Meditations on First Philosophy
Much of what Descartes sets out to prove in the Meditations is of little interest today, so I will deal only with those arguments that still have some relevance.
Mlle.X. “Mais, moi, M. Descartes, qui ne pense pas?” - Pound
Undoubtedly Descartes’ proof that he exists is the single most famous proof in the history of philosophy. One is tempted to muse, that’s all very fine for him, but what about the rest of us? We can’t be sure that he existed. In fact we can’t even know that we exist except whenever we doubt it or deny it or are unsure about it.
The stated goal of the Meditations is to find something, anything, that can be proved beyond doubt. Descartes begins the process by placing everything in doubt and dramatizes the procedure by depicting himself as the central character doing the doubting. One thing he doubts is whether anything at all exists and he responds to that doubt by his discovery that in the act of doubting at least he, the doubter, exists.
At first blush this proof looks pretty unassailable. But after a while doubts creep in. What exactly is this existence that can so firmly be attributed to the doubter? And what is this doubter about whom we can assert something so infallible? Is the proof completely free of logical error?
Existence: In order for Descartes’ Cogito to work existence has to be defined so narrowly that it has no meaning at all beyond something else that the doubter seems to be doing at the same time he is doubting.
In the First and Second Meditations Descartes gives three examples of what it means to exist.
The first is roundabout. It is actually a definition of what it means not to exist. If something does not exist it has only been imagined or dreamt and nothing more. So to exist means to not just be imagined or dreamt. Descartes calls this: being external to myself or outside of me. In the ordinary use of imagining or dreaming to be imagined or dreamt means that in some state different from the imagining or dreaming I will not encounter what I imagined or dreamt or it will not appear to me in the same way as it did when I was imagining or dreaming and I will be aware of a difference. Because he proposes that I may always be dreaming or always be deceived by a mauvais génie, Descartes does not mean imagining or dreaming in this way, for in that case there may not be a state different from imagining or dreaming and indeed such a state may be inconceivable. However, if the state of not dreaming is inconceivable then the distinction between existing and not existing has no meaning because in that case we cannot conceive or define what it means for a thing to exist in a non-dreamt sort of way. If the non-dreaming state is conceivable but, because of the mauvais génie, I just never enter into such a state, then there really is no difference between the things around me just existing-as-I-once-used-the-term and existing-in-a-dream. (Of course, a distinction would have to be made between the permanent state of dreaming and what we had once called dreaming, which would have to be called “dreaming within dreaming.”) The one is an exact duplicate of the other. (This appears to be one of Husserl’s points in his Cartesian Meditations.) However, there is still no clear definition of what it means to exist.
The second example comes from when I misapprehend one thing for something else as in the dummy I misapprehend for a man. In that case the man is really a dummy. The man does not exist because the dummy does exist. I have simply mistaken the one for the other.
The third example of what it means to exist is to occupy a particular space and point(s) in time, to be extended and to exclude other things from one’s location in space and time. Descartes cannot mean this in his argument because a thing can fill all these criteria of physical existence and still not exist if I am also dreaming the space and time it occupies.
When he actually proves that he exists Descartes adds a fourth example of what it means to exist. If I must by definition exist whenever I doubt my own existence, then existing must at least occur when doubting occurs. To exist at least partly means to doubt. Equally it at least partly means to conceive, to affirm, to deny, to want and not want, to imagine and to sense, for these are all variations of doubting or activities I also engage in or can engage in when I doubt. In the ball of wax example Descartes adds understanding to these mental phenomena for I cannot doubt the existence of a ball of wax or even conceive of a ball of wax unless I understand what that thing is through all the changes it undergoes before my senses.
The existing that Descartes has proved himself to be doing fits none of these examples except the fourth. It does not fit the first because Descartes could be dreaming and doubting at the same time. The second example is not clearly relevant to a totalizing doubt about whether anything at all exists. Descartes himself rejects the third as still subject to doubt.
But the concept of existing in the fourth sense adds no content to the concept of doubting. Existing means nothing except doubting or something you do while you are doubting. If it did Descartes would be able to provide a description of existing while not doubting. The bottom line is "existence" effectively means nothing; in the Cogito it is a meaningless term.
