Descartes Colloquium, Brazilian Academy of Philosophy
Faculdade da Cidade, Rio de Janeiro, May 9th 1996

Translated by Pedro Sette Câmara

“La verdad es lo que es
y sigue siendo verdad
aunque se diga al revés.”

(The truth is what is and does
continue being the truth
even though one says it is not)

Antonio Machado

Descartes assures us that the sequence in the Meditationswhich takes him from the questioning of the outer world to the discovery of the Cogito axiom isn’t just a logical model, a hypothetical articulation of thinkable thoughts, but rather an actual experience, a narrative of thoughts which were thought. But, how trustworthy might have been his self-observation? Can we take the faithfulness of what he reported for granted? Furthermore, can we take for granted the paradigmatic universality of that sequence of thoughts, admitting it will happen in equal or similar fashion, with similar or equal results, to anyone who undertakes the reexamination of the architecture of one’s beliefs from its very foundations? Is it possible for a man to have a similar experience, or, on the contrary, it was Descartes who, in fact, had an entirely different experience, allowing himself to be deceived and taking as a description what is purely an invention?

The possibility of doubting our sensations, our imaginations and our thoughts is something anyone can testify. The possibility of putting the whole set of our representations on hold, reducing the “world” to a vanishing hypothesis, is also sure.

But, after performing all these operations, Descartes assures us of having found, at last, the certainty of doubt: doubt is a thought, and in the instant I doubt, I cannot doubt that I am thinking the doubt. The self-confidence in the metaphysical solidity of the thinking ego comes forth as a powerful psychological compensation for the lack of confidence in the reality of the “world”.

Even though very keen to describe the thoughts which precede the state of doubt, Descartes is oddly evasive when it comes to the state of doubt itself. Actually, he doesn’t describe it: he only affirms it, and, jumping immediately from description to deduction, he begins to draw the logical consequences which the verification of this state imposes on him.

Let us do what Descartes did not. Let us try to stop the impulse of consequentialist automatism, and keep ourselves for a moment at the description of the state of doubt.

In the first place, it is not a state — a static position in which a man may rest unchangeably, just as he gets sad or in contemplation, still or lying down. It is rather an alternation between a “no” and a “yes”, an impossibility of resting at one of the terms of an alternative without the other coming to dispute the primacy. For either “no”, or “yes”, once accepted as definitive terms, would immediately eliminate doubt, which consists of antagonistic coexistence and of nothing else. But this antagonism isn’t static: it is mobile. The doubtful mind goes endlessly from on term to the other, without reaching a support point where to rest and “be”. And, since each of the terms is the other’s denial, the mind would not be able to rest at any of them without, for an instant, denying the other: precisely at that instant, the mind is not in doubt – it is either affirming or denying, it is affirming one thing and denying the other, even though it may not be able to persevere in the affirmation or in the denial without thinking of a thousand reasons to abandon either. And, in the instant of affirmation or denial, doubt suppresses itself as such and fights for its establishment as affirmation or denial; but it fails, and it is of this failure that doubt is made of. What follows is an inevitable conclusion: a doubt that does not doubt itself, a doubt that, suspending the alternation, imposes itself as a “state” and thus remains, is impossible. In taking doubt as a “state”, omitting that it is an alternation between two antagonistic instants, Descartes reifies it and takes it as a certainty: “I cannot doubt that I doubt in the instant I doubt”, a sentence Descartes takes as an expression of the most conspicuous obviousness, is actually the expression of logical nonsense and of psychological impossibility. What would be more correct to say is that, in doubting, I doubt about everything, doubt itself included. Doubt is not a state: it is the succession and coexistence of antagonistic states, it is a not being able to be.2

What leads Descartes into error is the fact that he takes doubt for denial, or, better yet, for hypothetical denial. I can, actually, make up a hypothetical denial and repeat it indefinitely. I can even amplify it – hypothetically, of course – to a point where it embraces what I consider to be the whole extension of my knowledge. But I cannot “doubt” about my own knowledge without, at the same time, restating it, in the sense that this is the only way through which one is able to alternate its affirmations and denials in the vicious circle of doubt.

Set forth in these terms, the cogito axiom is no more than a new and more obscure enunciation of the old argument of Socrates against the skeptical: it is impossible to deny without affirming the denial, thus without affirming something. But, from this point of view, the Cartesian discovery is reduced to very little: far from having given a new foundation, critical or negative, to the world of knowledge, he makes nothing but demonstrating again, through the crooked path of a false psychological self-description, the logical primacy of affirmation over denial. However, the acknowledgement of this primacy is, simultaneously, the denial of doubt as a founding act. The discovery of Descartes is a non-discovery, it is the discovery of the impossibility of discovering anything through a way defined, more than any other thing, by an intolerable self-contradiction.3

But, with that, I have solely demonstrated that doubt, as such, cannot serve as a critical basis. I did not expose yet the bases which, in their turn, make doubt possible. And this is the decisive point, for if there is something “behind” doubt, it is this something, and not doubt, which constitutes the firm support point Descartes looked for, and which he naïvely believed to have found in the acknowledgement of doubt.

