true that for Descartes an idea is an intelligible rather than sensible representation. As opposed to Gassendist images, which are particular, it is universal. Gassendi has been charged by later commentators with having failed to see how different a theory of ideas he was dealing with, or worse, with having just failed to understand it. But of this Gassendi is innocent, for he correctly saw that Cartesian knowledge would be abstract in just the pejorative sense in which Descartes had castigated Aristotelian-Scholastic knowledge as abstract (and in which Bernier, as we have just seen, was to castigate Gassendi's conception of time). Were there Cartesian ideas, Gassendi argued, they might lead to what is true of a thing, but failing to be an image of it, they would not provide real knowledge of it. How so?

Gassendi does not object to Descartes's conclusions, at least not to his principal ones concerning the existence of God and the immortality of the soul.¹⁵ Instead, in a way that adumbrates Locke's reaction to Cartesianism, he attacks the arguments for Descartes's conclusions, and especially the kind of argument they exemplify. Thus, for example, we find Gassendi taxing Descartes for claiming that his metaphysical demonstrations are more certain than those of mathematics when they do not even get general assent. The following is paradigmatic of the offending sort of argument:

1. Things I consider as conceptually distinct are really distinct.

2. I consider thinking substance and corporeal substance as conceptually distinct. Therefore,

3. Thinking substance and corporeal substance are really distinct.¹⁶

The first premise expresses a necessary connection between concepts that is apprehended by a metaphysical or intellectual intuition; the second expresses an instantiation of the same concepts apprehended by the same intuition. Now, to be sure, Descartes rejects this syllogized version of his reasoning that Gassendi attributes to him.¹⁷ For Descartes, the conclusion of what is really an intellectual induction is the first premise, drawn on the basis of the second premise and the conclusion. But it is precisely this induction and the intuition on which it rests that Gassendi rejects and that lies at the heart of his critique of Descartes.¹⁸ For Gassendi, as for Locke, experience is the only source of knowledge. For Locke, as for Gassendi, we have no intellectual intuition and thus his rejection of the innate ideas claimed by Descartes and others is of a piece with his rejection of the ideas Malebranche claimed to see in God. Gassendi shows the way, effectively by rejecting the nonsensuous half of Republic's Divided Line.

Consider his reaction to the Cartesian method of doubt, a topic on which as a skeptic he is especially sensitive. In the Disquisitio¹⁹ he argues that contrary to what is claimed for this method, it is impossible to free the mind from its prejudices. The attempt to do so is, we might say, (1) psychologically, (2) methodologically, and (3) epistemologically futile. Let us consider each of these aspects, which are important even quite apart from the issue of intellectual intuition.

(1) The psychological futility of Cartesian doubt is of a piece with a skeptical position advanced since antiquity. Gassendi's point is that we just cannot avoid making judgments that we are accustomed to making such as that the intersection of two lines produces angles equal to two right angles or that a certain round object shines. And his point perfectly accords with the recommendation of traditional skeptics that we suspend judgment whenever it is possible to do so. For it is possible to suspend judgment in their view only with respect to metaphysical issues, that is, the essences of things. In all other cases, that is, in sense perceptions, nature breaks the skeptic by constraining assent.

Gassendi is often portrayed as having missed, not just Descartes's theory of ideas, but more especially the nature of his method of doubt. Gassendi failed to distinguish, so the story goes, between the intrinsic doubt of convinced skeptics and the instrumental doubt Descartes uses against them, between real and practical doubt on the one hand, and the merely theoretical doubt advocated by Descartes on the other. Descartes did not really doubt he had a body, but only feigned such doubt in order better to support his belief that he did. But in fact Gassendi shows remarkable sophistication in replying to this charge of oversight as it was first raised by Descartes himself. To be sure, “the distinction ‘between the acts of daily life and the inquiry after truth’ is totally justified.”²⁰ The skeptics reject indifference in questions of daily life (and, we might add, in precisely the same conservative way Descartes did with his provisional morality).²¹ With respect to the inquiry after truth they distinguish between phainomena, “things which appear to the senses,” which they accept, and noumena, “the things which are understood by the mind,” which they do not.²² Thus the dogmatist Descartes gets it exactly the wrong way round. This assertion is no argument, but it at least shows that Gassendi understands perfectly well what is at issue.

(2) Gassendi's charge under the rubric of what I have called the methodological futility of Cartesian doubt is rather ill-defined.

Secondly, even if I granted that the mind is liberated and is like a tabula rasa on which no judgment has been traced, you assume that “it can deduce some conclusions from principles”;²³ but this too appears to be impossible. Clearly if it has no preconceived notions [praejudicia], then it does not have any principles; for principles, as they are here understood, are statements [enunciationes]; and statements are kinds of judgments in which something is either affirmed or denied. Hence these principles will be judgments, and inasmuch as they are conceived in advance, they are preconceived notions. Therefore, if the mind has no preconceived notion in it, neither will it have any principles from which it may deduce something … since no new evidence appears which convinces us that any different relationship than the one above is to be enunciated, it follows that the statements about reality will be neither new statements, nor opposite ones, but the same ones as before, and so the same preconceived notions will crop up again.²⁴

Essentially he seems to be saying that even if the doubt Descartes recommends were possible, it could not yield the result he thinks. Gassendi claims that nothing really new emerges from the method of doubt as a deconstruction of all the propositions we have held; that is, having broken down these propositions (enunciationes) into their constituent elements, we then put them together again in exactly the same way. Now, of course we know that the Cartesian restoration of the house of belief may look room-for-room identical; but its novelty is the supposed foundation on which it rests. And it may be to this very supposition that Gassendi objects.

In addition, Gassendi may be arguing “holistic empiricism.” Though the mind is initially a tabula rasa, and though all knowledge is empirically derived, we need principles to derive principles. We may have here, from the principal source of seventeenth-century empiricism, a prototype of more recent arguments against concept empiricism. Epistemologically, it takes a pair of tongs to make a pair of tongs.

Finally, Gassendi may partially misunderstand Descartes, whose principles are not statements, that is, things we put together, but common notions or innate ideas, that is, structures that are given. This is the realism to which Descartes, whatever the details of his theory of innate ideas, is clearly committed. In these terms, however, the issue is the very fundamental one we have been discussing of whether we have the intellectual intuition to apprehend these structures that give us the essence of things.

(3) Under the heading of the epistemological futility of Cartesian doubt, Gassendi reiterates a point from above and focuses it on a familiar anti-Cartesian theme, namely, that the precious criterion of clarity and distinctness is a psychological phenomenon only.

Thirdly, granted that the mind may retain some principles from which it may draw conclusions, you assume that they are “not obscure and uncertain, but very evident and certain.” But, finally, this too is impossible, namely that there should be different principles from the ones that already were, that is to say some that are self-evident and certain, and the majority obscure and uncertain … after this liberation from preconceived notions, it remains equally inclined to assent to false principles as to true ones in the event that the former should turn up first and appear to be fact; and you cannot induce any reason why in this state of equal inclination false principles must come to the mind seeming only apparent and obscure and therefore uncertain rather than seeming evident and therefore certain. For the mind will be as it were at a crossroads, and it will be a matter of chance whether one set of principles offers itself, or another, and whether it accedes to the first, to the ones that make a stronger impression by the mere fact that they appear to be real, rather than to the others.²⁵

For Gassendi, the conviction reached by the Cartesian method of doubt is only psychological, with no guarantee of objective validity for either his concepts or his reasoning. For all we know, clarity and distinctness may be reliable, but how do we even know, asks Gassendi, which ideas are clear and distinct?

Gassendi's answer to the skeptical challenge he poses is a naturalistic empiricism. Thought has an empirically derived structure. It consists solely of material images produced in the brain, stored in the memory, and processed by analogy, composition, division, augmentation, and division.²⁶ The argument for the central role of these images and against the Cartesian intellectual intuition of essences is one found often enough in the materialist tradition. People affected by drugs or dreaming are furthest removed from the senses and least dependent on corporeal memory, but they are also least capable of thought.²⁷ Locke was to raise the same consideration as part of his argument against the Cartesian thesis that the soul always thinks: “This I would willingly be satisfied in, Whether the Soul, when it thinks thus apart [i.e., when we are asleep], and as it were separate from the Body, acts less rationally than when conjointly with it, or no: If its separate Thoughts be less rational, then these Men must say, That the Soul owes the perfection of rational thinking to the Body: If it does not, 'tis a wonder that our Dreams should be, for the most part, so frivolous and irrational; and that the Soul should retain none of its more rational Soliloquies and Meditations.”²⁸ The rest of Locke's argument will be discussed below.

