philosophy is for everyone
and not just philosophers

philosophers should know lots
of things besides philosophy

Feature articles

PhiloSophos knowledge base

Philosophical Connections

Pathways to Philosophy programs

University of London BA

Pathways web sites

Philosophy lovers gallery

GVKlempner: complete videos

PhiloSophos home

Pathways to Philosophy

The Circularity of Defending Reason
in the Cartesian Circle

by Justin Robert Woods

Rene Descartes has been accused of making a circular sequence of reasoning in his Meditations on First Philosophy. The circularity is seen where Descartes appears to claim that knowledge of God requires knowledge of God. Although Descartes makes a courageous and expected attempt to escape circularity, his arguments leave the reader either denying a great portion of his philosophy (along with Descartes), or accepting that one must stop the method of doubting at his favourite tautology where clear and distinct perceptions guarantee truth because they are clear and distinct. However, Descartes is not alone in making circles out of reasoning. The most famous attack against his Ontological Argument is itself circular and neither logic nor Relativism can keep it in its undeserved position against the great founder of modern philosophy. Let us first look at the cause of so much complaint about Descartes' reasoning. We will start by introducing the Ontological Argument and then Descartes' own defence. We will conclude by exposing the problems of claims of circularity that Descartes himself has missed as well as the problem of defending reasoning in general.

Descartes claims at the beginning of the third Meditation, and after summarising his certainty that he exists because he is a thinking thing, that clear and distinct perception guarantees truth. He states that in order for him to be certain of his own existence, he must be able to know something of what is required for certainty. He recognises that his 'certain' claim of existence asserts a clear and distinct perception. Descartes reasons, therefore, that certainty of the truth of perceptions is reliant upon clarity and distinction and therefore, formulates the general principle '...whatever I perceive very clearly and distinctly is true.'[1]

The circularity of Descartes' reasoning is revealed by it being entailed by such clear and distinct perceptions that they are themselves reliant upon certain knowledge of God's existence for their truth, yet such knowledge itself requires true perceptions. This later claim occurs in the following passage:

'Now, however, I have perceived that God exists, and at the same time I have understood that everything else depends on him, and that he is no deceiver; and I have drawn the conclusion that everything which I clearly and distinctly perceive is of necessity true.'[2]

Here Descartes is stating that he perceives 'that God exists'. Knowledge of God's supreme goodness entails that he would not deceive Descartes into making clear and distinct perceptions that were not true. Therefore, because of this knowledge, which is only attained through the true perception that God exists, Descartes appears to conclude that all other clear and distinct perceptions are true. This shows that the premise, 'I have perceived that God exists' is reliant upon the conclusion, 'Everything which I clearly and distinctly perceive is true.'

According to Walton,[3] a sequence of reasoning is 'circular' if one of the premises is reliant upon the conclusion for its truth. This is indeed, then, a circular argument. God's existence must be known before we can know the truth of our perceptions, but in order to prove God exists we need to rely upon the truth of certain premises used to infer such existence, and such premises cannot be known to be true before we are certain that God exists. One aspect of this problem is identified by Arnauld in the Fourth Objections:

'I have one further worry, namely how the author avoids reasoning in a circle when he says that we are sure that what we clearly and distinctly perceive is true only because God exists. But we can be sure that God exists only because we clearly and distinctly perceive this. Hence, before we can be sure that God exists, we ought to be able to be sure that whatever we perceive clearly and evidently is true.'[4]

Arnauld points out that if we must prove that God exists before we can validate our clear and distinct perceptions, how is it that we can use such perceptions to prove that God exists? Before Descartes can prove that his perceptions are clear and distinct, he needs the guarantee that certain knowledge of God's existence would provide. Arnauld seems quite aware that circularity has been committed. He is referring to the passage that isolates the requirement of God's existence to validate such perceptions, where Descartes states:

'...I must examine whether there is a God, and, if there is, whether he can be a deceiver. For if I do not know this, it seems that I can never be quite certain about anything else.'[5]

The problem with this is how Descartes then goes on to establish the existence of God. 'This is, indeed, what he next proceeds to do, but by way of offering proofs.'[6] The problem with using the proofs that he does is that only certain knowledge of God's existence can guarantee the truth or certainty of anything, and therefore, '...we have no grounds for accepting the premises or validity of these proofs.'[7] For example, Descartes states that God is good and would not deceive us, and that God provides us with a faculty of correcting errors. Descartes is proposing that we know God first, in order to validate the proofs that would lead to knowing God through inference. God's existence is the proof of God's existence. The circularity here is quite evident.

