Descartes and Malebranche on Thought, Sensation and the Nature of the Mind

Antonia Lolordo. Journal of the History of Philosophy. Volume 43, Issue 4. October 2005.

One often-discussed aspect of Malebranche’s philosophy is his critique of Descartes, and in particular, of Descartes’s argument that the nature of the mind is better known than the nature of body. This critique appears, in more or less identical terms, in almost all the works of Malebranche which touch on the theory of mind: in the Search after Truth, its Elucidations, the Dialogues on Metaphysics and Religion, and the various Lettres in response to Arnauld. Malebranche’s critique is often thought to be highly effective: for instance, Nick Jolley calls it a “powerful internal critique of Descartes” and takes Malebranche to have “argued convincingly that Descartes’s famous thesis that the mind is better known than body is inconsistent with some of his central commitments concerning the nature of knowledge.” I find such claims rather surprising. On closer examination, Malebranche’s objections on this point are readily answerable from within the Cartesian system.

Thus, I shall argue, Malebranche’s objections to Descartes’s position are not particularly effective. Nor do they operate in isolation from more general features of Malebranche’s philosophy. In particular, they do not operate in isolation from his denial of the intelligibility of created powers, an aspect of his occasionalism; and from the denial of innatism embraced in the doctrine of the vision in God.

I begin by reminding the reader of what Descartes says about the nature of the mind. In section II, I give an account of Malebranche’s objections, together with some further illumination shed by Arnauld’s critique and Malebranche’s response to it. The general form of Malebranche’s objection is that if we had an idea of the mind or its nature, we would know a priori the range of possible modes of the mind; but we lack any such a priori knowledge; so we cannot have the idea of the mind which Descartes claims we have. The specific version he gives focuses on our ignorance of the range of possible sensory modes of the mind. In section III, I consider the general form of the objection, and argue that it fails as an internal critique of Cartesianism, for it relies both on the denial of innate ideas and on the denial of created powers. In section IV, I consider the specific form of the objection, which focuses on sensations, and argue that this fails too. For it is entirely irrelevant to Descartes’s point whether an idea of the mind’s nature engenders knowledge of possible sensations or not. Here, Malebranche has simply misunderstood the Cartesian account of sensations.

“On the Nature of the Human Mind, and That It Is Better Known Than Body.”

Two of Descartes’s theses are relevant here: that we know that the essence of the mind is thought; and that we know the mind, or its nature, better than we know body. I begin with the first thesis:

The extension of a body … may take on various different modes … But considered in itself, the extension itself-the subject of these modes-is not a mode of the corporeal substance, but an attribute which constitutes its nature and essence … [and similarly,] thought itself, as the internal principle [internum principium] from which these modes spring and in which they are present [ex quo modi … exurgunt et cui insunt], is not conceived as a mode, but as an attribute which constitutes the nature of a substance…

One crucial point is what this passage says about the relationship between substances and their natures or principal attributes, on one hand, and their modes on the other. Thinking and extended substances are the subjects of their modes; thought and extension are attributes constituting the essence of those substances, “internal principles” from which these modes spring. The notion that modes spring from principal attributes or natures involves both ontological dependence and something I will call causal dependence. Ontological dependence is straightforward. Modes cannot exist without the substance to which they belong, but a substance can exist without any particular mode or collection of modes. Causal dependence is a little more complex, so let me leave it until after we have seen Descartes’s second thesis.

Descartes has what are sometimes taken to be two distinct lines of argument for the claim that the nature of the mind is better known than the nature of body. The first starts from the claim that bodies are perceived by the intellect, not the senses, and proceeds to the conclusion that the mind is better perceived than anything else:

I know now that even bodies are not strictly perceived by the senses or by the faculty of imagination but by the intellect alone, and that this perception derives not from their being touched or seen but from their being understood; and in view of this I know plainly that I can achieve an easier and more evident perception of my own mind than of anything else (AT 7.34/CSM 2.22-23).

The second line of argument holds that whenever I know an attribute of a particular body, I thereby know a corresponding attribute of the mind, “which it has in virtue of knowing the attributes of a thing” (AT 7.360/CSM 2.249). This argument has struck many commentators, going as far back as Gassendi, as puzzling. Why does knowing more attributes of a thing imply knowing its nature better? If-as Descartes says elsewhere-we know substances through their natures rather than their attributes, why should it even help to know more attributes of mind? I shall suggest an answer below, in connection with Descartes’s account of innate ideas, but for the moment let us return to the notion of causal dependence.