Whatever Gassendi's arguments against Cartesian intuition, the question might plausibly be raised as to why images would be epistemologically inadequate to the Cartesian task. However ideas are produced, why do they not give us the essences of things? The crucial consideration here is Gassendi's nominalism, found explicitly in his first work, the Exercitationes of 1624,²⁹ and controlling his thought thereafter. In stating it, Gassendi is as blunt as Locke was to be in stating his: “nothing can be found that is not a unique thing.”³⁰ Nor at this level of philosophical analysis can there be much by way of argument apart from an appeal to parsimony. To invoke or ignore universals is the first step of a research program for which the confirmation or disconfirmation is very remote. The upshot, in any case, is that ideas qua images are particulars and contain no essences. They can contain only what by experience and reasoning we put into them. Gassendi clearly saw the importance of this difference from Descartes, which is worth pursuing textually in some detail.

Descartes in Meditations 3 had argued that the idea of God must be innate since (1) it is not derived from the senses; (2) “nor is it likewise a fiction of my mind for it is not in my power to take from or add anything to it.”³¹ Gassendi objected to this, claiming that in fact the idea of God was partly derived from the senses and partly composed by the mind, for obviously an idea can be added to. Indeed, this is precisely what we do in coming to know God more fully.³² Locke gives a similar account of our idea of God. The mind “enlarges upon” ideas of existence and duration, knowledge and power, which are ultimately derived from experience, making them “boundless,” that is, (potentially) infinite.³³

In his response to Gassendi, Descartes invoked “the common philosophical maxim that the essences of things are indivisible. An idea represents the essence of the thing, and if anything is added to it or subtracted from the essence, then the idea automatically becomes the idea of something else.”³⁴ To rebut Descartes's response, Gassendi turns against him another of his own responses. Recall the moral of the piece of wax story from Meditations 2. We know the piece of wax, not through the senses, nor through the imagination, but by the mind alone (mentis inspectio). Gassendi reads this to mean that we are supposed to know the wax itself, the substance of the wax or its essence, and objects that all that we know of it are its accidents, that in fact we cannot conceive of the wax apart from any extension, figure, and color.³⁵ Similarly, we do not know that the self is essentially thinking, only that it as a matter of fact thinks. In his reply, Descartes seems to give up the game: “I wanted to show how the substance of wax is revealed by means of its accidents. … I have never thought that anything more is required to reveal a substance than its various attributes, thus the more attributes of a given substance we know, the more perfectly we understand its nature. Now we can distinguish many different attributes in the wax. … And there are correspondingly many attributes in the mind.”³⁶ This seems to mean both that ideas can be enlarged upon, that the essences of things are not indivisible, and that our knowledge is limited to the accumulation of accidents or appearances. And Gassendi is quick to exploit this opening to full advantage.³⁷

The opening seized by Gassendi is only an apparent one, however, for Descartes's substance-attribute language cannot be understood in the Aristotelian terms Gassendi clearly assumes. Instead, individual things like the piece of wax are modes of extension in the sense of being instantiations of it, that is, ways that extension can exist (thus the French façon d'être as the translation of modus). And the better we know extension, the one essence of the material world, the better we know individually extended things like the piece of wax. The connection between extension (extensio or res extensa) and extended things (extensa) is the connection between the axioms and the theorems of geometry; it is the deductive connection that replaces the connection of inherence in the Aristotelian substance ontology.³⁸ Thus can Descartes claim in response to Gassendi's original objection that by coming to know God better we alter our idea of Him: “once the idea of the true God has been conceived, although we may detect additional perfections in him which we had not yet noticed, this does not mean that we have augmented the idea of God; we have simply made it more distinct and explicit, since, so long as we suppose that our original idea was a true one, it must have contained all these perfections.”³⁹ The process of making explicit what is only implicit in an idea is that of deduction, the possibility for which … qualifies the idea as both innate and distinct.

Gassendi just denies that there are such ideas:

As for you, when you say that “the idea represents the essence of a thing,” it seems that I may infer not incorrectly that if there are any things whose essence you do not know, you do not have an idea of them. Therefore I ask you: do you know the essence of the sun, of the moon, or of some other star? I suspect that you will not say that you do. For what do you know about them besides their size, shape, movement, distance, light, brightness, heat, their power to generate growth, to warm, to move, and other such things, if there are any. But the very essence, the nature, the inner substance which lies underneath these is totally hidden from you.⁴⁰

To be sure, there are ideas, but from the nominalist point of view they are of a very different sort.

Actually, you do have ideas of things, but not the kind you claim to have. In fact ideas of things exist only to the degree that we know them. And since we know their accidents with a distinct knowledge, but not their essences, which we divine as it were or conceive indistinctly as lurking under them, therefore, there is a distinct idea of their accidents, but not of their essences, which we comprehend indistinctly underneath them. From which it results that the clearer and more precise idea we have of something, the clearer and more precisely we know several of its accidents. Since experience shows that the ideas in our minds are like images of things, and the images are not of a thing's essence, but of its accidents, it follows that just as the image of a certain man reproduced in a picture is all the more perfect if the symmetry, the arrangement, and the representation of a great number of parts is more elaborately worked out, and each of the individual traits which are in the separate parts is more carefully reproduced, so the idea of any thing becomes all the more perfect if it portrays more of its accidents, or more of the things surrounding it, as it were, in a more ordered fashion, with greater skill, and more lifelike.⁴¹

GASSENDIST THEORIES OF SPACE: APOTHEOSIS AND ANNIHILATION

Gassendi's account of space and time begins with an attack on the received view that all being is divided into substance and accident.⁴² Although it is primarily Aristotle who receives Gassendi's attention, Pythagoras, Epicurus, Lucretius, and others⁴³ are noted as well, since Gassendi is concerned with a range of views based on the principle that all being is either in itself or in another.⁴⁴ With the additional premise that all being is either corporeal or incorporeal, space and time could plausibly be regarded on the Aristotelian view only as corporeal accidents, so that if there were no bodies there would be no space or time. Gassendi's contention is that even without body there would be unchanging place and flowing time,⁴⁵ and that therefore a category beyond substance and accident must be admitted.

Space (locus, i.e., unspecified place) is a “quantity or some sort of extension,”⁴⁶ an incorporeal tridimensionality: “the length, width, and depth of the walls of some water contained in a vase would be corporeal; but the length, width and depth of the walls of the vase if the water and every other body were excluded from it would be spatial.”⁴⁷ Note that the corporeal quantity is not the water itself, but its dimensions. One wonders how these corporeal dimensions differ from the incorporeal dimensions of the space it occupies. The distinction itself that Gassendi wants to draw is clear enough, and we shall turn to it at length below. What is unclear is the motivation for the distinction. A thing might be said to be extended either because it is in space, in which case corporeal extension is superfluous, or because it has corporeal extension, in which case its space is superfluous. Gassendi's text thus suggests two, more parsimonious views, each of which was later picked up and developed by his followers.

One view tended to emphasize space and time as conditions for existence—affections as Gassendi called them⁴⁸—and then to use some quality other than extension such as solidity or impenetrability to distinguish body from space. This is the active view of space⁴⁹ that dominates the Neo-Platonic tradition and to which Gassendi himself clearly is inclined. There are strategic reasons that make this route attractive to him, but quite independently of these he subscribes to a localization pattern. “There is no substance and no accident for which it is not appropriate to say that it exists somewhere, or in some place, and exists sometime, or at some moment.”⁵⁰ Cureau is an obvious proponent of this view, and Launay notably arrives at it. In order to secure divine immutability, argues Launay, God must be conceived as in space; besides which, “it is impossible for the human mind to conceive a being that exists and that is not in any place; for everything that is in itself necessarily is someplace.”⁵¹

The other view suggested by Gassendi's text tended to reject spatial dimensions as primitive and instead to derive them from corporeal dimensions. This represents the passive view of space that traces to Democritus.⁵² Bernier, for example, may have been led to it by considerations such as those just raised. Why must a body be in space in order to exist, he asked, when space is not its “productive cause” and is “of an entirely different nature”? If space can exist without body, why should body not be able to exist without space?⁵³ If a body has corporeal dimensions different from spatial dimensions, he may be arguing, there is nothing to preclude that a body should exist and yet not be in space. We shall return to both of these tendencies, and the views they embrace, below.

Meanwhile, because these Gassendist views on space are sometimes difficult to sort out, a roadmap through the two tendencies may be of use here before journeying through them. Both tendencies were initially driven by a theological problem that had already been raised by Gassendi himself, namely, that if space is uncreated and independent of God, as it is on his view, then God is not the author of all things. Each is inclined to its own solution. Launay's solution, on the one hand, was obvious. Roughly put, spatiality is a feature of God, not something different from, and rivaling Him. This view may be called the apotheosis of space. In this, Launay was preceded, if not influenced, by Cureau. With Bernier, on the other hand, the tendency was to minimize, and finally to eliminate altogether, the ontological status of space and thus any theological problem it posed. This view may be called—literally, as we shall see—the annihilation of space. An additional contribution of the proponents of either the apotheosis or the annihilation of space was to advance the dialectic beyond anything in Gassendi's text. They each did so in two ways: (1) by giving far greater emphasis to the importance of views on space, and (2) by applying the theological objection to the Cartesians' views on space. In this, both Launay and Bernier were preceded, if not influenced, by Charleton. Bernier's views in particular were anticipated by La Grange's novantique criticism of the Cartesian views of Rohault. In turn, Bernier was attacked in Régis's defense of those same Cartesian views. To conclude this section, finally, two reflections will be offered in an effort to establish the contest for the historical and philosophical importance of these views on space. It is important to begin sorting out who may have influenced whom on space as absolute or relative, and to indicate that the stakes concerned no less than the principle of sufficient reason itself. But first, Gassendi's view and his attempt to answer the theological objection to it must be investigated.