It is important to hold on to our understanding that the concept of existence in the Cogito is meaningless because, due no doubt to Descartes’ talent for exposition, it certainly feels like “existence” does mean something at the end of all this mental self-flagellation. There seems to be a sort of halo of existence surrounding the core of thinking while you’re sitting comfortably in front of your fire. This halo extends from the glow of the dying embers to a certain feeling you have even if you were a blind and deaf paraplegic doubting your existence. The feeling, however, is empty because existing says nothing about you that doubting has not already said.
You can reinforce your understanding that the concept of existence in the Cogito is a meaningless concept by reflecting that Descartes’ only real example of contrasted existence and non-existence relates to spatio-temporal existence. He elides subrepticement from spatio-temporal existence while he is doubting no more than un certain sourire when he concludes his proof.
The Ego: Descartes would have much more to say about the exact nature of what it is he has proved to exist in his treatment of mind as a non-extended substance and so the topic of the Ego deserves to be discussed under a separate head. At this point it is sufficient to observe that the Cogito has said no more about what has been proved to exist than that it doubts. Specifically Descartes has not proved that the doubter is a soul or a disembodied spirit or, depending on how one defines "mind," mind. Indeed it has not been proved that the doubter is not anything more than a material substance as Descartes would go on to define "material substance" in the Fifth Meditation. Now despite his numerous protestations to a strict adherence to clarity and rigor, Descartes slides swiftly into any number of assumptions, as one might call them, idées reçues, which add significant and unjustified content to the bare concept of a doubter. As early as the Fourth Meditation he speaks of his idea of a human mind as something which is not extended in height, length or depth, does not participate in anything belonging to the body and is distinct from any corporeal thing. None of this can be concluded in anyway from what has been proved about the doubter, certainly not as of the Fourth Meditation. Why can't the doubter be something corporeal? In fact why can't it be a doubting machine? Even in Descartes' day some enterprising clock maker probably could have built a machine that assured anyone who was willing to pay attention that it sincerely doubted everything and such a machine's pronouncements would exhibit no less conviction than Descartes' own.
Mersenne raises exactly this problem in the Second Objections to which Descartes replies by merely quoting his statements in the Sixth Meditation and repeating that the distinction between mind and body is clear and distinct and that therefore the mind must exist separately from the body. The only argument is a challenge to Mersenne to prove that the distinction is not clear and distinct. At a certain point Descartes' use of the "clear and distinct" excuse, able geometer though he was, begins to sound a but like American politicians pledging to protect "our freedoms" and appealing to "the inherent goodness of the American people." Certainly the fact that you can doubt that the body exists but presumably you cannot doubt that you are doubting is of itself not a clear and distinct distinction between mind and body. For Descartes has not proved that the doubter was not a body all along just like those anguished robots with perceptions and memories in Philip K. Dick. Although Descartes lacks a distinction of what "to exist" means other than "to be located somewhere in space and time," nevertheless his insistence on the existence of the mind or soul without the body gives an idea of what he would like to understand "The soul exists" to mean other than and no more than "Something is now doubting." For Descartes the touchdown would be, as he states from the very beginning of the Meditations and reiterates in the Second Replies, to prove that the soul would continue to exist after the death of the body. But without any discussion of what that means Descartes cannot reliably be counted on to even know where the end zone is. Does that mean that after our body dies we would find ourselves as if by teleportation sitting on a throne on a cloud somewhere? Are the movies right and material objects would pass right through us? Would we have some sort of disembodied awareness and perception of things without corporeal apperception and the capacity to cause changes in the material world, somewhat on the order of the disembodied camera in Being John Malkovich? It is as if after breaking the first tackle in proving that doubters must exist Descartes straightaway fumbles.