Another accusation of circularity objects on the same lines as the one above, but refers to a particular 'proof' for God's existence: knowledge of the existence and essence of the self:

'You are not yet certain of the existence of God, and you say that you are not certain of anything, and cannot know anything clearly and distinctly until you have achieved clear and certain knowledge of the existence of God. It follows from this that you do not yet clearly and distinctly know that you are a thinking thing, since, on your own admission, that knowledge depends on the clear knowledge of an existing God; and this you have not yet proved in the passage where you draw the conclusion that you clearly know what you are.'[8]

Here, Mersenne in his 'Objections' refers to the circularity of the same two claims. On the one hand, he says, ' draw the conclusion that you clearly know what you are.' (i.e. a thinking thing), and on the other hand ' do not yet clearly know that you are a thinking thing.' Mersenne's objection could be put in the following form:

P1. One cannot know anything clearly and distinctly until one is certain of God's existence.

P2. In order to know God exists with certainty, one must be a thinking thing.

P3. To know with certainty that one is a thinking thing requires knowledge of God's existence.

C. Therefore, to know for certain that God exists, one must know that God exists.

The objection points out that all of the conclusions Descartes claims to have reached, such as knowledge of the existence and essence of the self as a thinking thing, the essence of matter, the essence of God, '...all depend on knowledge of a non-deceiving God.'[9] Yet, Descartes asserts that these conclusions are made prior to knowledge of that which they depend upon for their truth and certainty.

In reply to Arnauld, Descartes refers to part of his second objections. Here Descartes appears to make two separate attempts to avoid circularity. Firstly, he claims that when we are convinced of a perception being true, '...if this conviction is so firm that it is impossible for us ever to have any reason for doubting what we are convinced of, then there are no further questions for us to ask: we have everything that we could reasonably want.'[10] Such convictions are held to be as true as the most perfect certainty. Descartes adds that such convictions cannot come from the senses but only the intellect. He seems to say that some perceptions are held to be certain and true based on the strength of 'belief'. One example given is the Cogito argument, 'I am a thinking thing and therefore exist.' Beliefs are such that the mere thought of them entails certainty. 'Hence we cannot doubt them without at the same time believing they are true; that is, we can never doubt them.'[11] Here, Descartes attempts to break the circle by allowing some perceptions to be held certain without God's guarantee, because they could not possibly be doubted. The problem with this is that it sacrifices the claim that nothing is certain until knowledge of God's existence is proved. Clearly, some things are certain independent on whether or not God exists.

Secondly, he states that there are other very clear perceptions of the intellect, but they are perceived only by attending to the arguments '...on which our knowledge of them depends.'[12] Their certainty is assured only when we are in such attendance. When we are not attending to them, they can be doubted. It may be that we have a memory of the conclusion that was reached by such arguments, which are for the time being forgotten. As to the conviction of certainty we should attribute to such conclusions, Descartes refers us to the end of the fifth Meditation. Here, he states that, ' long as I remember that I clearly and distinctly perceived it, there are no counter-arguments which can be adduced to make me doubt it, but on the contrary I have true and certain knowledge of it.'[13] Descartes is making a distinction between what we perceive clearly and what we remember having perceived clearly. In order for something to be knowledge, it must at least have duration in the mind. This means that even when not concentrated on, it is still knowledge with the potential to be known. Some beliefs, on the other hand, can be certain without God's guarantee for their truth, but only for as long as they are concentrated on. Such beliefs are clear and distinct, and therefore true, when thought of and not reliant upon a non-deceiving God that would assure us of their truth. Their truth is ascertained by their being clear and distinct when concentrated on.