GASSENDI'S VIEW AND THE OBJECTION TO IT

To distinguish corporeal and incorporeal extension, Gassendi asks us to imagine that God has annihilated everything below the Aristotelian lunar sphere, the result of which would be the preservation of its original dimensions without any corporeality. For example, a point on the sphere would be a certain distance from the one opposite it, namely, the diameter of the sphere. He next imagines that God creates and then destroys an infinitely large world. The space that remains—in truth, actual space—would have three properties: (1) immensity—space is without limit, although the world occupies only a part of it; (2) immobility—the world or any part of it may be moved without the space it occupies; (3) incorporeality—the space occupied by an object offers no resistance to bodies penetrating it or abiding with it (corporeis penetranteis … compatienties). This way of expressing its incorporeality distinguishes space from incorporeal entities such as God, intelligences, and the human mind, the last of which, at least, is a “real and genuine substance with a real and genuine nature.”

With the characteristics he has given it, Gassendi's space is indeed liable to the theological objection that it is a thing uncreated and independent of God with the result that He is not the author of all things. To this Gassendi had three kinds of reply. One was to minimize the ontological status of space. As opposed to positive incorporeal things like minds, space cannot act or be acted upon and is characterized only by its penetrability. One wonders how Gassendi can have it both ways. Indeed, the tendency among those who like Cureau and Launay made space and time conditions for existence was to resolve the problem by making space and time properties of God.

A second reply from Gassendi was the argument that he meant by space nothing more than what was admitted as the imaginary spaces by the “majority of sacred doctors,” who are undeterred by the objection, “alleging that it is nothing positive, neither a substance nor an accident, under which heading all things created by God are subsumed.”⁵⁴ These spaces are called imaginary, he insists, not because they depend on the imagination as do chimeras, but because we imagine their dimensions as we do the dimensions of bodies falling under the senses. The way in which these imaginary spaces are apprehended is crucial to their status, but it is just this that Gassendi left unclear. Nor was it made any clearer by Bernier, who at first gave as the reason these spaces are called imaginary that we are unable to conceive of them according to their whole immensity.⁵⁵ Later he gave as the reason that we conceive their extension or dimensions in the fashion of corporeal dimensions.⁵⁶ But then he tried to reply to the Cartesian argument, to which we shall return below, that if there were nothing between the walls of an empty room they would touch.⁵⁷ Here, Bernier interpolates a remark not to be found in Gassendi, namely, that indeed there is nothing between them, that is, “nothing corporeal, nothing that falls under the senses.” So either Bernier contradicts himself, or imaginary space is somehow imaginable but not perceptible.⁵⁸ Nor had Charleton offered any improvement: “not that they are merely Phantastical, as Chimaera's; but that our imagination can and doth apprehend them to have Dimension, which hold an analogy to the Dimensions of Corporeal substances, that fall under the perception and commensuration of the sense.”⁵⁹ All of this sounds suspiciously like an important Cartesian view to which we shall also return, at great length, below. Descartes, Malebranche, and others held that we can perceive, not just extended things, but extension in general and that we can do so independently of any sense perception of it. As we shall see here, this is why Bernier came finally to reject the notion.

John Sergeant also rejected it, but for a different reason. He thought that the notion of real imaginary space was either incoherent or contradictory and that what its proponents were talking about existed only in the imagination. In his view they ought to have said that it is the imagination that is infinitely extended beyond the world; “but this is so notorious a Banger, that they say not this neither.”⁶⁰ That is, if imaginary space is real, the mind is extended. This is not the last of the bangers over the mind's location with respect to its object that will be encountered. The imagination does extend beyond the world in the same sense, according to some, that things beyond the world are in the mind. Indeed, … [Arnauld argues] that Malebranche makes both God and the individual mind materially extended in order to explain perception of things at a distance. We shall also then see that the question of perception at a distance distinguishes the theories of ideas of the gods (Malebranche) and giants (Locke).

In a third reply to the theological objection, Gassendi argues that space as he conceives it with its three characteristics poses less of a threat than those essences admitted as eternal, uncreated, and independent of God and that are the eminence (praecipuum) of substances and accidents. This is the problem in spades that Gassendi saw in allowing space as an element of things, for a doctrine of independent essences puts the intelligible component of things beyond divine control. As Charleton put it, “To hold [an essence] uncreat and independent, is obliquely to infer God to be no more than an Adopted Father to Nature, a titular Creator, and Author of only the material, grosser and unattractive part of the World.”⁶¹ Gassendi does not have this problem since he denies the distinction between essence and existence, as much for created things as for God, with the result that no essence is eternal except God's, which is eternal but not essentially so. Thus God according to Gassendi is freer and more powerful even than according to Descartes. For the latter, all truth, including eternal truths, even such as 3 + 2 = 5, depend on the divine will; but the Cartesian God is constrained by His own immutability. While He could have willed that 3 + 2 be other than 5, once having willed so He cannot will otherwise. Whether immutability is sufficient to the constraining task that Descartes sees for it is a question to be considered in chapter 4. Meanwhile, we can say that the Gassendist deity is under no such constraint.⁶²

APOTHEOSIS

The simplest way of resolving the problem of space as a competing divinity is the way taken by many of those for whom an eternal essence loomed as a competing divinity, namely, by making it an aspect of the single divinity. To this solution Gassendi himself showed more than passing partiality. The localization pattern found in Cureau, however problematic, certainly suggests it. Launay commits himself to the view, though not without a certain ambivalence. He follows Gassendi's characterization of place (lieu, locus) as an incorporeal tridimensionality allowing penetration by bodies, and then immediately describes it as nothing other than the virtual or eminent (eminentielles) parts of God's immensity.⁶³ He realizes that “to admit real spaces” [espaces] that preceded the world and that will follow it as an uncreated being if it is annihilated is a “difficulty” (inconvenient)—presumably the theological difficulty above—but he immediately replies to it that spaces so conceived accord quite well with God's immensity.⁶⁴ The way such spaces accord with divine immensity, it would seem, is that they just are the divine immensity, that is, space is actually a feature of God and the virtual or eminential parts of it that are distinguished by their capacity to receive different bodies are what he calls place. Here Launay uses an image that dates to a pseudo-Hermetic text of the twelfth century, the Book of Twenty-four Philosophers: God is a sphere whose center is everywhere and whose circumference is nowhere.⁶⁵

However unclear this conclusion might be throughout Launay's essay on space, none other is possible by the end. Launay repeats the argument that God could create innumerable worlds or enlarge this one and that, to place these worlds, ultramundane spaces must be admitted.⁶⁶ These spaces must be infinite, otherwise we would be unable to conceive or represent the divine immensity (p. 102). But Lauany slides from its conception or representation to immensity itself. We must be careful not to make God material, he says, but there is no danger in making God spiritually extended (p. 203). This gives us the significance of the problem in seeking whose solution Launay had “stirred all the dust in the colleges,” namely, how to make place immobile, as it must be in order to mark the beginning and end of locomotion; for otherwise a thing could change place without moving or move without changing place (p. 80 passim). His answer was divine immobility: Because God is by His nature incapable of change and hence immobile, space as a system of virtual places is immobile (pp. 80-81). Conversely, this conception of space shows how God is immobile and unchanging even if He should move the world or create others elsewhere—unlike “our adversaries” (unnamed) who are forced to say that God must move in these instances as the soul moves when the body does (p. 103).