Hobbes and Gassendi ("...le noeud de la difficulté n'est pas de savoir si l'on existe, mais ce que l'on est" and "Vous montrez bien que vous êtes, mais non point ce que vous êtes." (Pléiade pp. 470 & 471.) also raise this objection though Hobbes' version is somewhat marred by a category mistake in equating a mind with a promenade (a mistake Descartes pounces on.) Arnauld (Pléiade, pp. 423 ff.) provides the clearest statement of the objection which every one of Descartes' interlocutors recognized. Arnauld's version has the merit of being a a purely logical question as to whether Descartes did in fact prove what he set out to prove. It is not burdened with the materialist metaphysics Hobbes and Gassendi implicitly defend.
Descartes' most tantalizing and potentially strongest argument that mind and body are really distinct comes in his reply to Arnauld (Pléiade, pp. 439 ff.). The first step of his argument is that a substance, a, is really distinct from another substance, b, if a complete and adequate description can be given of a without reference to b or to any property of b. The next step is to show that a complete description has been given of mind (something that thinks, doubts, wills, understands etc.) without reference to the body, for at the point in his demonstration where mind has been proved to exist a complete and adequate description has been given of mind even though his doubt whether the body even exists has not yet been lifted. This chain of reasoning sheds some light on why Descartes keeps insisting, much to the befuddlement of his interlocutors ("...personne n'ayant encore pu comprendre votre raisonnement..." (Pléiade p. 519)) , that he has a clear and distinct idea that mind and body are separate. It has a superficial sort of logical charm. It functions much like his proof of the existence of God. In fact one could call it Descartes' ontological proof of the existence of mind. It asserts a real distinction between two things if the definition of each thing does not contain terms that pertain to the other thing. And since, at this stage, Descartes believes he has already proved that mind exists, it must exist separately and distinctly from body. But this strongest version of Descartes argument is not a checkmate. It just rearranges the pieces on the board. Imagine an image on a TV screen. A complete and adequate description of the image of Tanya Danielle now appearing on the screen in front of me can be written down (a blonde model with shoulder length hair and awesome boobs) that does not reference the steel and glass chassis of the TV or the movement of electrons on the LCD. Yet the image of Tanya Danielle and the physical state of the screen at the time are in an important sense identical. Unless the screen is there in a certain state of excitation the image does not exist. One can say that the image is reducible to the extended object. In the same way the thoughts of the doubter that Descartes has proved to exist may very well be reducible to states of the doubter's body. It may very well be that the description of a thought, t, in Descartes' mind corresponds to a bodily state, s, such that for any unique tn there would be one and only one sn. Restated as reducibility rather than definitional distinction, the problem remains for Descartes that he still has not proved the distinction between mind and body, because when he proved that the doubter existed he may in fact have proved all along that a doubting body existed. This reduction may not be achievable, but it is conceivable and as long as it is conceivable Descartes has not met his burden of proof. The Sixth Objections encapsulates this point (Pléiade pp. 523-524) where it is observed, in the language of the time, that a distinction formed in the imagination does not imply a real distinction in the things themselves.
Descartes adds one more qualification to his understanding of the doubter in the Fourth Meditation to the effect that, since it doubts, it is somehow incomplete and dependent, thereby introducing - clearly and distinctly of course - the idea of an entity which is complete and independent, namely our old friend, God. Such sleight of hand is unworthy of the great geometrician. Surely if God is perfect he should be a perfect doubter as well.
One of the more entertaining proofs that mind and body are distinct comes in Descartes' battles with Gassendi, where in exasperation at what he views as Gassendi's caviling, Descartes observes that he very well could have been an elephant all along even while he doubted. The sad truth is that Gassendi was right. Descartes could have been an elephant all along.
