[T]here is an ambiguity in the word "idea". "Idea" can be taken materially, as an operation of the intellect, in which case it cannot be said to be more perfect than me. Alternatively, it can be taken objectively, as the thing represented by that operation; and this thing, even if it is not regarded as existing outside the intellect, can still, in virtue of its essence, be more perfect than myself. (8)(1)
This is a puzzling passage. It's pretty clear, from this passage, what an idea is in the material sense. But what on earth is this second kind of 'idea' that Descartes claims to identify, the idea in the objective sense? Apparently it is something which exists in the intellect, may or may not also exist outside the intellect, and can be in some manner (but only in virtue of its essence) more perfect than an actual thinking substance even if it doesn't exist outside the intellect. I daresay most readers will be hard pressed to think of any entity that fits this description and also could with any propriety be called an 'idea'. Let's call ideas in the objective sense (whatever they may be) "ideaso" and ideas in the material sense "ideasm". Descartes gives us to understand that the ideao is the thing (often an external, physical object) that we think about, and also that this thing exists in the mind in some sense: "[T]he idea of the sun is the sun itself existing in the intellect" (102) (though he assures us that this is only objective and not formal existence). What could he mean by this? I'm going to consider three alternative interpretations:
(a) That Descartes means that the sun - a very large, hot, luminous, physical object - literally exists in my mind.(2)
(b) That my ideao of the sun is a mental object of thought which represents the sun - a mediate object of awareness between me and the real sun.
(c) That to say something exists in the understanding (or intellect, or mind) is nothing more than to say that it is understood or thought of.
In like manner, we sometimes speak of things existing in the mind in ordinary language: e.g., you are 'in my mind' or 'in my thoughts' even when you are away; Santa Claus 'exists in the minds' of many young children; Mary is dead but she will 'live on in our memories'. Does this mean that Mary literally continues to live, albeit in some non-physical realm? Or does it rather mean that an idea or mental representation of Mary continues to exist in our minds? Or is it no more than a colorful and slightly misleading way of saying that, while Mary does not live on, we continue to remember her? These suggestions correspond to the above three interpretations of Descartes; and as it's pretty clear that the third is the correct one ("x exists in our memories" just means "we remember x"), one is tempted to think (even to hope) that Descartes' "objective existence in the intellect" can be interpreted along analogous lines.
This sets up my problem. I'm going to argue below that Descartes does not intend "objective being in the intellect" either in the sense of (b) or in the sense of (c). Amazingly enough, Descartes literally means that the sun exists in the mind when we think about the sun.
Here is how I understand representationalism as a theory of perception (the 'Theory of Ideas', as Reid calls it): According to this theory, in perception,
(1) we are never directly aware of external objects.
By "direct awareness" I mean that which is not based on some other knowledge. Thus, if perception furnishes us with knowledge of the external world, it must be by some process of inference.
(2) but we are directly aware of some mental phenomena (mental objects, states, or occurrences), which are called "ideas".
(3) the ideas in some way 'represent' external objects.
They might represent in virtue of resembling their objects. Or, at the least, our knowledge of ideas must function as both cause and justification of our beliefs about external objects and the features of our ideas correlate with the features we take the external objects to have.
This theory is what Thomas Reid took Descartes, Locke, and nearly every other philosopher to task over, claiming among other things that it leads inevitably to skepticism. And it is indeed not difficult to find indications of the Theory of Ideas in Descartes' Meditations. In the Third Meditation, Descartes tells us that some of his thoughts are "as it were the images of things" (meaning things outside him, as the examples make clear), and these are what he means by "ideas". (37) And when he then goes on to question the common sense beliefs he has hitherto held concerning these things, he makes it clear that these beliefs amount to that external objects transmit to him their likenesses (38) or "ideas or images of themselves" (40), and that consequently his ideas "resemble" the external objects that cause them (38, 39). This is the way Descartes interprets realism; this is how he phrases the trust in one's senses that everyone naturally has. And although in this passage he is rather questioning this belief in the reality of the external world than asserting it, it stands to reason that once, by the end of the Meditations, he has laid to rest all the doubts raised in the beginning, it is this form of realism that is again left holding the field. Indeed, if Descartes were a representationalist, it is hard to think how he might more directly and clearly have expressed his view; and it would be perverse to suppose that, if he were not a representationalist, Descartes should have spoken in so misleading a manner.