The problem with this is why should we accept the faculty of remembering as unflawed and the certainty of its memories? We may have been deceived by the possible Demon that our memories are true when they are false. Knowledge of the good, non-deceiving God would render such deception impossible, but again, such memories of clear and distinct perceptions are argued to be independent of God's existence. We must still ask, how do we know God exists and is a non-deceiver when we still have to rely upon the same disputable proofs for inferring his existence? Therefore, past memories of having perceived clearly and distinctly are themselves still possibly mistaken.

One interpretation[14] has it that when we establish, on the basis of clear and distinct ideas held before the mind at time t(1), that a non-deceiving God exists and therefore guarantees the truth of the memories of past perceptions, a thinker at time t(2) can recall to memory the ideas of t(1). Such memories can be relied upon for their truth without being attended to and with no reliance upon knowledge of God's existence at the time their truth is ascertained. But this whole argument still relies upon the certain knowledge of God's existence. At the beginning of the argument, we are asked to establish the existence of God based on clear and distinct perception. What guarantee do we have, excluding God, to rely on in order to establish such a perception? In short, why establish God's existence as a clear and distinct perception? Descartes tells us that we can be certain of this because it is clear and distinct. But, it could be that we are mistaken in our assessment that it is clear and distinct. Even if we cannot doubt it, it does not necessarily mean that it is true or that we believe it — a point that must hold great weight, since Descartes pays so much legitimacy to 'beliefs'. The circle, therefore, remains.

Descartes' attempts to avoid circularity are twofold. The first attempt breaks the circle by allowing some perceptions to be certain without God's guarantee, but this sacrifices the claim that nothing is certain until God's existence is established. The second attempt relies upon the memories of having perceived clearly and distinctly, but this begs the question, why establish that God exists based on the memory of having perceived this in the past, thus relying upon memories not having been tampered with by an evil demon, in the state of not relying on God's essence to prevent such occurring? Furthermore, ideas thought to be clear and distinct may not be such.

Therefore, Descartes has the ability to avoid circularity at some considerable sacrifice to his philosophy, with the first attempt, but there seems to be no hope in this with the second attempt. Furthermore, Descartes' method of absolute doubt seems to be selective as he does not doubt the legitimacy of logical reasoning, memory, thought itself, or that thought implies a thinker. So, a further circularity is entailed by claiming that these things can never be doubted, but we must employ them to doubt everything in order to discover that which is certain.

Now to look closely at the claims of circularity and the arguments a Relativist will mistakenly rely upon. Descartes places a lot of emphasis upon the power of reason in establishing knowledge of the world; so much so that he refuses to include it in his famous Method of Doubt. On what immense authority, then, does Descartes justify reason? Could Descartes use reason itself as such an authority? If we use reason itself to justify reason it seems clear enough that we are being circular in our reasoning. We are saying that in order to justify reason we can appeal to a method of justification, and that this method is reason. We are using the same unjustified method to justify itself. However, this last statement identifying 'circularity' is itself the result of reasoning. Is not such an accusation of circularity itself circular? Is it even possible, then, to find reasonable self-justification circular by also using reason? Failing the defence of reason by reasoning based on circularity is itself circular and as a type of 'reasoning' cannot therefore be made consistently. So, it is not the case that any attempt to defend reason must fail through circularity, when circularity requires reason itself for its identification. If reason fails to be a tool for holding certain beliefs, then rational inferences that identify circularity fail also to be held in such a condition. Circularity may well be identified in ways that reason plays no part; that is, it is not self-contradictory to admit such. However, the author can think of none here. The Cartesian Circle, then, is not a circle. In fact, until another method of justification can be discovered, no sequence of reasoning is circular where circularity is defended by reason.