At this point, however, it is not clear whether for Launay God is space or is in (all of) space. Despite his clear adherence to the localization pattern, Launay's conception of space effectively undermines the active, absolutist alternative, for in the final analysis Launay must deny that space is independent of the things it contains. As for Gassendi, so for Launay, not only are matter and space different, but corporeal and incorporeal dimensions are different. They are specifically different kinds of being, as he says: “Extension is the genus of occupied [plein] and void place,” the specific difference between them being, as Lucretius pointed out, tangibility (p. 97). Thus, Launay may be able consistently to maintain that time (or the virtual parts of God's eternity) endows things with their duration and continuation in existence, but not that space grounds their extension. If anything, corporeal extension is prior to incorporeal. As we have seen, Launay thinks that without an infinite incorporeal extension there can be no account of immensity; similarly, without it there would be no account of “the sphere of activity of an angel or a rational soul, which act only in determinate spaces because their power is limited. To make one thing coexist with another, to render it closely present to all the parts of an extended body without giving it extension is a thing inconceivable and impossible.”⁶⁷ Two lines of argument thus converge in the conception of immensity. God to act everywhere must be everywhere; and since He is everywhere He is immobile (there is nowhere else for Him to go) and thus the reference with respect to which other things move. But this is to give up the absolutist conception of space, not just in the trivial sense that space as a feature of God is no longer an independent being suo modo. For God is in space in the sense that He stands in all possible spatial relations. He is to the left of everything and to the right of everything, and He is space in the sense that with respect to those relations all other spatial relations can be defined. There is no space apart from things in spatial relations. On this reconstruction of Launay's position, then, he eliminates Gassendi's extra set of dimensions in favor of a relational theory of space. Indeed, … the only difference between it and Malebranche's relational theory is the difference between real and intelligible extension. This is the enormous difference, however, that places them on opposite sides of the grande bataille. However eclectic he may be, Launay does not for a moment flirt with the Cartesian theory that makes extension the essence of matter.

A striking feature of Launay's treatment of space is his unabashed application to the Cartesians of arguments that he takes from Gassendi and elaborates. For, in identifying space, matter, and extension and thus eliminating the void, the Cartesians hold a view that is “much bolder and more dangerous than that of the Peripatetics, who at least do not claim that the void is inconceivable,” which amounts to denying divine omnipotence. Affirming omnipotence gives Launay his principal argument, which takes two forms. First, God can destroy all that He has created (pp. 76-78); He can also destroy a part of what He has created, such as the contents of some container whose walls He keeps immobile. Ex hypothesi there will not be body between the walls of the container, yet there will be extension between them because they are immobile and would touch only if they moved. Even if as a result of God's act they did move, they could do so only over time, and since God could destroy the contents instantaneously, there would be at least an interval during which there would be extension without body. This form of Launay's argument of course does not really address the Cartesians, for whom the void is not just a physical, but a conceptual impossibility.

A second form of the argument is better in this regard since it involves or at least adumbrates the independence principle that was to figure so importantly in the vacuum-plenum debates. The principle, to be found already in Descartes's Principles,⁶⁸ is that what is conceivable apart is really distinct as an individual. As we shall see, just what is conceivable apart, with respect to space and its contents and to many other issues, was very much a matter of debate. For the present argument, at any rate, we are again asked to consider an emptied container, for example, a triangular room. Its inner surface “is distinguished only by the mind from the exterior (surface), with which it makes the same body,”⁶⁹ and must therefore have the same shape that the exterior surface has. But in order for the inner surface of the room to be triangular, there must be some extension between its sides. That is, the triangle requires the void in order to be preserved as the individual that it is. Perhaps another way we can put this issue for the Cartesians is that if God were to destroy a part of extension as Launay supposes, He would falsify geometry, for the perpendicular drawn from the base to the apex would not have the extension required by Euclid, but none at all. Thus, those Cartesians who save omnipotence in Launay's sense do so only by making geometry depend on the divine will, for if the altitude of a given triangle is equal to its area divided by half its base, this is only because God wills to create the space between its apex and base, that is, to create that particular triangle.

That a ground for geometry is the interesting issue here is suggested by a pair of arguments Launay raises that Bernier later picked up in the Abrégé. This time God is supposed to create three contiguous worlds in a row and then to annihilate the one in the middle while keeping the two others immobile. The Cartesians would have to say that though they are at a distance these worlds touch and moreover that God could re-create the world between them without moving them. Second, if God were to create a pile of spherical worlds, the Cartesians would have to say that they all touch each other not just at one point, but at all points in the way a pile of cubes would, which “seems to me so contrary to common sense and the demonstrations of mathematics” that the void must be admitted (pp. 101-2). Even these arguments do not get at the core conceptual issue, however, since they assume what for Descartes is unintelligible, namely, that God could create only three extended objects. … [For] Descartes there is only one really extended object—what he called res extensa. It is infinite, and cannot be conceived otherwise without doing violence to geometry. For Descartes, the globes Launay is talking about are, if anything, phenomenal entities that depend on res extensa for their essence. With respect to them the relevant question is whether they could be conceived without any extension at all or whether they must have an extension that satisfied just the Euclidean requirements that Launay cites. That is, the issue is whether things have essences (of a certain sort) as the Cartesians held, or are bare as the Gassendists held. As we shall see, here and in the next section, Bernier offered arguments that relate more obviously to this issue.

ANNIHILATION

Bernier's account of Gassendi's views on space is problematic in the extreme. There are three versions (the proto-Abrégé, Abrégé 1, Abrégé 2) and there are significant differences among them as to the material from the Syntagma that Bernier chooses to include or omit, the ordering and emphasis of the material, and the material of his own that he interpolates. Although it is impossible to spell out the historical details, it seems clear that Bernier was responding to Cartesian arguments on the topic whose general thrust can be made out. These arguments may well have had the additional result of changing Bernier's own views. Whatever its source, in at least one case the change in Bernier's own views altered the way in which he presented Gassendi's. Whether driven to it by the Cartesians or not, Bernier came to a view of space that was more Gassendist than Gassendi's own view.

The most obvious departure is the emphasis given the topic of space by its placement at the very outset of the proto-Abrégé and Abrégé 1. For Bernier space is of primary importance. “The first thing we must do in turning our eyes toward the universe is to conceive a vast and immense space, infinitely extended everywhere in length, depth and breadth, the field⁷⁰ of the Almighty's works and the general place of all that is or may be produced.”⁷¹ Gassendi's treatment of space, however, had occurred only in the second of the seven books that comprise the first of three parts of the physics, which occurs between the logic and the ethics; that is, it is rather buried away.⁷²

Indeed, Gassendi seemed not to have recognized the significance of his views on space. The Animadversiones initially plumps for the void on the basis of physical arguments concerning motion, rarefaction and condensation, and saturation, but connects it to an analysis of space only by way of a worry about the substance-accident dichotomy (pp. 169-77). The account of space, which is actually quite close to what later appeared in the Syntagma, figures only as a digression, as Gassendi calls it, from the discussion of time, which space is used to explain since it is a parallel concept (pp. 610-22). Indeed, the account of space is separated from the discussion of atoms and the void by over four hundred pages of text. It was Charleton, six years before the Syntagma, who picked out this material and coupled it with the earlier topics and other related material in the first book of his Physiologia. This chapter follows the material in the Animadversiones fairly closely, consisting of translation, paraphrase, and summary. Occasionally Gassendi does not quite get the argument right, as when he seems to think that it follows from the possibility of God's annihilating and then re-creating the world that the space in which He does so must be immense (p. 67). The crucial premise of the argument, made explicit by Gassendi but ignored by Charleton, is that the world God might create could be infinitely larger than the present one He might annihilate (p. 615). But generally, Charleton's espousal of the Gassendist cause, although ruinous in other respects, undeniably advanced it in the case of the analysis of space. For one thing, Charleton included Descartes in his rebuttal of an argument that Gassendi had attributed only to the Aristotelians, namely, that the sides of a container whose contents have been annihilated must touch since there is nothing between them.

The principal thrust of Charleton's argument for the void is that all the arguments against it fail. The argument in particular from Descartes, White, and the other Aristotelians fails because the substance-accident dichotomy is gratuitously restrictive. “When any Cholerick Bravo of the Stagirites Faction shall draw upon us with this Argument … we need no other buckler than to except Place and Time.”

Gassendi's arguments in the Syntagma on behalf of the void are aimed directly against the Aristotelians and, if at all, only indirectly against Descartes to the extent that the identification of matter and space, and thus the elimination of the void, relies on an ontology of substance and accident. Like Charleton, Bernier picks up these arguments and turns them specifically against the Cartesians, elaborating them beyond any warrant in Gassendi's text. The clearest example of this is the argument with which he closes the chapter on space: “I might add that those who are unwilling to recognize space in the way in which we allow it seem reduced to an extreme predicament, which is to allow a body of infinite extent that is perhaps eternal, independent and incapable of being destroyed—space, body and extension being the same thing in their view—which is no cause for concern on our view because on it space is neither substance, nor accident, nor anything capable of action or passion … and is but a pure capacity for receiving bodies. But I would be ashamed to pause further on this.”⁷³

Four years later, in 1678, Bernier's shame had vanished and, even if he does not name them as such, he directly attacks the Cartesians. Those who confuse space and body are reduced to the extreme predicaments of “allowing a corporeal substance that fills all possible spaces, or rather, which is itself space and which is consequently of infinite extension, and of maintaining (for fear of being obliged to allow any void) that God with all His power would be unable to destroy or annihilate the least part of that substance and that it is therefore independent of God.”⁷⁴ But even this he thinks gives them too much attention. By 1684, however, a separate section is devoted to the predicaments of these “moderns” as he calls them. Not only are they taxed with the above theological objection, but Gassendi's objection that the Aristotelians reject the void only by relying on an ontology of substance and accident is now directed against the Cartesians as a great predicament in physics. If God were to empty a room of its contents while preventing both anything else from entering and its walls from moving, they would say that since there was nothing between them the walls would touch.⁷⁵

An additional feature of Bernier's treatment of the above argument shows him to have been involved in the polemics of the period. In all three versions of the theory of space, he argues that the existence or the mode of existence of one body cannot “absolutely” depend on the existence of another body, that the shape of the room cannot depend on air or anything else that it may contain. The result is that God might begin by producing the air that the room will contain and then the room; but He might equally well produce the room first and then the air it contains.⁷⁶ This of course is the independence principle invoked by the Cartesians, among others, that we soon shall see Cordemoy to have used against them in just this way on behalf of atomism.