This notion can be clarified by contrasting Descartes’s account of modes with a version of the textbook scholastic theory that he, along with the other so-called mechanists, rejects. Versions of that theory held that qualities, as well as substances, have essences. For example, opium causes sleep because it has, as a real quality, a dormative virtue-something which is essentially the power to dull the senses, which in turn tends to lead to sleep. It has traditionally been held that mechanists reject this sort of “dormative virtue” explanation because they view it as a pseudo-explanation. However, it is important to see what exactly is supposed to be wrong with the explanation. To explain opium’s ability to dull the senses by appeal to its possessing a real quality, the power to dull the senses, rules out any further explanation of the power in terms of the substance itself. The power stands on its own, as a primitive which is additional to and irreducible to the substance. Hence there is no answer to the question of why the power to dull the senses goes around with the other qualities and powers of opium. As far as this explanation goes, opium might just as well cause excitement and coffee dull the senses.

The intuition against the view of real qualities, then-an intuition which Descartes shares-is that behavior ought to be explained in terms internal to the substance itself, so we can see why this substance has the modes or range of modes it does. To respond to this intuition, Descartes avoids reifying qualities by reinterpreting them as modes or ways of being of substance. Thus his essences play a stronger theoretical role than the essences of certain scholastics. In particular, they are far more determinative and explanatory of the qualities and behaviors of the substance. This is what issues in the causal demand on the relationship between modes and substances-the demand that modes cannot have any causal powers independent of the powers of the substance itself (albeit modified in one way or another). And this is really just a way of making clear what is involved in the claim that all so-called qualities, accidents and powers are in fact merely modes of a substance. Thus Descartes’s account of the substance-mode relationship embodies both an ontological claim and a claim about the reducibility of powers.

Malebranche’s Objections

Malebranche distinguishes between knowledge by idea-which we have of bodies and their properties-and knowledge by consciousness or inner sensation. Each person has knowledge by consciousness of her own mind or soul (thus knowing that a thinking being exists) and of its modifications (thus knowing that she possesses certain modes of thought). Knowledge by consciousness cannot, however, reach to the nature of the mind.

Thus, since there is no idea of the mind or its nature available to us, we are precluded from any positive grasp of the mind’s nature, and know only some modifications of it. For although God has an idea of the human mind-an idea which is the exemplar or archetype of created minds-he does not choose to make it accessible to us. Knowledge through consciousness is sufficient to show that the soul is immortal, spiritual, and free. Moreover, we have indirect knowledge that the mind is immaterial. Since this is all the knowledge of the soul we need for practical purposes, God need not give us knowledge through idea of the soul. In fact, he chooses not to. For if we had such an idea, we might be “inclined … too much to view the soul as separated from the body,” which would diminish the strength of the mind-body union with unfortunate consequences for temporal life (Search; OCM 1.452/LO 239).

So much for the consequences of our lacking an idea of the mind or its nature. Why does Malebranche think we lack such an idea? Mainly because he thinks we are ignorant of a number of things which we would know if we did have an idea of the mind. He argues for this in the course of giving some objections to Descartes’s claim that the nature of the mind is better known than body. Indeed, it is important to recognize that although Malebranche’s explicit concern is to rebut the claim that mind is better known than body, his implicit attack is on the claim that we have any knowledge at all of the mind by way of idea. For not much in Descartes’s system relies on our having better knowledge of the mind than of body. If we simply had complete knowledge of both, this would be good enough for his purposes. Malebranche is well aware of this fact. His attack thus directs itself toward the claim that we have direct, introspective knowledge of the mind’s nature.

The attack consists of three main objections:

(1) The objection from indirect knowledge of incorporeality. Our knowledge of the mind’s incorporeality, Malebranche says, proceeds from the following reasoning. Modes of bodies are all reducible to figure and motion, which can be measured, so that bodies and their modes can be represented mathematically. However, we can form no precise mathematical idea of sensations, emotions and the like. Hence the modes of the mind cannot be modes of body, but rather modes of some other, unextended substance-the immaterial mind.

Now, Malebranche says, if we had a clear and distinct idea of the mind’s nature, then the idea of the mind would be represented to us directly. We would not need to go through this sort of reasoning. So it follows that we must lack a clear idea of the mind’s nature, and the mind cannot be better known than body (Dialogue 1; OCM 12.34-25/Jolley 7).

Descartes’s response, I take it, would simply be to deny the claim that we need an indirect procedure, like seeing that sensations are not measurable, to know that the mind is incorporeal. Rather, all we need to do is recognize that thought excludes extension-a recognition which simply requires consulting our ideas of the mind and of body. Neither objection nor reply is particularly compelling here.