Furthermore - and this is perhaps the point of greatest moment - the overall thread of the Meditations does not seem to make much sense unless one assumes the representational theory of perception; but if one assumes that then it does make sense. Descartes claims to have set out to prove that knowledge of our own minds and of God is more certain and more easily had than knowledge of the physical world. (2,16) How does he do this? Roughly: In the beginning, he tries to call everything into doubt. Then he notices that his beliefs in his own (that is, his mind's) existence and his beliefs concerning his mental states can not be so doubted; they are certain. Then he demonstrates God's existence based on some characteristics of his idea of God; and finally, he derives the existence of external objects from the existence of God together with some facts about his ideas of external objects. Now, if Descartes did think that our awareness of bodies is indirect (inferential), and based on our awareness of some mental entities (I here make no distinction among mental states, events, and objects), then this procedure is entirely logical. If one accepts the representative theory of perception, then it is obvious that a problem of explaining how we can know about physical reality at all immediately arises, that some argument must be given, and that this argument must proceed from some features of our ideas. And if the argument Descartes comes up with is really how we know about external objects (so that atheists would not have such knowledge), then the conclusion Descartes claims as his primary result of import does indeed follow: knowledge of the mind and of God is more evident than knowledge of bodies.
If, on the other hand, Descartes did not accept the representative theory as described above, then his overall procedure in this work makes very little sense: For if we reject (1), then the motivation for seeking arguments for the existence of the external world disappears; moreover, it would be false to say that knowledge of the soul and of God is more evident than knowledge of the physical world. Although it might perhaps be said to be more certain (conceivably, our perceptual awareness might be direct but still uncertain), it could not be said to be more evident, or more easily known, than facts about corporeal bodies, since, if our awareness of bodies is direct, nothing could possibly be more easily known.
And if we reject (2), then it would make no sense to apply to our knowledge of our own minds for the evidence of the existence of external objects. Instead, whatever it is that we know directly (for there must be some such thing if there is to be any knowledge at all) is what we should look to for the premises from which to derive the existence of bodies. If either we do not know ideas directly, or ideas are not mental entities, then it would be wrong of Descartes to claim that he has shown how knowledge of the mind is more evident than knowledge of body. He has done so only in virtue of showing that knowledge of bodies depends on knowledge of our ideas (beginning with our idea of God).
Finally, if we reject (3), then though it might well remain that knowledge of the soul and of God is completely evident while knowledge of bodies is not, there just would not be any knowledge of bodies, since Descartes would then have no argument left for determining their properties. Thus, all three essential elements of the Theory of Ideas are required to make good Descartes' thesis that knowledge of the mind is more evident than knowledge of bodies: First, that knowledge of bodies is not completely evident; second, that knowledge of some mental phenomena is completely evident; and third, that the properties of bodies can be determined from the properties of these mental phenomena. And although it is not necessary here to suppose literal resemblance between ideas and physical objects, there certainly is still enough here of what Reid had to criticize in the 'Theory of Ideas'.
There are some fairly obvious problems with representationalism, however - problems with it first of all as a theory of perception and secondarily as an interpretation of Descartes. As Berkeley taught us, an idea can not possibly resemble anything other than another idea; and we might also add that a physical object can not resemble anything other than another physical object. Of course, Berkeley turns this point against physical objects, but we might rather turn it against ideas instead. To elaborate the point: Suppose there is an object that has a certain size, a shape, a color, a weight, a consistency, and so on; and there is another object which has no shape at all, no size, no color, no weight, no consistency (if you can imagine such a thing). Is there any sense at all in which such objects might properly be said to resemble each other? Plainly, these two objects are as far from resembling one another as two things could possibly be. Even if they had different shapes but they at least both had shapes, they would resemble each other more.
And this must be the nature of the case with respect to physical objects and ideas, if indeed both these kinds of entities exist. Berkeley takes it for granted that it is ideas which have such observable properties as size, shape, and so on, but let us assume, pace Berkeley, that it is physical objects which have these properties. At least as far as Descartes is concerned, it is bodies and only bodies that can have primary qualities (though we may have to leave out secondary qualities; perhaps nothing has secondary qualities; let's not get into that). A mental phenomenon of whatever kind can not have any of these qualities. So, if an idea is a mental phenomenon, it can not resemble a body in the least.