However, leaving circularity of circularity aside, if we take the Relativist view we run into similar problems. To the Relativist, '...rationality[15] stands alongside and equal to any other means of forming beliefs, and alongside any sets of beliefs within which certain of these are held to be beyond rational critique.'[16]

Why on earth should the Relativist be concerned that certain of their beliefs be held beyond 'rational' critique? Is it that such avoidance justifies their beliefs? From the description above, the Relativist is seen to understand the reasoning of such things as 'categorising' and 'excluding' to construct definition: their beliefs are alongside, but not in the same category as 'reason', and some beliefs are excluded from the category of 'rational critique'.

Furthermore, one could argue that a 'rational critique' included the observance of the reasonable apparatus of language use. Surely, the Relativist must use language to express ideas and beliefs, otherwise how do we know that there are any? If the Relativist uses language, then surely they must abide by the rules of grammar, syntax, and such. The question is: what kind of beliefs do they hold in order that such rules essential to expression are abided by? The Relativist, and any other person who wishes to use language, must believe in the reasoning behind the structure of language if they wish to employ it. Moreover, the laws of logic are also rational. Is it possible, then, that the Relativist can avoid the law of the excluded middle, or the law of contradiction? How many sides does a Relativist believe a triangle has? Does the Relativist believe that they hold belief 'A' as well as belief 'not-A', or neither belief? The Relativist's beliefs are subject to such laws that govern reason and language, and therefore the content of the belief may be irrational, but the means of expressing it and the laws that govern its nature are indeed rational. Therefore, the Relativist is grounded in reason.

Descartes, on the other hand, was correct in claiming that reason needs no justification. However, his justification for this claim was based on clear and distinct perceptions. 'Reason, then, is what is made evident in those clear and distinct perceptions on the basis of which new beliefs can be formed from old ones.'[17]

For Descartes, the only justification required is that of our applications of reason, and not of reason itself, so that we may distinguish what only appears rational from what actually is rational. The authority by which this measured is the clear and distinct perception. But how on earth can we be certain that what appears to be a clear and distinct perception is such? Descartes leaves us having to accept the authority of clear and distinct perceptions as fundamental to our system knowledge and reasoning. Although, reason needs no justification, if we do not believe in this higher authority, then we must ask Descartes what justification there is for accepting it beyond belief. Descartes does not justify clear and distinct perceptions except as they are believed, therefore, perhaps the solution is to state that whatever is clearly believed is true. One thing is for sure, though: the Cartesian Circle argument is going to be believed but never justified.


Descartes, R., Meditations on First Philosophy (edited by J. Cottingham, Cambridge, 1996)

Honderich, T. (Editor), The Oxford Companion to Philosophy (Oxford, 1995)

Townsend, A. (editor), Descartes and the Defence of Reason (Melbourne, 1998)


1. Descartes, R., Meditations on First Philosophy (edited by J. Cottingham, Cambridge, 1996), p.24

2. Ibid, p.48

3. Walton, D. N., 'Circularity', in The Oxford Companion to Philosophy (edited by T. Honderich, Oxford, 1995),p.135

4. Descartes, R., Meditations on First Philosophy (Cambridge, 1996), p.106

5. Ibid, p.25

6. Townsend, A. (editor), Descartes and the Defence of Reason (Melbourne, 1998), p.163

7. Ibid, p.163

8. Descartes, R., Meditations on First Philosophy (Cambridge, 1996), p.102

9. Townsend, A. (editor), Descartes and the Defence of Reason (Melbourne, 1998), p.46

10. Descartes, R., Meditations on First Philosophy (Cambridge, 1996), p.104

11. Ibid, p.105

12. Ibid, p.105

13. Ibid, p.48

14. Townsend, A. (editor), Descartes and the Defence of Reason (Melbourne, 1998), p.164

15. 'Reason' and 'rationality' are used interchangeably in this text.

16. Townsend, A. (editor), Descartes and the Defence of Reason (Melbourne, 1998), p.162

17. Townsend, A. (editor), 'Descartes and the Defence of Reason (Melbourne, 1998), p.166

© Justin Robert Woods 2003