An important question is raised by Bernier's attack on the Cartesians. How is it that he thinks the Gassendist theory can avoid the difficulties he aims against the Cartesians? For although the Cartesian extension is corporeal, it does not satisfy the Gassendist condition for being something positive, namely, that it have the power of acting.⁷⁷ On the contrary, active matter of the Gassendist sort ought not to have been and was not intelligible to any Cartesian. Indeed, as I shall try to show, for Descartes and some of his followers, the matter identical to extension is incapable even of real motion. Even for those like Malebranche who thought matter capable of real motion, matter had no dynamical properties. It was this conception of matter that figured so prominently in arguments for occasionalism, for example. The most basic dynamical property of impenetrability was necessitated for the Gassendists because they distinguished matter from extension. Bernier's argument, therefore, that the Cartesians introduce a “positive nature capable of acting” because it cannot be destroyed by God⁷⁸ either is no objection at all or tells at least as well against the Gassendist theory. In addition, by appealing to the independence principle, Bernier threatens to make not just space independent but bodies as well. If the room and the air it contains can exist apart because they are individuals, then they can exist apart from God. To this the alternative seems to be Spinozism: They are not individuals because they cannot exist independently of God, of whom they must be a feature or a part. Why would Bernier subject Gassendi to the contagion of these issues?

At first Bernier seems to ignore, or at least minimize these difficulties. The proto-Abrégé, for example, does not raise the theological difficulty at all as far as Gassendist space is concerned. Instead, Bernier there advances two arguments that space need not be created. One is an argument from the pseudo-Archytas modified to read that everything must be created in some place, but that place cannot be created in some other place since this would open an infinite regress.⁷⁹ The other is the argument that if the earth were withdrawn from the place it occupies, there would be no need to create that space in order to replace the earth, and thus there was no need to create its space in order originally to create the earth. (Neither argument figures as such in Abrégé 1 or Abrégé 2.) In Abrégé 2, however, the objection is raised in no uncertain terms. Bernier at length applies Gassendi's worry with respect to eternal essences directly to the Cartesians. On behalf of Gassendi's theory, meanwhile, he is content merely to counter with the reply that space thus conceived is no more problematic than the sacred doctors' imaginary spaces.

A clue to the evolution of Bernier's reaction to this problem is his growing reservation over Gassendi's theory of space. Consider the characterization of the void as imaginary space, which differs markedly in Abrégé 1 and Abrégé 2. To be sure, both make the point that it is so characterized by the theologians not because it is chimerical, depending only on the imagination, but because we imagine its extension or incorporeal dimensions “after the fashion of corporeal dimensions.” But, following an ambiguity of emphasis in Gassendi's text, Abrégé 1 makes the point in order to emphasize the existence of this space lest having been distinguished from incorporeal substance it be thought of as nothing at all. Abrégé 2 makes the point in the course of an apology for Gassendist space vis-à-vis Omnipotence; all that is admitted is what the theologians meant by imaginary space, which was uncreated and independent of God and which they yet allowed as orthodox. The change seems explained by the threat expressed in Abrégé 1 to the reality of space, which by 1682 becomes the full-blown doubt that space seems properly to be “a pure nothing”: “only with difficulty could a being other than God be admitted which is eternal, immense, independent, indestructible … penetrable, and immobile, which are nonetheless the properties [that had been] attributed to space.”⁸⁰ What I am suggesting, then, is that Bernier felt free to attack the Cartesians on a point on which Gassendi's theory was no less vulnerable since he himself was hesitating about just that problematic aspect of it.

Indeed, Bernier's reservation becomes a penetrating nominalist critique of Gassendist space. In addition to the theological difficulty, it is objected that space (1) is incorporeal yet has parts; (2) is imperceptible hence should not be admitted except for very strong reasons; and (3) when imagined as empty is not imagined as a being, whether corporeal or incorporeal.⁸¹ The criticism is sharpened in Doubts 2, where such a space is described as the chimera of those who delight in deceiving themselves, and where the characterization of space as a being “in its own fashion” is ridiculed as both obscure and useless against the above objections. The root difficulty, clearly expressed in both sets of doubts, is that the Gassendist space is an abstract entity. To the argument that between the walls of an empty container there must be some distance Bernier now replies that if that distance is “a certain line, or spatial, invisible and incorporeal length which makes the walls distant from each other, it is a pure fiction”—to be distant they need only not touch.⁸² What is true of equality is true of distance; “they are abstract terms, which like all others of this sort, lead us to error if we conceive something abstract or separated from the concrete” (pp. 387-88).⁸³ For two things to be equal there need not be some “distinct entity which is the equality” to make them equal. They need only each have a certain size. And when one measures the distance between two things, there is no thing that is being measured any more than there is a capacity in an empty room that can receive things. Space in fact is nothing at all: “I maintain that [these allegedly infinite, etc.] spaces are not, do not exist, are not a being, are not a thing” (p. 386). Empty space is best referred to by a negative judgment (p. 385). For example, nothing is in this room, where ‘nothing’ functions as a “particle” (p. 392) or syncategorematic term, just as there need be no thing called “darkness” spread out in a room in order for it to be dark—there just need be no light in it.

It was not likely that this radically nominalized space, despite its drawing near a relational view, would be found congenial by the Cartesians, who in one way or another take a position of extreme realism with respect to space. Nor was it. Régis for one agreed that the void as a room from which God has removed the air should be conceived by the same kind of negative judgment by which the dark is conceived; but just as the dark is a lack of light in air capable of being illuminated, so the void is a lack of air in a room capable of having it, and since the void has some quantity it is some matter or other.⁸⁴ Régis goes on to argue against “others” (presumably, than Bernier) who nonetheless seem to hold his view of the void. They maintain that since space as a mere negation or privation would presuppose some subject, it might be viewed as pure nonbeing (pur néant). Régis's main argument that pure nonbeing has no properties, while the void has extension, fails against Bernier, for whom nothing has extension. There are things that are extended and qualified in other ways, but apart from which there is no extension or anything else. The distance between two things is not itself a thing but a property of the things distant that is not different from them. Thus, Régis just begs his realism against Bernier, just as he begs the question against Gassendi who rejects the substance ontology.

Bernier's conception of space as a pure nothing was anticipated by La Grange, who in his criticism of the Cartesians departed from Aristotle to allow not only the absolute but the natural possibility of the void.⁸⁵ In particular he is concerned with the views of Rohault, the natural sense of whose expression of them, according to La Grange, is that matter is infinite, uncreated, and independent of God. But he is also concerned with Descartes's view, which he sees as in effect coming to the same thing, namely, that a plurality of worlds is impossible because there is no place for another beyond the actual one (chap. 28). To combat both Rohault and Descartes, he thinks that either real, extraterrestrial space must be denied, which is not easy, or space though real (veritable) must not be something positive (rien de positif) or have real extension. The latter is supposed to become plausible when we realize that there is no space between things that touch, but that there is between those that do not because a third thing can be placed between them. “Thus space, properly speaking, is a certain capacity for receiving a body because there is no space in which a body cannot be put” (p. 403). So far this sounds like the Gassendist passive container. Indeed, the space now occupied by the world was the same before the world was created, and is neither substance nor accident. But it is not a being suo modo—“it is not a being at all. It is nothing [ce n'est rien]” (p. 404). The line he takes in explaining this view, however, is not at all as clear as Bernier's. Like Bernier, his inclination is to talk of space in terms of relations of distance. Thus he supposes that beyond the created world God could create a stone at a certain distance from it and then, without any real extension between them, alter that distance, and so on. But he gives distance a status Bernier clearly wants to deny it. Although space is nothing, “this nothing is real in its fashion, i.e. it is something that is such as it is imagined to be and thus is not imaginary.” (And therefore those who call it imaginary space are badly mistaken.) If we were to imagine real extension where there is none, he explains, it would be imaginary. But space is imagined as it is, that is, not as a being but nonetheless real. He sees two senses of the term ‘real’: real in the nature of things and real in representation. Space is real only in the latter sense, or can be as when we conceive space between two things that are not in contact. (By contrast, the space imagined between things that are in contact is imaginary.) Space thus conceived as nothing can be uncreated, infinite, indestructible, and immobile (pp. 407-8).