Objection: Even assuming that sufficient content can be given to the notions of existence and of the self such that the conclusion of the Cogito can somehow be understood as meaningful and that this content does not impugn the validity of the proofs, nevertheless a severe independent objection can be raised against Descartes’ demonstration. This objection is independent of how or whether we understand “existence” or how or whether we understand “the self.” It is best stated by Descartes’ own interactive method:
I have determined not to accept anything as true that has not been proved. I now understand that I would be contradicting myself if I doubted all the things that appeared to me to be true and also doubted my own existence, for in the act of doubting I exist. Satisfied that I have secured a beachhead, however small, I jump into bed, jerk off and sleep the sleep of the philosophically secure. The next morning I rush to my armchair determined to advance my meditations from the point I had reached yesterday. Let’s see. I was doubting everything. Check. I doubted my own existence. Check. But by doubting I existed. Check. But wait a moment. Did I really exist yesterday? I am determined to admit that anything that has not been securely proved may be the work of the mauvais génie. If the image of the fireplace in front of me could simply be the nefarious illusion produced by the mauvais génie perhaps my memory of sitting in front of this very fireplace at a previous time is also a nefarious illusion. Perhaps these memories have been placed in my mind by some strange drug or by hypnotic suggestion? But didn’t I prove yesterday to my satisfaction that, under pain of self-contradiction, if I doubted I must exist? However, today seems to be a different story since I can now doubt whether I existed yesterday without contradicting myself. How can this be the case? What has changed? Well, yesterday I was doubting and I am doubting again today. The only difference is that yesterday I was doubting yesterday and today I am not doubting yesterday. Today I am doubting today. And today I cannot be certain that I really was doubting yesterday. My memory of my doubting, and consequently my memory of my existing, may also be the work of the mauvais génie. But I do know I exist now, for I am doubting now. My mind is all awhirl. I retire to clear my thoughts. Perhaps the solution will become evident after breakfast and a porn film. (Pause) OK, I’m back. Maybe this is the solution. How do I know I really existed before breakfast and hence whether I was really doubting? If I wasn’t really doubting before breakfast then maybe I really existed yesterday. Hold on. That doesn’t work. I am now doubting whether I was really doubting before breakfast. That means I am now doubting whether I really existed before breakfast. And I would not contradict myself to ascribe my memory of my preprandial existence to the mauvais génie. But that means I now have no valid proof that I existed either yesterday or before breakfast. Take a deep breath, man, you can solve this. Oh no! It is not a contradiction for me to assert now that I did not exist before I took my breath. Or before I wrote the end of this sentence. Or at the moment preceding my typing the period of this sentence. Or at the instant preceding the contact between my finger and the computer key for the final keystroke of this sentence! It appears that the validity of the Cogito has temporal limits. That proof works only at the instant that the doubter is doubting. Furthermore, that instant seems to have no greater extent than a point in space. We would in fact have to use the mathematics of limits just to give it the narrowest intellectual standing. For my money this is as close as you can get to the failure of Descartes’ Cogito. For today we may legitimately doubt whether Descartes really doubted anything in 1641.
Note 1: It is no fault that Descartes does not define what it means to doubt for, unlike Spinoza, he does not propose to proceed by definitions, axioms and deduction, and he does provide a pretty clear illustration of what it means to doubt by his own example. Any serious confusions regarding the nature of doubting would throw the whole of the Meditations into chaos and would raise some interesting questions about the nature of philosophical proof. Indeed anyone who doubts doubt will be doubting what he himself is doing and so may not be doubting at all (Disentangle this one. It sounds more like a Donne poem than Cartesian philosophy.) At the end of the Second Reply Descartes, protesting all the way, agrees to Mersenne's request to transcribe his proofs in axiomatic form (The only proof actually so transcribed is the proof of the existence of God). The result is a proof by stipulation since the Axioms already contain what he sets out to prove.
Note 2: With the Cogito Descartes reintroduced a unique form of proof into a Western philosophy enamored of logical deduction, a form of proof that had lain dormant for centuries. It is a peculiarly interactive form of proof that seems to thrive on if not require statement and response such as can be found in dialogues or, as in Descartes' case, meditations, where in effect he plays the role of each of two interlocutors. How insolite this type of proof really is can be demonstrated by rewriting the Cogito as a syllogism:
1) Whatever doubts exists.
2) Descartes doubts.
3) Therefore Descartes exists.
Or more accurately:
1a) Whatever doubts exists.
2a) I am now doubting.
3a) Therefore I exist.