(2) The objection that knowing more modes of the mind than of body does not imply knowing the nature of the mind better. Here Malebranche’s attack begins from the claim, discussed above, that whenever I cognize something other than my mind I thereby know that my mind can cognize that thing, so that I know more attributes of my own mind than of anything else. Descartes sometimes says that this claim shows that we know the nature of the mind better than the nature of body. It is this further claim which is Malebranche’s target. He objects that while I do indeed know by inner sensation more attributes of my mind than of anything else, this does not, contra Descartes, imply anything about my knowledge of the mind’s nature:

I can count that there are in my mind three properties: that of knowing that two times two is four, that of knowing that three times three is nine, and that of knowing that four times four is sixteen. If you wish, an infinity of properties in me can thus be counted, since these three properties are different from each other. But I deny that the nature of the things thus capable of being counted can be known dearly. (11th Elucidation; OCM 3.167/LO 636)

For knowledge of attributes does not as a general rule imply knowledge of essence. Instead, on Malebranche’s view, in order to know the nature of the mind-or even what is essential to its particular acts-we would have to know “what the soul’s dispositions consist in which make it readier to act and represent objects to itself.” In fact, however, “we cannot even conceive what such dispositions could consist in.” This is to be contrasted with the bodily case, where “we have no difficulty in seeing what constitutes the animal spirits’ readiness to be distributed in the nerves in which they have already flowed many times,” or “no difficulty in discovering that as the passages in the nerves become larger and their fibers relax in a certain way, the spirits can easily pass through them” (11th Elucidation; OCM 3.168/LO 637).

In other words, in bodily cases we can, after identifying a certain disposition to act, reduce that disposition to the non-dispositional mechanism which underlies it. This constitutes an explanation of the disposition and the behaviour it issues in-one proceeding only in terms of theoretically acceptable modifications of extension in motion. However, in the case of the mind, we have no reductive characterization of the relevant disposition or power. I come back to these worries about dispositions or powers in the next section. For it will become clear that this sort of worry lies behind Malebranche’s third, and initially most compelling, objection as well.

(3 ) Objections relying on sensations as examples. “We are able to say that we have a clear idea of the body because in order to know the modifications it can have, it suffices to consult the idea representing it,” Malebranche says. For instance, in knowing that the essence of body is extension, we know a priori that bodies can take on various geometrical shapes. However, we lack any such a priori knowledge of the range of possible modes of the mind:

… we have no idea of our mind which is such that, by consulting it, we can discover the modifications of which the mind is capable. If we had never felt pleasure or pain we could not know whether or not the soul could feel them. If a man had never eaten a melon, or seen red or blue, he would consult this alleged idea of the soul in vain and would never discover distinctly whether or not it was capable of these sensations or modifications… (11th Elucidation; OCM 3.164/LO 634)

Malebranche adds some further examples, and a comparison with the case of body:

If we had never sensed pain, heat, light, and such, we would be unable to know whether the soul was capable of sensing these things, because we do not know it through its idea. But if we saw in God the idea corresponding to our soul, we would at the same time know, or at least could know all the properties of which it is capable-as we know, or at least can know, all the properties of which extension is capable, because we know extension through its idea. (Search; OCM 1.451/ LO 237-38)

If we had a clear idea of the mind, we would be able to figure out, just on the basis of that idea, which modes the mind was capable of taking on. But we are unable to figure this out, as the examples of pleasure, pain, heat, light, colors, and tastes show. So we have no clear idea of the mind, and hence cannot know the mind as well as body.

Moreover, if we had a clear idea of the mind and knew its nature, we would know the relations between different modes of the mind. But, Malebranche says, we cannot discover the relations between the various modes of the mind, and hence we must not have a clear idea of it:

We can be said to have a clear idea of a being and to know its nature when we can compare it with others of which we also have a clear idea, or at least when we can compare the modifications of which the being is capable … We cannot discover clearly the relation between pleasure and pain… (11th Elucidation; OCM 3.167-68/ LO 636)

Nor, to cite another example, can we determine the relation between different colors (or different color sensations): we know that one shade of violet is darker than another, for instance, but we “do not know clearly either by how much, or in what being darker … consists” (Ibid.).

A third sensation-based objection begins with Malebranche arguing that we cannot even identify the modes of the mind directly, by consulting our grasp of the mind’s nature. Rather, in order to know what is a mode of the mind, we must also consult the idea of body. The Cartesians themselves, Malebranche claims, grant that in order to know that sensible qualities are modes of the mind, one must examine the nature of extended substance and thereby learn that extended things cannot be colored:

… even if one is actually feeling pain or seeing color, one cannot discover through simple perception whether these qualities belong to the soul. One imagines that pain is in the body that occasions it, and that color is spread out on the surface of objects … (11th Elucidation; OCM 3.164-65/1.0 634)

This indirect procedure would be unnecessary, Malebranche says, if we had a clear idea of the mind’s nature.