La Grange's rather Hobbist theory of space perhaps gives us a reading of another of Bernier's texts of obvious relevance to his account of space, namely, his chapter on place. The basic structure of the chapter remains unchanged throughout the three Abrégés. The problem Bernier deals with, as well as his eventual solution to it, is at least partly linguistic. Place must be immobile otherwise a thing might move without changing place or change place without moving. Thus the place of a thing cannot be the immediate surface of what surrounds it. This Aristotelian view has the additional inconveniences that a tower in the wind, for example, must move, and that the universe as a whole has no place. Place, then, is just occupied space, which when unoccupied is the void. Thus the basic structure of the three Abrégés. The tone and detail of the arguments, however, differ markedly. Abrégé 2, for example, compresses the material, treats it rather less enthusiastically, and suppresses two rebuttals appended to the earlier editions. One rebuttal replies to the Aristotelian argument that place is not a volume occupied by a body because, as the body itself has volume, two volumes would then occupy the same place. The other replies to the argument that space must be regarded as material since it is a divisible quantity. Both arguments are rebutted with the distinction between divisible corporeal and indivisible incorporeal extension. Once again, these changes in Abrégé 2 are explained by Bernier's Doubts.

In his Doubts Bernier is prepared to endorse the Aristotelian view of place as more plausible than Gassendi's view, at least to the extent that it reflects the commonsense conception of place and melds with his own nominalist conception of space. Place as the surface of a surrounding body seems clearly known, he thinks, by everyone, including children and even animals (a room is the place of a bed, a trunk the place of some gold, etc.). To the objection that such a view leaves the world without a place, Bernier with Aristotle grants the objection as harmless and then argues that the space in which we make distinctions beyond the world is only imaginary. It is a pure fiction that is regarded as real, seemingly in La Grange's sense, only because it is formed through constant experience of things that really are in place. The reality is only in the representation. Properly speaking, the void itself is (a place where there is) nothing. Thus neither this world nor some world God might create beyond it are contained in the void.⁸⁶

Part of Bernier's worry over place is merely linguistic insofar as the representation or description of motion as change of place is concerned. But part of it would seem to hinge on the deep ontological question of the status required by what is described. And it is here that Bernier's nominalism becomes obvious and perhaps obviously problematic. He tells us that it came to him while thinking about modes that most if not all of them are indefinable, being themselves what enter into definitions of things. What Bernier has to say suggests that modes are known only on the basis of acquaintance: (1) modes are clear and evident by themselves and one need only have eyes to see what they are (p. 408); (2) this is plausibly true of some (e.g., pain) but perhaps not all (e.g., action) of Bernier's examples (p. 407); (3) modes are primitives that cannot be defined without circularity or synonymy (p. 406). Consequently, motion and rest, qua modes of bodies, cannot be defined and are known perfectly and definitively in the fashion of Diogenes. The upshot is that the basic objection to the Aristotelian view of place, viz. that it allows a thing to move without changing place and to change place without moving, is overcome, for it is based on a definition of motion as “a successive application of a body to the parts of the bodies surrounding it” (pp. 401, 410). In fact, only to allow the possibility of motion as “the passage of a body from one place to another” did Gassendi, following the ancients, admit space—that strange entity now viewed by Bernier as virtually contradictory—“a being which [though] incorporeal has parts … which subsists in itself and is not a substance, a being which is everywhere and is nowhere” (pp. 410-11). Better, then, to accept the commonsense view of place and refuse all definitions of motion and all the paradoxes with them.

Bernier thus in a sense avoids difficulties at the linguistic level, but it is not clear just what else he has done. It is one thing to deny to space any independent ontological status. It is quite another to provide an account that without it nonetheless grounds everything that Bernier must regard as true of material things and their motion. After all, the void was admitted, not to provide a definition of motion in vacuo, as it were, but, for among other reasons, to avoid the difficulty of the plenum vis-à-vis motion. For Bernier to reject the void and regard motion as a primitive is, at least historically, an unusual procedure. For he not only regards the space in terms of which motion is conceived as merely imaginary, but he rejects every definition of it because “definitions explain the nature of the thing” defined. Those who define motion treat it as a thing that is passed from one moving thing to another, a mistake that he thinks lies at the root of their mistaken view about collision and conservation of motion and of the difficulties they see in projectile motion. Although he never quite puts it in these terms, Bernier's procedure is to move the problems from the level of things to the level of how we talk about them. The procedure may be fair enough, except that the move as effected by Bernier converts nominalism from an ontological view into a lexicographical one. Even if, as some have come to believe in recent years, the world is the language we use to describe it, there is still a difference between physical or metaphysical problems and merely linguistic ones. It is one thing to say there is nothing to talk about; it is another to refuse to talk about it.

FURTHER REFLECTIONS

I shall conclude this section with two further reflections. One has to do with the principle of sufficient reason, which in the end will restore the larger context for the philosophical significance of these Gassendist analyses of space. The other, with which I shall begin, concerns the question of historical influences. Toward the end of his chapter on space⁸⁷ Gassendi has an argument for the existence of space suo modo that may be important in sorting out who was influenced by whom on the question of space. The argument is intended to rebut the view that all being must be substance or accident and that space, since it is dependent on what is in it, must be an accident—a thing can change place as it can change color; hence place like color is an accident of it. Gassendi responds that though place can join and separate from (accedat, & abscedat) the thing located without the destruction of that thing, and hence may seem to be an accident of it, place in fact does not approach or recede (neque accedere, neque abscedere) but is immobile; it is the thing located that moves. That is, what remains through change in the case of motion is space, not the thing in it which ex hypothesi has changed place. Thus, once again it is the absolute impassivity of space that emerges as crucial in Gassendi's analysis. In this instance, however, either Gassendi just begs the question on behalf of an independent space or what he says is true of all qualities. When a thing changes color, it changes, not its color: The apple, not greenness, becomes red. Thus, he says that, if anything, substance should be attributed to space insofar as it is successively occupied by bodies (presumably as a thing successively can have different colors). But space cannot be a substance either, “which ordinarily is understood not only as what exists through itself, but also and especially as something corporeal and material, or what has the faculty of acting and abiding, which surely are incompatible with place.”

The proto-Abrégé and Abrégé 2 ignore this argument altogether. Abrégé 1 perhaps hints at it, arguing that place has properties not generally attributed to substances: immobility, incorporeality, and inability to act.⁸⁸ It is tempting to attribute Bernier's lack of attention to the argument to his own disenchantment with the conception of space as a separate being.⁸⁹ Whatever the explanation, Bernier cannot have been anyone's source for the argument. Launay has a version of the argument, but one that is even less successful than Gassendi's in motivating the rebuttal: “If our adversaries say that place [lieu] is an accident, because it can be or not be without its subject, viz. the thing placed.”⁹⁰ That is, Launay begins with a characterization of space according to which it is, if anything, a substance. Charleton, on the other hand, has a version very close to Gassendi's text that he uses “to authenticate this our Schism.”⁹¹ Whether it was from him, or more probably Gassendi himself, that Newton picked it up, is difficult to know. But it does emerge in his De gravitatione et aequipondio fluidorum.⁹² Having proposed to overthrow the Cartesian philosophy with respect to extension, Newton rejects the trichotomy that extension be substance, accident, or nothing at all. It is not substance because it is not absolute in itself and is not active, and since it is conceivable without body and would not perish with a body that God might annihilate, it is not an accident; indeed it thus “approaches more nearly to the nature of substance.” Using Gassendi's term (affectio) he calls it a disposition of all being.⁹³

Locke has a celebrated passage in which he too rejects the substance ontology, and which, occurring where it does in his account of space, suggests that his source(s) would lead him to the absolutist conception. Despite the suggestion, … it is Bernier whose views here shed most light on Locke. By contrast to Bernier's line on space, the absolutist view is for a commonsense empiricist like Locke plainly a priori unsatisfactory. It is weird (the soul of dwarfs in the bodies of giants),⁹⁴ ill-defined (the question whether God is in space), and extravagant not just in being unparsimonious but in admitting the kind of entity it does. If substance and accident are unintelligible, a fortiori are space and time as a third category.

Finally, the Gassendist analysis of space raises a question with respect to the principle of sufficient reason. If space is infinite and homogeneous, and if this world occupies only a part of that space,⁹⁵ then why does it occupy the part of space that it does rather than some other; why does God create where He does rather than elsewhere? This is a very old problem with an enormous and difficult literature,⁹⁶ to which Leibniz's argument against the Newton-Clarke version of Gassendi's view is only the best known contribution. There are several philosophically plausible ways an upholder of the view might deal with the problem. One of them, however, is not through an appeal to a distorted authority. But I am afraid this is what the Gassendist position comes to, if there is a Gassendist position at all.