As a syllogism, the Cogito is not a proof simply because one critical step is an assumption (the Minor Premise, (2) and (2a)). So it is not clearly and distinctly demonstrated, which it must be for the proof to be successful. In demonstrating the Minor Premise the aforementioned interactivity comes into play. For the proof of the Minor Premise comes in the form of a response to a challenge from the universal doubter to the effect that, in doubting everything, this doubter cannot deny that he is doubting. This sort of dialogue is the anti-pyrrhonist cousin to a dialogue that could occur with an epistemic nihilist who takes it upon himself to deny everything. The related response would be that in denying everything, the epistemic nihilist cannot nevertheless deny that he is denying. This sort of demonstration calls out for specific mental or verbal actions on the part of somebody. When those actions are rewritten as steps in a syllogism they turn into assumptions in need of proof.
This sort of proof first saw the light of day in those famous passages of the Platonic dialogues where Socrates demonstrates to various nubile young boys that, in the course of making one assertion, they really are proving its negative. In the hands of some philosophers proofs like these have been called dialectical after their origin in dialogue. I find the term "interactive" to be much less laden with terminological controversy.
Interactive proofs fell into eclipse among Popish theologians particularly following the embrace of Aristotle's logic which occurred around the 13th century or thereabouts. Notably Anselm's ontological proof and Descartes' revival of a version of that proof aspire to interactivity along the lines of the Platonic dialogues, for they begin with the challenge to conceive of a perfect (or greatest or whatever) entity. If what is conceived is then written down as a step in a syllogism, it shows a like characteristic, namely as an assumption in need of proof. It is a degraded form of interactive proof because one could very well be performing no more than empty mental horseplay in the attempt to conceive of a perfect being. Descartes' proof of the existence of God is for practical purposes indistinguishable from Anselm's (The only scholastic Descartes acknowledges reading in the Meditations is the Jesuit, Suarez, whose voluminous writings are said to contain a version of the ontological proof), but the Cogito, for which Descartes is justly celebrated, appears to have no precedent in Western philosophy. It is simply a brilliant and absolutely original application of this characteristic form of argument to a new topic without the unjustified assumptions about what can be conceived that ruin the ontological proof.
Nevertheless, in the two major proofs of the Meditations, i.e. the Cogito and the proof of the existence of God, Descartes' methodology is pretty much the same, that is it relies on interaction. It is distinct from logical deduction and, when successful, carries a high degree of conviction.
Descartes is quite clear that his method of proof cannot be rewritten in syllogistic form. He argues that the major premise is an assumption that requires proof, for we only accept universal statements such as the major premise as a result of generalizing from particular instances. The truth of statements regarding particular instances can be established only by a clear and distinct intuition (Pléiade pp 375-376). In another passage he calls his presentation as geometrical and says that by this he means that he will take care not to advance any proposition until he has made clear, and presumably proved, all the things on which that proposition depends: In his "démonstrations très exactes, je me suis vu obligé de suivre un ordre semblable à celui don’t se servent les géomètres, savoir est, d’avancer toutes les chose desquelles dépend la proposition que l’on cherche, avant que de n’en rien conclure." But this description is vague enough that we can take it to mean no more than a general order of conclusions. For example, he won't undertake to prove that material things exist before he proves that he himself exists. Finally (Pléiade pp. 387-388) Descartes distinguishes between analytic and synthetic argument. he characterizes syllogistic deduction as synthetic and his own method as analytic.
Note 3: Much of the rest of the Meditations is disappointing because it rests squarely on the proof of the existence of God. All this other stuff is true you see because God assures us it is true and God would certainly not be such a vaurien as to deceive us. The significant exception is the theory of extended and mental substances, not so much because Descartes manages to prove that there are two substances as from the inherent interest of the theory itself both in the avatars of that cunning fellow, the mind, and in the fact that the concept of extended and measurable substance would become the basis for ontological discussion in much subsequent natural philosophy.
Note 4: The infelicities in the Meditations do not impugn Descartes’ project of putting everything into doubt. They do show the confusions inherent in casting that doubt in terms of existence, viz. doubting whether things exist. [Next]