In response to this, Arnauld claims that Malebranche has simply misunderstood the Cartesian position. Descartes never argued that we know that colors are a mode of the mind by first knowing that purely extended things cannot be colored. There is simply no question about whether ideas of color are modes of the mind: all ideas are modes of the mind. The issue is whether there are colors in things as well as in the mind-whether, that is, colors are mere sensations or instead resemble some thing or quality in the world. Knowledge of extension comes into the picture only to rule out there being colors in bodies. It requires no reference to extension to know that colors are modes of the mind-only to know that they are materially false modes.

This initially seems like a natural response for Arnauld, who, like Descartes and Malebranche, rejects the scholastic view that colors are formally in things and objectively in the mind. Arnauld reads Malebranche as simply having misplaced where extension comes into the picture. We do not need to invoke extension to show that there are objective colors in the mind, Arnauld in effect says, but only to show that there are no formal colors in things.

But notice now that Arnauld’s reading has Malebranche thinking that Descartes is worried about whether colors are objectively in the mind-something which was not an issue for anyone, scholastic or not. That is, he reads Malebranche as focussing on the ontology of sensations, not the ontology of sensible qualities. Worse, Arnauld’s reading has Malebranche completely missing the real point of Descartes’s position, the denial of formal colors-that is, the denial that there is anything resembling or identical to objective colors in the material world. Thus Arnauld’s reading has Malebranche making a very basic mistake in understanding one of the aspects of Descartes’s programme with which Malebranche himself concurs. This is implausible, especially given that Malebranche repeats the objection in several places over several years.

Hence we should look for another reading. I suggest starting from Malebranche’s claim that we tend to imagine that pain is in the body that occasions it, just as we tend to imagine that colors are spread on the surface of bodies. I take it this means that the natural view is that our sensations have representational content which corresponds accurately to the way the world is. Now the Cartesians reject this natural view, holding instead that sensations are materially false. However, Malebranche can argue, the Cartesians cannot and do not claim that they discovered the material falsity of sensations just from considering the nature of the mind. They had better not claim this, for if it were apparent from the mind’s nature alone that sensations had no objective reality then it should not be so natural to think they do. Instead, the Cartesians had to discover the material falsity of sensation by first discovering that nothing with the formal character corresponding to their apparent objective reality could be intelligible in a body whose nature was extension. And Malebranche takes this to amount to a concession that we cannot fully understand what sensations are in terms of the essence thought, and thus a concession that the Cartesians have not really grasped the essence of the mind.

I shall argue that this objection fails as an internal critique of Descartes, since sensation cases are beside the point when we are talking about what flows from the mind’s nature. However, I take it that my reading makes the objection more comprehensible than Arnauld’s did. It also makes it more closely allied with the central thrust of Malebranche’s objection.

I have classified this-as well as the two preceding worries-as an objection relying on examples of sensations. But the question is whether any of these three objections works if rephrased in terms of purely intellectual ideas. Now, it is clear that the first two do not. The problem of the objective reality of secondary-quality sensations does not carry over to pure ideas, since there are shapes in bodies, as well as in the mind, on Descartes’s view. And we can, it seems plausible to claim, clearly determine the relations between different geometrical figures, or between substance-ideas and mode-ideas. One might think that Malebranche’s claim that we cannot, just by consulting the alleged idea of the mind, know whether the mind is capable of tasting a melon, seeing red, or the like, also carries over to purely intellectual examples. Perhaps, for instance, we would be incapable of telling whether the human mind could grasp the theorems of Euclid’s geometry if we had never had ideas of geometrical extension. Hence one might think that Malebranche’s most plausible objection remains. Simply having an idea of the mind’s nature tells us nothing about the range of possible thoughts or possible objects of thought. And-since no one involved here denies that knowing that bodies are extended tells us a great deal about the range of possible modifications of bodies-this constitutes an objection to the claim that the mind is better known than body (or indeed that it is known anywhere near as well). I address this worry in the following section, beginning, somewhat indirectly, with natures and powers.

Natures, Powers, and Dormative Virtues

In order to give a defense of Descartes, then, we need to see how having an idea of the mind’s nature, thought, provides us with knowledge of the possible range of modes of the mind. It need not tell us all the possible modes of the mind, but it must tell us enough of them to defend Descartes’s core claim that our idea of the mind is complete, that is, the claim that our idea of the mind suffices to let us know that the mind is a substance.

I argue that such a defense is in fact not far to seek. To see why Descartes need not take Malebranche’s objection seriously, we must simply pay attention to what the idea of the mind or its nature is supposed to consist in-in particular, pay attention to the connection between the idea of the mind’s nature and the doctrine of innate ideas.

I suggest reading the claim that the nature of the mind is thought as the claim that the mind is by nature intentional. In fact, since Descartes holds that there is no misrepresentation in idea (only in judgment), we can say, more strongly, that the claim is that the mind is essentially that which has cognitive grasp of the true. This is why Descartes says that the success of the cogito suggests “the general rule that whatever I perceive very clearly and distinctly is true” (3rd Meditation; AT 7.35/ CSM 2.2.4). It is not the fact that the cogito can be performed, on the basis of clear and distinct perception, which suggests the truth rule. Rather, it is the content-the idea of the mind-which we grasp when we perform the cogito which suggests to us that whatever we perceive clearly and distinctly is true.