On behalf of ultramundane space Gassendi quotes, and Bernier translates, two brief texts from Augustine's City of God that deal with the topic. An argument is insinuated to the effect that unless space is infinite, the divine substance is limited by its ubiquity; that is, because God is everywhere, His place must be an infinite extension or He is not infinite. Limited space limits God. The upshot is that the divine substance must be conceived as though it were extended and diffused throughout infinite space.⁹⁷ The qualification as though is included “lest we imagine that the divine substance is extended in the manner of bodies.” But the difference seems only to be that bodies are spatially (and temporally) limited. “As corporeal extension is said to be extended because it is not merely in a single point but is spread out through several parts of space, so the divine substance is held to be as it were extended [quasi extensa] because it exists not merely in a single place but in many, or rather in all places.”⁹⁸ In these texts, at least, God is not said to be space, but only to be in space.

The context for the Augustinian texts on which the above dialectic is based is a discussion of God as creator. The immediate problem is the creation of the world in time: Why did God create at one time rather than another? This question was raised theoretically by the Epicureans⁹⁹ to show that God did not create at all. The problem is over mutability in God and sufficient reason. That God should act at one moment rather than at another indistinguishable from it means either He acts because of some change in Him or for no reason at all. To avoid what seems to be the only alternative to the Epicurean impiety, namely, that the world is created by God but without a beginning in time, Augustine argues that the question is no more legitimate than the question why He creates here rather than elsewhere. “For if they imagine that there were infinite stretches of time before the world existed, an infinity in which they cannot conceive of God's being inactive, they will, on the same showing, imagine infinite stretches of space; and if anyone says that the omnipotent could not have been inoperative anywhere in that infinity, it will follow that they are compelled to share the Epicurean fantasy of innumerable worlds.”¹⁰⁰ Gassendi quotes and Bernier translates only the italicized material. The second text is also taken from its context: “Will they [who acknowledge that God is everywhere in His immaterial presence] say His substance is absent from those spaces beyond the world, that He is enclosed in the space of this world, which is so small in comparison with that infinity?” Augustine's point is that if God were in these infinite spaces, then the plurality of worlds would follow; hence He is not because as the very title of the chapter indicates, there is no space outside the world.¹⁰¹

Gassendi and Bernier do not attempt to legitimate infinite and homogeneous space by direct appeal to Augustine, because they themselves do not even raise his problem of sufficient reason. Their distorted texts are cited to show, independently of that problem, that Augustine found the notion of infinite ultramondane space intelligible, acceptable, or whatever. This is reflected in the various uses to which Bernier puts the texts. The proto-Abrégé follows Gassendi more or less closely as above. Abrégé 1 seems to use Augustine's texts less to argue that space is in fact infinite, or that beyond the occupied space of the world there is unoccupied space, than just to make the distinction between corporeal and incorporeal extension. It also ignores the problematic attempt to distinguish the senses in which bodies are in space and God is in space. Abrégé 2 merely reproduces the texts, introducing them with a remark about support for the view of the divine substance as spread out as it were in the imaginary spaces beyond the world, which is a view Bernier thought was held by Democritus, Epicurus, Lucretius, Nemesius, and the theologians.

What, then, to make of the distortion? Outright dissimulation is prima facie implausible, of course, and in any case unnecessary given the availability of acceptable ways of dealing with the problem. Just a mistaken reading is also improbable, but perhaps accounted for by the conceptual improbability, or difficulty at least, even in seeing a problem concerning sufficient reason. For the radical empiricist creation is a matter of sheer, unconstrained volition, or in nontheological terms, existence is fortuitous. Ultimately there is no reason why now and not earlier, why here and not there. That the sphere of what is not—of where God could have created but did not—should be infinite only serves to emphasize the ultimate lack of reason for the finite sphere of what is, of what He did create.¹⁰² At this point the contrast with the Cartesian position is sharp. Here too extension is infinite, but here it is also real, whether material as for Descartes or intelligible as Malebranche would have liked it to be. If extension were merely finite, God would be less than God. In Platonic terms the world would be less than full and God would have been jealous and less than good. In Cartesian terms He would be a deceiver and the Euclidean picture of the world would be false. Given that He is not a deceiver, the existence of extension, whether created or uncreated, is far from fortuitous. Thus the significance of the Cartesian identification of matter and extension, which was sufficient but hardly necessary for the mathematization of nature. For with this identification the question, why here and not there, does not arise; for matter is everywhere it could possibly be.

Notes

1. See letters to Mersenne, December 18, 1629, AT [Descartes, Oeuvres, ed. C. Adam and P. Tannery. 1st ed. 1897–1913; 2d ed., CNRS, Paris: Vrin, 1964–75], 1:97; January 1630, ibid., pp. 112-13; March 4, 1630, ibid., p. 127; May 6, 1630, ibid., p. 148; or even as late as December 1638, AT, 2:464-65.

2. AT, 3:388-89.

3. July 22, 1641; AT, 3:416.

4. To Colvius, April 23, 1643; AT, 3:646.

5. To Mersenne, February 23, 1643; AT, 3:633.

6. To Noel, December 14, 1646; AT, 4:585-86.

7. The episode is recounted in Bougerel, Pierre Gassendi, pp. 306-8. See Jones, Pierre Gassendi, who gives more details on these personal relations (pp. 66-69) and also gives the substance of Gassendi's criticisms of the Meditations (pp. 135-88).

8. For the outline of my account, I have drawn heavily on the work of O. R. Bloch, to whom my debt will be apparent. Bloch first dealt with the topic in “Gassendi critique de Descartes”; he later elaborated many of these themes in La philosophie de Gassendi: nominalisme, materialisme, et métaphysique.

9. CSM [Descartes, The Philosophical Writings, trans. J. Cottingham, R. Stoothoff, and D. Murdoch. Cambridge: Cambridge University Press, 1985], 2:268; AT, 9:198.

10. Richard A. Watson takes Foucher's argument to be one of two major reasons for “the downfall of Cartesianism.” See his book by that title.

11. CSM, 2:234; AT, 7:337-38.

12. New Essays, pp. 131-32.

13. Des vrayes et des fausses idees; Oeuvres, 38:199-200.

14. Bloch, “Gassendi critique,” p. 232.

15. CSM, 2:179; AT, 7:257.

16. Disquisitio Metaphysica; Opera, 3:297a.

17. CSM, 2:100; AT, 7:140.

18. Bloch, “Gassendi critique,” pp. 220-26.

19. Opera, 3:279a.

20. Ibid., 3:286a.

21. Discourse on Method; CSM, 1:122-25; AT, 2:22-28.

22. Opera, 3:286a.

23. Cf. Gassendi's syllogistic reconstruction of Descartes's reasoning in Meditations 1 (3:279a).

24. Opera, 3:279b; Selected Works, p. 166.

25. Opera, 3:280a; Selected Works, p. 167.

26. Bloch, “Gassendi critique,” p. 227.

27. Opera, 3:299b.

28. 2.1.16; 113.

29. Opera, 3:95-210.

30. 3:159a; Selected Works, p. 43. Cf. Locke: “All Things, that exist, being Particulars” (3.3.1; 409).

31. CSM, 2:35; AT, 7:51.

32. CSM, 2:212; AT, 7:304.

33. 2.23.33-35.

34. CSM, 2:255-56; AT, 7:371.

35. CSM, 2:189-90; AT, 7:271-72. See also CSM, 2:190-99; AT, 7:273: “When you go on to say that the perception of colour and hardness and so on is ‘not vision or touch but is purely mental scrutiny’, I accept this provided the mind is not taken to be really distinct from the imaginative faculty. You add that this scrutiny ‘can be imperfect and confused or perfect and distinct depending on how carefully we concentrate on what the wax consists in.’ But this does not show that the scrutiny made by the mind, when it examines the mysterious something that exists over and above all the forms, constitutes clear and distinct knowledge of the wax; it shows, rather, that such knowledge is constituted by the scrutiny made by the senses of all the possible accidents and changes which the wax is capable of taking on. From these we shall certainly be able to arrive at a conception and explanation of what we mean by the term ‘wax’; but the alleged naked, or rather hidden, substance is something that we can neither ourselves conceive nor explain to others.”