Furthermore, this reflection seriously calls into question the very existence of 'ideas' (in the sense of intermediary objects of perceptual awareness). For when I look at the sun, all I can find that I am aware of is a round, luminous, yellow disc. If this thing is not the 'idea' of the sun, and if my 'idea' of the sun is some other object that has no shape, color, or position whatsoever, then, I must confess, I do not know what this object can possibly be. I am sure that I at any rate have never seen any such thing. I do not deny that I have sometimes been aware of things having no physical properties, by introspection or a priori reasoning - but never in perception. In perception I have never been able to detect anything non-physical; and in vision I have never been able to observe anything non-extended. Again, I do not deny that my awareness (that is, my being aware) of the sun is distinct from the sun, and that I am aware of this awareness, in the sense that I know I am aware of the sun. But if there is supposed to be some third object, which I am aware of before I am aware of the sun, then I cannot guess what this object is or what it is like; and I think any one who reflects will find the same thing.
Considerations like these plague the representative theory of perception. Could Descartes have been blind to them? This would mean that he failed to appreciate his own mind/body distinction - that he forgot either that mental phenomena have no spatial properties, or that things with spatial properties are utterly different from things without spatial properties - only so could he have thought that ideas could be 'likenesses' of bodies. Furthermore, even if he did overlook these points, he must also have expressed himself very misleadingly when he wrote the replies to Caterus: "By this I mean that the idea of the sun is the sun itself existing in the intellect..." (102) If he were a representationalist, and he were being at all careful, he should rather have said that the idea of the sun is a representation of the sun that exists in the intellect, though certainly the sun itself cannot be in the intellect. So this provides some motivation for looking for another interpretation of Descartes.
Arnauld's reading of Descartes is that to "exist objectively in the intellect" just means nothing more than to be thought of (note that this makes it a slightly metaphorical expression). In this interpretation, when I think of the sun, there is the sun existing outside me, in physical space, and there is my act of thinking of the sun; there is no third object.
Now the first problem with this is already evident from the above. If this account applies to our awareness in general (particularly perception), then Descartes' procedure in the Meditations makes no sense. In order for him to claim that he has shown knowledge of the mind to be more certain and evident than knowledge of bodies, there must be some intermediary object of awareness between us and the bodies; and then it will follow that this object is known with greater certainty and obviousness than bodies. And so there must be a third object: for there is the real sun, there is the mediate object of awareness, and then there is the mental operation of being aware of it.
The only evidence I can find in favor of this reading is that Descartes does say that understanding something is sufficient for having an idea of it in your mind (160). This would certainly make sense if having an idea of x just meant understanding "x". However, it would also make sense if Descartes had a theory according to which having an idea in one's mind was the only way humans can have of understanding things. If Descartes thought, for example, that it was just obvious that you can only understand what x is if you have a representation of x in your mind, then it would make sense to say that "whenever I express something in words, and understand what I am saying, this very fact makes it certain that there is within me an idea of what is signified by the words in question". Likewise, if he thought that the only way of understanding something was literally having the thing enter your mind, he could say the same thing.
Furthermore, this interpretation makes Descartes' distinctions between ideasm and ideaso (8), and between the sun and the idea of the sun (102) superfluous and only misleading - for in Arnauld's view, there are only two things, the sun and my awareness thereof (whereas Descartes' two distinctions together result in three categories). My awareness of the sun is my ideam of the sun. So if my ideao is to be anything distinct, it must be the sun itself. But it is very odd to call the sun an idea.
Perhaps, then, my ideao of the sun is to be interpreted as the sun qua object of thought; that is, the sun insofar as I am aware of it. But on this reading, it seems that the ideao really is, just as Caterus said, the sun with some extraneous label applied to it - but Descartes assiduously distanced himself from this view.
We arrive at what most will feel the most ridiculous interpretation of Descartes' notion of objective being in the intellect - the literal one. Now there is initially something to be said for this approach, if it can be made intelligible - other things being equal, a philosopher ought to be assumed to have said just what he meant.