The claim that the nature of the mind is to grasp the true might at first seem unhelpful. Indeed, Malebranche himself, who grants that we have indirect knowledge of the mind’s nature but no idea of it, could grant that cognition of reality is the essence of the mind. But how does knowing that the mind’s nature is cognition of the true help tell us about the possible range of modifications of that nature? My answer has two parts, the first of which is simply clarificatory.

It is well known that Descartes’s term ‘idea’ applies both to acts and to their objects or contents. An idea is both a content and the act by which attention is directed toward that content, i.e., both a cognitive object and an act which is individuated by that object. Now, since ideas are modes of the mind, it follows that (at least some) modes of the mind can be understood as both acts and as cognitive objects as well. Thus the question, how does having an idea of thought tell us about the possible range of modes of the mind?, can be rephrased as the question, how does having an idea of thought tell us about the possible objects of thought?.

This brings us to the second, more substantive, part of the answer. Once we grant that the question is about how grasp of the mind’s power to cognize reality is supposed to issue in grasp of particular bits of reality, then we can see that Descartes gives us the answer in the form of his doctrine of innate ideas, ideas built into thought itself. For instance, consider Descartes’s famous reply to Regius’s anti-innatist arguments:

When [Regius] says that the mind has no need of ideas … which are innate, while admitting that the mind has the power of thinking (presumably natural or innate), he is plainly saying the same thing as I, though verbally denying it. I have never … taken the view that the mind requires innate ideas which are something distinct from its own faculty of thinking. (Comments, CSM I 303/AT VIIIb 358)

This passage is generally cited in attempts to explicate Descartes’s account of innateness. But it tells us just as much about what Descartes takes the mind’s power of thinking, its nature, to be. Innate ideas are not distinct from the faculty of thinking. Rather, they are built into the mind’s basic power, not requiring any causal intervention by God, the human will, or external bodies or minds. Hence in knowing the nature of the mind and finding out that the mind contains various ideas true independently of the world (such as the idea of God or of a triangle), I thereby know some of the particular truths that it is natural for me to grasp. If the innate ideas are not distinct from the power to think, then in having a clear and distinct perception of the mind’s power to think I thereby also grasp at least some of the innate ideas, and hence grasp at least some of the modes of the mind.

At this point, we can begin to see what Malebranche thinks is wrong with this understanding of the mind’s power to think and the modes in which it is realized. Malebranche is hardly going to be willing to take seriously Descartes’s claim that the idea of the mind or its nature is the idea of a power, given his general hostility to created powers-a hostility which goes beyond the common early modern hostility to powers ajoined to substances. This point will become clearer by taking a brief look at Malebranche’s concerns about the doctrine of innate ideas, as given in the 10th Elucidation defense of the vision in God against innatism:

I am amazed that the Cartesian gentlemen who so rightly reject the general terms nature and faculty should so willingly employ them on this occasion. They criticize those who say that fire burns by its nature or that it changes certain bodies into glass by a natural faculty, and yet some of them do not hesitate to say that the human mind produces in itself the ideas of all things by its nature, because it has the faculty of thinking. But with all due respect, these terms are no more meaningful in their mouth than in the mouth of the Peripatetics. (10th Elucidation; OCM 3.434/LO 622)

(This passage, incidentally, makes it quite clear that Malebranche sees a close connection between the Cartesian doctrine of innate ideas and the Cartesian way of understanding the nature of the mind.) In order to see what is going on in this passage, we need to look at what the comparison of fire burning by its nature is doing here. The claim that fire burns by its nature is a paradigmatic formal causal claim. Now, we have a tendency to think of the concept of formal causation as one which disappears from use within mechanism, but this is not entirely true: it was still an ongoing topic of dispute at the time when Malebranche was writing the Search. Thus Malebranche’s objection here is, in effect, that to claim that the nature of the soul is thought is simply to lapse back into the unexplanatory language of Aristotelian science.

While Malebranche does not use the terminology of formal causation in this passage, that formal causation is at issue here is made clear by Arnauld’s response. In his polemic against the Search, Arnauld insists that “one must not ask … why mind is capable of thought, for it is the nature … of mind to think” (VFI 2; OA 35.185/Kremer 7). This is an example of his third rule of method, that one should not seek reasons ad infinitum. The fifth, closely related rule is not to confuse questions which should be answered by giving a formal cause with questions that should be answered by giving an efficient cause. For instance, “when we have arrived at the point of knowing the nature of a thing, there is nothing more to look for or ask for regarding the formal cause” (VFI 2; OA 35.183/Kremer 5-6). All we need look for now is the reason for the thing’s actual existence, namely, God’s creation. This clearly suggests that knowledge of the mind’s nature is knowledge of the formal cause of its range of operations, so that to go on asking why the mind thinks once we know its nature would be failing to see where explanation comes to an end. Once we see that the mind’s nature is to think, then we can understand particular thoughts as limitations of that form. This is how an explanation of the mind’s power to cognize reality in terms of its nature is supposed to go.