36. CSM, 2:248-49; AT, 7:359-60.

37. Opera, 3:352b-353a.

38. Bracken, “Problems of Substance.”

39. CSM, 2:256; AT, 7:371.

40. Opera, 3:352a; Selected Works, p. 222.

41. 3:353b; Selected Works, p. 223. Among those who have been sensitive to Locke's debt to Gassendi, François Duchesneau has produced perhaps the most extensive treatment of the question, but one that resulted in a conclusion precisely opposite to that drawn here. He emphasizes Gassendi's rejection of intellectual pictures, basing his case on texts from Bernier, Digby, and Gassendi's theory of vision. On this view, the perception of an idea is for Gassendi the mind's act of grasping the intelligible in the sensible given. I am rather persuaded by Bloch, who argues just the opposite from the model of Gassendi's theory of vision: “Vision is in no way the intuition of a cognitive content given in sensation, but a reconstruction by the mind of the reality of things from a content which is, not the [intelligible] translation of that reality [in the mind as per Duchesneau's interpretation according to the Aristotelian tradition], but the [causal] effect of it” (La philosophie, p. 20). Thus ideas are not intellectual, but they are yet pictures, which ceteris paribus are caused by what they picture. Of this difficult issue, much more below.

42. Syntagma; Opera, 1:179 ff; Selected Works, p. 383ff.

43. 1:180b.

44. 1:179b.

45. locum constantem, & Tempus decurrens (1:182a).

46. Quantitatem, extensionemve quandam (1:182a).

47. 1:182a-b; Selected Works, p. 385.

48. 1:179a.

49. Sambursky, “Place and space.”

50. 1:182a; Selected Works, p. 384. See Abrégé 1, vol. 1: “it is inconceivable that a substance should be and that it should not be in some place” (p. 21). See also p. 19: Space must be uncreated since nothing is created unless in a place, and “it would be ridiculous to say that place is created in another place since this would go on to infinity.”

51. Essais physiques, p. 103.

52. Sambursky, “Place and Space.”

53. Abrégé 2, 2:399-400.

54. 1:183b-184a; Selected Works, p. 389. Also, 1:189b. Brundell points out that Gassendi here clearly drew upon the Jesuit Conimbricenses (Pierre Gassendi, p. 66).

55. Proto-Abrégé, p. 7.

56. Abrégé 1, 1:13.

57. Principles 2, 18; CSM, 1:231; AT, 8:50.

58. Abrégé 1, 1:13-14. The remark is dropped from Abrégé 2.

59. Physiologia, p. 68.

60. Method, p. 42.

61. Physiologia, p. 69.

62. See Heyd, “Philosophy”; Osler, “Providence” for the significance of voluntarist theology in this period.

63. Essais physiques, p. 76.

64. Ibid., p. 79.

65. Launay actually says “immense being” (p. 102). Gassendi, Opera, 1:190b.

66. Essais physiques, p. 101.

67. In fact, the incorporeal extension of the soul is somewhat greater than the corporeal extension of the body that imprisons it, since the soul of a dwarf can fill the body of a giant.

68. HR [Descartes, The Philosophical Works, trans. E. S. Haldane and G. R. T. Ross. Cambridge: Cambridge University Press, 1968. Ist ed., 1911], 1:242-43; AT, 8:24.

69. Essais physiques, p. 75.

70. In the heraldic sense: la table d' attente.

71. Proto-Abrégé, p. 4: “The first thing a physicist must do … is … to consider this space as the general place of all that has been produced … and that God may draw from His omnipotence.” See Abrégé 1, 1:7-8; Abrégé 2, 2:1-2.

72. Bernier explains in Abrégé 1 that Gassendi thought that logic, the traditional first part of philosophy, was, if harmless, without great use: “if the eye sees, the ear hears … without any precepts, the understanding can reason well, seek the truth, find it and judge it without the aid of logic” (au lecteur, unpaginated). Abrégé 2 returns the logic to be outset of the work, although the treatment of space remains at the beginning of the physics.

73. Proto-Abrégé, p. 14.

74. Abrégé 1, 1:25.

75. Abrégé 2, 2:5, 10-11.

76. Proto-Abrégé, p. 13; Abrégé 1, 1:17-19; Abrégé 2, 2:12-13.

77. See Syntagma; Opera, 1:184b.

78. Abrégé 2, 2:10.

79. Archytas had argued that everything is in some place, hence because of the same regress, place must be nothing. See Sambursky, “Place and Space.”

80. Doubts 1, p. 25.

81. Ibid., pp. 25-27.

82. Doubts 2, p. 387.

83. The Cartesian Antoine LeGrand later argued that “that which is Nothing can never constitute the Distance of Bodies.” Nor could distance be “founded in the Bodies themselves,” for then relations of distance would never change (Entire Body, pt. 2, p. 2).

84. Systême, 1:286.

85. Les principes, p. 410.

86. Doubts 2, p. 398. With this conception of place Bernier can also answer the puzzle about Archytas's arrow: It would indeed go beyond the world but not into the void. With it he can also restore Archytas's argument to its original intent, which he does in effect arguing against Gassendi's view: How must everything be in a place when space, supposedly a real being, has no place?

87. Opera, 1:184a-185b. The text is taken verbatim from the Animadversiones, p. 614.

88. 1:21.

89. He also adds that (1) nothing can be its own place, because a body that moves does not carry its place with it, but leaves one and moves to another; (2) common sense tells us that place is different from the thing placed, and that since place is what receives bodies, it is both prior to and supposed by them.

90. Essais physiques, p. 79.

91. Physiologia, p. 66.

92. Which its editors date between 1664 and 1668 (p. 90).

93. Unpublished papers, pp. 99-100; trans. pp. 131-32. cf. Syntagma, Opera, 1:179a. The term does not appear in this connection in Animadversiones and therefore not in Charleton's account of space. I am thus inclined to agree with Westfall's contention that Newton here “drew his discussion of space and time directly from the Syntagma” (“Foundations,” pp. 172-73, n. 5).

94. Locke of course is interested in metempsychosis (see his notes on Bernier) but only for purposes of thought experiments. He does not say that the soul of Castor in fact is ever in the body of Pollux, but only asks what we would say if it were.

95. Opera, 1:182a.

96. It has been systematically treated and made remarkably tractable by Sorabji.

97. praeter eandem concipimus infinitatem quasi extensionis, (Opera, 1:191b); comme diffuse & d'étendue (proto-Abrégé, p. 7).

98. Ibid.

99. Cicero, De natura decorum 1.9.21.

100. P. 434, corrected.

101. De Civitate Dei, p. 638. Augustine's points are faithfully conveyed by André Martin, however (Philosophia, 3:95).

102. As Launay put it, except for contemplating and loving Himself, God is inactive in those times and places in which He does not create, and if what He creates does not have the full temporal and spatial extension it might have, this is because He is “free and independent and consequently master of His actions” (Essais physiques, p. 104). Curiously, Launay initially gets the point of the Augustinian texts right when he argues that there are not innumerably many worlds. But later he falls in with the Gassendist interpretation when he argues that God could create innumerably many worlds or enlarge this one as it pleased Him, and since He cannot create a world without a place to put it in, ultramundane spaces must be admitted (Essais physiques, pp. 16, 101). The latter text raises the problem of sufficient reason at least obliquely. God must be where He acts, according to Launay, thus must be beyond the world in order (to be able) to create there.

Works Cited

Arnauld, A. Des vrayes et des fausses idées, contre ce qu'enseigne l'auteur de la recherche de la vérité. 1683. In Oeuvres, vol. 38.

_____. Oeuvres. Paris, 1775-83.

Bloch, Olivier René. “Gassendi critique de Descartes.” Revue philosophique de la France et de l'Etranger 156 (1966): 217-36.

_____. La philosophie de Gassendi: Nominalisme, matérialisme, et métaphysique. The Hague: Martinus Nijhoff, 1971.

Bougerel, J. Vie de Pierre Gassendi. Paris, 1737.

Bracken, Harry M. “Some Problems of Substance among the Cartesians.” American Philosophical Quarterly 1, no. 2 (April 1964): 129-37.

Descartes, R. Oeuvres, ed. C. Adam and P. Tannery. 1st ed. 1897-1913; 2d ed., CNRS, Paris: Vrin, 1964-75.

_____. The Philosophical Writings, trans. J. Cottingham, R. Stoothoff, and D. Murdoch. Cambridge: Cambridge University Press, 1985.

Duchesneau, François. L'empirisme de Locke. The Hague: Martinus Nijhoff, 1973.

Gassendi, P. Opera. Lyon, 1658.

_____. The Selected Works, trans. C. Brush. New York: Johnson Reprint, 1972.

Jones, H. Pierre Gassendi, 1592-1655: An Intellectual Autobiography. Nieuwkoop: B. DeGraaf, 1981.

Leibniz, G. W. New Essays on Human Understanding, trans. P. Remnant and J. Bennett. Cambridge: Cambridge University Press, 1981.

Locke, J. An Essay Concerning Human Understanding, ed. Peter H. Nidditch. Oxford: Clarendon Press, 1975; reprinted with corrections, 1979.

Watson, R. A. The Downfall of Cartesianism. The Hague: Martinus Nijhoff, 1968.