Surely, the reader wants to object, it can not make sense for a physical object to literally exist in my mind. But to this I answer: this only seems obvious to modern readers because we do not share Descartes' exotic metaphysics, in which things can have various grades of existence. We today usually think of existence as a black-and-white matter: a thing either simply exists, fully and actually, or does not exist at all. And of course if one thinks that, then the "existence of things in the mind" must either mean complete existence, in exactly the same manner as we think that the sun exists presently (but unicorns don't), or else must be somehow a metaphorical expression, a figure of speech calling for interpretation. And since we naturally think there's a difference between the ontological status of the sun and that of a unicorn, we think that a unicorn does not exist at all, whether we think about it or not. Furthermore, in its normal mode of actual existence, the sun has to be in physical space; so it can not exist in the mind.
But in Descartes' ontology, things are capable of having different grades of existence (165) (he considers this "completely self-evident" (185)). Further, he makes it clear that the way in which things exist in the intellect is one of the lower grades of existence:
Now this mode of being [= objective being in the intellect] is of course much less perfect than that possessed by things which exist outside the intellect; but, as I did explain, it is not therefore simply nothing. (103)
To motivate Descartes' view at least slightly, consider the status of universals: does the property of squareness exist? Well, in a sense one wants to say it does, for we don't want to say discussion of such abstract objects is sheer delusion or fantasy. But naturally one hesitates to say that universals exist in exactly the same way that the concrete physical objects around us do (this is what people find objectionable in Plato's Forms). Well, similarly, for Descartes, substances exist in a more full-blooded manner than mere modes do.
So the ideao of the sun is the sun, having this very attenuated kind of existence that things get to have in our minds. And perhaps, when a physical object exists in this non-actual mode, it does not require to be in physical space. This provides the answer to the main objection.
This interpretation explains a lot:
(1) Why does Descartes repeatedly say that the sun itself exists in the intellect, then adding that it does this only objectively, "i.e. in the way in which objects normally are in the intellect"? The least he could mean by the last qualifier ("in the way objects are normally there") is that he is not referring to some metaphorical kind of 'existence in the intellect'. He also isn't, though, referring to formal existence, which is the high-grade kind; rather, the way in which objects are normally in the intellect is in this attenuated manner.
(2) We can now escape the problems with representationalism - for if the idea of the sun is the sun, albeit in this low-grade existence, then it becomes possible that the idea of the sun should resemble the actual sun (i.e. the fully existing sun, outside the mind). As to how the mind can contain other objects, even in an attenuated form of being, this may just be one of the unique powers of mental substances as distinguished from other kinds of entities - that is, intentionality.
(3) How could Descartes think the Ontological Argument was valid? The traditional assessment of the Ontological Argument is that its mistake is to treat existence as a property. But how it does this is seldom made clear. Whether existence is a 'property' or not, might it not still be the case that the definition of God includes that he exists? Why not?
Well, here is what is wrong with it in my view: the argument claims that "God does not exist" is contradictory, for when the atheist makes such a statement, he is at once speaking of a most perfect (and hence, most real) being, and also saying that such a being has an imperfection. If God does not exist, it is said, then he would be less perfect than if he did. (65-6) The reason this is wrong is that if God does not exist, he is not therefore imperfect in any manner; for there simply is nothing there to ascribe any imperfections to. It is nonsensical to discuss what would be true of God on the supposition that there is no God.
But in Descartes' metaphysics, the ontological argument really would be valid: the essential step is made when it is established that God exists in the understanding. If this be granted, and the God existing in the understanding literally is God, literally existing, then there is something there to ascribe properties to whether or not the atheist's position is correct. That is, on Descartes' view, the atheist's view can not really be that God does not exist at all; the atheist must merely be saying that God exists only in the understanding but not outside the mind. Thus, God exists after a fashion, but his existence is of this low-grade kind. And if this is right, then even on the assumption that the atheist's position is true, it still makes sense to talk about the nature of God, and we could still intelligibly say that he was perfect or imperfect (and thereby really be talking about God).
(4) Why does Descartes think he needs an argument for the existence of external objects?
On the present view, what we know immediately in perception is our ideas of physical objects: that is, we know directly that physical objects exist tenuously, in the mind. What we need an argument for is the conclusion that these same things also exist outside the mind. So we can here make sense of Descartes' project of giving proofs of the existence of external objects.
1. Page references are to Descartes' Meditations (AT VII).
2. This is what Yolton refers to as "a wildly impossible view" (John Yolton, Perceptual Acquaintance (Minneapolis: University of Minnesota, 1984) p. 38).