Arnauld and Malebranche thus seem to agree about what is supposed to be happening in Descartes’s system. Now, the comparison of the mind thinking by its nature to the fire burning by its nature may be rhetorically powerful, but it is merely rhetorically worrying for Descartes. For even if the comparison succeeds, it does not show us anything more about Descartes’s system than the applicability of some old-fashioned terminology he often avoids. There is an important ambiguity in the analogy between fire and the mind: is the power to burn the entire nature of fire, or merely a part of its nature or something adjunct to it? If the latter, then Descartes would reject the explanation-and the form of explanation it embodies-for the same reasons he rejects the dormative-virtue type explanations of real qualities discussed in section 1 above. However, Descartes could then also reject the comparison between the fire’s power and the mind’s power, so there would be no difficulty. But if the former-that is, if we take the fire’s power to burn and melt glass as supposed to constitute its entire essence-then Descartes can rule out the explanation on the grounds that it is incompatible with our knowledge that bodies are essentially extended and only extended, hence lacking activity. He need not-and in fact does not-reject the entire pattern of explanation, as his discussion of the scholastic idea of heaviness in the letter to Elisabeth of May 21, 1643 shows (CSMK 2.19/AT 3.667-68). He can merely say that traditional scholastic explanations go wrong because they apply a sort of explanation appropriate for the mind to the case of body, where it is inappropriate.

Malebranche’s tactic here is to assimilate Descartes’s account of the mind to the scholastic accounts of body which both he and Descartes ridicule, in order to show an internal tension within Descartes’s system. But this is not as easy as it seems, since the worry that Descartes himself has-the worry that motivates his rejection of accidents and qualities construed as anything more than modes-is a worry about powers adjunct to substances and not a worry about powers per se. Now, Malebranche must recognize this, and also recognize that thought is not a distinct faculty which the mind possesses but the whole nature of the mind. Thus he must in the end be objecting to positing any basic or unreduced powers, not merely to understanding these powers as accidents or real qualities. Indeed, in another context Malebranche denies primitive powers without qualification, on the grounds of their unintelligibility:

There are many reasons preventing me from attributing to secondary or natural causes a force, a power, an efficacy to produce anything. But the principal one is that this opinion does not even seem conceivable to me. … I cannot find in me any idea representing to me what might be the force or the power they attribute to creatures (15th ‘Elucidation; OCM 3.204-205/LO 658).

The connection between the denial of created, basic powers and occasionalism is immediate:

Some philosophers prefer to imagine a nature and certain faculties us the cause of the effects we call natural, than to render to God all the honor that is due His power (15th Elucidation; OCM 3.103/LO 657).

In effect what has happened here is that Malebranche has taken a basic thesis of mechanism, the denial of irreducible dispositions-a denial closely allied to the claim that matter is essentially inert-and tried to carry it over into the case of the mind. Malebranche cannot find a mental equivalent of impact laws any more than Descartes could. But while Descartes might conclude from this that causation in the realm of the mental works very differently from causation in the bodily realm, Malebranche seems to conclude instead that we have no idea of the mind. He takes our inability to explain mental causation in the terms developed for bodily causation as an illustration not of the fundamental difference between the two sorts of substance but as an illustration of our ignorance of the mind.

At this point, we have clearly moved quite far away from the “internal critique” of Cartesianism that Malebranche is sometimes thought to be offering. It is essential for the Cartesian version of mechanism to do away with unreduced powers and explain everything in terms of impact laws. But this does not carry over to the case of mental powers, and there is no clear reason why one should place demands from the science of body onto explanations of the mind if, like Descartes, one views the mind as an entirely different type of substance. There is no clear reason why a dualist should deny agency to the mind just because she denied it to body. So Malebranche’s objections seem to rely not on internal features of Cartesianism, but on his own suspicions about created powers. For once we grant Descartes that the mind is a power, he has a story to tell us in terms of that power which makes clear a number of possible modes of the mind. I have in mind here precisely those modes invoked in the a priori procedure of the Meditations: the ideas of God, of mathematical objects, of substance-mode ontology and the like. The procedure of the Meditations can thus be read as itself giving us an a priori psychology, starting from the idea of the mind’s nature, by giving us an a priori metaphysics.

The Sensation Cases

So much for the general form of Malebranche’s objection, that Descartes owes us an a prim psychology which he does not provide. One might still wonder about Malebranche’s examples of modes which are not intelligible in terms of the nature of thought. Nothing I have said so far helps us see, for instance, how we could know the relations between different color-sensations as a result of grasping the mind’s nature.

It is notable that all of Malebranche’s examples of modes which are not intelligible in terms of thought are examples of sensations. One might wonder why this is so. A tempting answer is that since Malebranche denies that ideas are modes of human minds, he does not want to posit ideas as modes even for dialectical purposes, leaving him with nothing but sensations as plausible candidates for modes of the mind. If this is so, then one would hope that Malebranche’s objection generalized to the case of intellectual ideas. Now, I have argued that it does not. Thus it is open to the Cartesian to respond that Malebranche has only shown that we cannot see how all the modes of the mind flow from its nature, leaving knowledge of a significant portion (all the innate ideas and what follows from them) intact. At this point, one might think that Malebranche has at best cast some doubt on whether we know the mind’s nature better than that of body-but that he has not achieved the more interesting goal of showing that we have no direct knowledge of the mind’s nature, or that the direct knowledge we do have is insufficient to show that the mind is a substance. Thus Malebranche would, as it were, have won the battle but lost the war.

I shall argue that the situation is not even this good. For the sensation cases themselves are not effective against Descartes either. Descartes can and should respond simply that the range of possible sensations is not supposed to be intelligible in terms of determinations of the mind’s nature. For on his account, I shall argue, sensations are properly referred not to the mind alone but to the “union and intermingling” of mind and body, so that we would need to understand the nature of this union in order to be expected to grasp the possible range of its sensations. Hence an inability to grasp apnorithe range of sensations of which the mind is capable does not affect the plausibility of Descartes’s case for our knowledge by idea of the mind’s nature, although it does show, as Descartes himself is willing to admit, that we lack intellectual grasp of the union of mind and body.

To see how this works, we need to look briefly at Descartes’s account of the union of mind and body. In the correspondence with Elisabeth, Descartes identifies three primitive notions-of thought, of extension, and of the mind-body union (CSMK 226/AT 3.691). Thought and extension, of course, are principal attributes of mind and body. But Descartes does not make clear here whether the mindbody union itself has a principal attribute, that is, is itself a substance. I wish to remain neutral here about whether the “real union” of mind and body constitutes a substance or not. All that is important for my purposes is that we have a notion of the union over and above our notions of mind and body.

Now, Descartes tells us-in the course of explaining that thought constitutes the nature of thinking substance-that “whatever we find in the mind is simply one of the various modes of thinking” and hence that “imagination, sensation and will are intelligible only in a thinking thing” (Principles 1.53; AT 8a.25). But later on in the same work, he spells this out further by saying that:

The mind is aware that [pain and other sensations] do not come from itself alone. And that they cannot belong to it simply in virtue of its being a thinking thing; instead, they can belong to it only in virtue of its beingjoined to something other than itself which is extended and moveable … namely what we call the human body. (Principles 2.2.; AT 83.41)

He immediately adds that “sensory perceptions are related exclusively to this combination of the human body and mind” (Principles 2.3; AT 83.41). Similarly, in the 6th Meditation Descartes tells us that a mind not closely united to a body would “not have confused sensations of hunger and thirst” but rather a purely intellectual understanding of the body’s needs, since sensations “must arise from the union and as it were intermingling of the mind with the body” (CSM 2.56/AT 7.81). These claims, taken together, make clear that Descartes holds that a mind not unified with a body could not have sensations. Sensations are modes of the mind only as the mind is unified with a particular body. If this is so, then it is clear why Descartes would not think that grasping the nature of the mind should by itself issue in understanding the range of possible sensations.

A number of questions arise immediately here. One line of questioning asks how Descartes could consistently both claim that sensations belong to the mind only as it is joined to a body and hold that sensations, being found in the mind, are simply a mode of thinking. Another line asks what it means to say that sensations belong to the union. The first line of questioning is most readily answered by distinguishing between two things Descartes might mean by ‘sensation’ or, perhaps better, two aspects of sensations. When I have a sensation of, say, heat, I am aware of something being presented to me, albeit confusedly. In this regard, a sensation is simply a thought, a mode of the mind. Indeed, there are confused intellectual awarenesses of content just as there are clear ones. Now, there is no difficulty in understanding how sensations, understood in this first way, can be modifications of thought and can be seen to follow from the nature of the mind, just like other presentations of content.

However, the sensation of heat is not simply a confused presentation. There is also a qualitative dimension to it, a feeling. Now, it seems likely that it is this feeling of heat which could not belong to a disembodied mind. While a pure intellect could have a confused thought just as easily as a clear one, it does not seem possible for that confused thought to be accompanied by a qualitative feeling. For recall that in the 6th Meditation Descartes tells us that sensations show that mind and body form a unit. A disembodied mind would not have sensations such as pain, but rather would be like a sailor in a ship and “would not feel pain when the body was hurt but would perceive the damage purely by the intellect” (CSM 2.50/ AT 7.81).

Thus the qualitative aspect of sensation is not explicated by knowing the nature of the mind; grasping the idea of thought does not tell us about the possible range of qualitative feelings. To this extent, there is something right in Malebranche’s critique. But we can now see that Descartes would never claim that grasping the idea of thought should explicate the possible range of sensation, since he holds that sensations must be referred to the union rather than to the mind alone. Indeed, Descartes further claims that the union cannot be grasped by the intellect but must be known through experience, so that the union cannot be understood purely a priori.

This claim is also crucial to resolving the second line of questioning, which asks what the union amounts to.’9 Although Descartes responds to this question quite explicitly in the Elisabeth correspondence, his response is generally found to be rather frustrating. He tells Elisabeth that the soul is conceived by “pure understanding,” while “what belongs to the union of the soul and the body is known only obscurely by the intellect alone,” but “known very clearly by the senses” (CSMK 2Z7/AT 3.691). The import of this claim becomes clearer if we consider how those things pertaining to body are known. For it is evident from the 2nd Meditation, as well as from this passage itself, that body itself is also known clearly and distinctly by the understanding. Thus the union, which is known clearly only through the senses, is a special case not open to the sort of philosophical understanding Descartes typically recommends. This is reinforced by Descartes’s subsequent claim that “metaphysical thoughts, which exercise the pure understanding, help to familiarize us with the notion of the soul,” but that it is only when such thoughts are dispelled by “the ordinary course of life and conversation, and by abstention from meditation” that we are taught “how to conceive the union of the soul and the body” (CSMK 227/AT 3.692). In other words, Descartes holds that the mind-body union is not something which we are capable of having intellectual grasp of, although we experience its existence throughout daily life. Unlike the mind alone, the mind-body union is not open to metaphysical understanding.

Now, I do not wish to claim that this rather oblique response to questions about the nature of the mind-body union should satisfy everyone. One might well think that Descartes has no right to take refuge in claims about the intellectual incomprehensibility of the union and thus avoid giving a straight answer as to its ontology. There remain genuine problems understanding how there can be a “real union” between mind and body which has qualities that are not simply qualities of either substance alone.

However, these are not the problems with which Malebranche’s objection concerns itself. The problem Malebranche poses is about the mind, not the union: he objects that, because we fail to grasp the modes of the mind a priori, we cannot understand the mind’s nature. And that worry does have a relatively clear answer for Descartes, one which lies in understanding what is included in the mind’s nature and what the appropriate modes to consider are. Although my consideration of Malebranche’s worry has brought us to the problematic doctrine of the union of mind and body, his worry is not about that union. Indeed, Malebranche seems not even to recognize that there is for Descartes a mind-body union relevant to understanding sensation. And thus Malebranche’s critique does not succeed in pointing out a Cartesian failure to give us the a priori psychology he owes us-although it may, somewhat indirectly, give us further reasons for worrying about the empirical aspects of Cartesian psychology.


I have presented Malebranche’s chief objection to Descartes’s claim that the mind is better known than the body. This objection is based on his denying that we have an idea of the mind-a denial which is supposed to follow from our alleged ignorance of the possible range of modes of the mind in general, and of the sensory ones in particular. Both the general and the specific form of the objection have relatively clear and plausible answers within Descartes’s system. Once we have a fuller understanding of what is involved in Descartes’s idea of the nature of the mind, we see that this idea does engender knowledge of a wide range of purely intellectual modes, via the innate ideas. One cannot, then, give up on the claim that we have an idea of the mind without giving up on innatism altogether at the same time. Of course, Malebranche may deny that any such idea is intelligible. But in the end this denial seems to rest solely on a refusal to admit created powers into his ontology. Moreover, the sensation cases which are so important to Malebranche’s argument are simply irrelevant as internal criticisms of Descartes. Malebranche is just mistaken in suggesting that having an idea of the mind must, for Descartes, imply having knowledge of the possible range of sensations. For Descartes, sensations-understood as qualitative feelings-are not limitations of the mind’s power to cognize reality, but belong instead to the union of mind and body, something which is not itself fully graspable by the intellect and hence not supposed to be part of Cartesian a priori psychology. Thus the plausibility of Malebranche’s objections depends both on certain features of his occasionalism and on a misunderstanding of the status of sensations within Descartes’s system. It is simply a mistake to take Malebranche’s objections as providing an “internal critique” of Cartesianism, let alone a particularly damaging one.