Descartes says that doubt is a certainty in the instant it is thought. But that is false: what is a certainty is the later reflection which affirms the reality of the experience of doubt. In the very instant of doubt, what happens is, as we have seen, an alternation between affirmation and denial, and hence the impossibility of affirming any state, if by state we understand, as one should, the coincidence of a factual judgement and the feeling which grants its negative or positive value, as in sadness, hurry, anger, hope etc. Doubt is not a state, for the simple reason that in it the feeling, which can be of anxiety, of hope, of curiosity, etc., does not coincide with a specific judgement, but emerges precisely from the impossibility of affirming or denying a judgement. It is rather a moment of suspension between states, an agitated void that contains the germs of various possible states – at least two – and never settles upon any of them without its own suppression. Therefore man never “is” in doubt: he simply passes by it, precisely as a transition between states. It is only when doubt is no longer a present experience and becomes the object of reflection that arises this certainty, purely retrospective and narrative: “I could not, up to this moment, establish myself in affirmation or denial.” Thus, there is not only a logical distinction but an actual separation between doubt as a present experience and doubt as an object o recollection and reflection – and it is the latter that is certain and indubitable4, not the first, even though Descartes takes one for the other and forwards to us as a direct intuitive evidence what is actually the object of later reflection. It is only this reflection that, in giving it a name, can endow with the oneness of a “state” that which is actually a succession of states which mutually suppress each other, or the coexistence of purely potential states, each of them being able to actualise itself only at the cost of the others’ exclusion. Endowing the void of alternation with the positive consistence of a “state”, Descartes at the same time transforms doubt into mere hypothetical denial, taking then as an actual psychological state what is rather simply the logical concept of a possible state.

To make things even worse, in the reflective affirmation of the reality of doubt are contained, as premises, two beliefs: one in the continuity of conscience between doubt and reflection, and other in the knowledge of the distinction between truth and falsehood.

1o Anyone who reflects about doubt is aware of still being “the same” who had the doubt; and if the act of doubting is formally distinct from the act of reflection, the conscious ego, in reflecting, knows he is the subject of two distinct acts – logically distinct and temporally distinct -, taking us to the conclusion that this ego is logically and temporally anterior to both acts and independent from them: it is not the act of doubting that founds the certainty about the ego, but, on the contrary, the certainty about the continuity of the ego is the sole guarantee that doubt was really experienced. For doubt, if it did not receive from later reflection the name that endows it with the apparent oneness of a state, would end up reducing itself to the mere succession of non-related affirmations and denials, successive hallucinations of schizophrenically plural subject, deprived of the empire of himself and dissolved in the atomistic flow of his states. In order to become the object of reflection, doubt is endowed with the artificial oneness of a name; and if just after that the mind forgets that this mind is a mere ens rationis and takes it for a substantial unity, then we are faced with a case of reflexive self-hypnosis in which a name produces magically, a posteriori, the reality of its object.

2o Being formally distinct, the two acts are empirically distinct as well, that is, distinct in time: first I doubt (that is, I come and go between successive affirmations and denials), then I reflect that I doubted (that is, I unify under the name “doubt” this multitude of antagonistic experiences). But the oneness of the ego, implied on this very reflection, and hence in the certainty of the doubt, is that continuity in time denominated memoryand recollection: memory, a premise for reflection, is logically and temporally anterior to reflection. Far from being able to base our confidence on memory, it is doubt who depends on it to have a logical foundation and to become possible in the realm of psychological facts.

But, if doubt depends on the guarantee it receives from the ego and from memory, then it has no founding capability. It is a founded thing, a secondary and derived certainty, it is the work of a more profound and unquestionable agent.

3o However, doubt implies something else. How is it possible to doubt? The possibility of doubt rests entirely over our ability to conceive things in a way different from the way they present themselves in a given moment. Doubt rests over supposition; it requires and implies the ability to suppose. Once things have presented themselves to the subject in one certain way, and not another, this second and supposed way can present itself to conscience only as a work of the subject himself, as an offspring of imagination or as a conjecture. In order to know he’s doubting, it is necessary for the subject to know that he supposed something and to thus acknowledge himself as the subject of not only two acts, as we have just seen, but of three: the act of doubting, the act of reflecting about doubt and, before both acts, the act of supposing or imagining. Imagination is, in addition to the continuity of the ego and to memory, the third requirement and the third basis for the possibility of doubt.

4o But, should the subject never notice any difference between things as they present themselves to him and things the way he imagines them, the subject would never be able to know that he supposed, since there would be no distinction between to suppose and to perceive. Therefore, the awareness of this difference is also a requirement and a basis for the possibility of doubt. To doubt, I need to distinguish, in my representation, what’s given and what’s construed, what’s received and what’s invented, that which I get already finished and that which I do and propose. Consequently, here is the assumption of the difference between the objective and the subjective, and, thus, the belief in the objectivity of the objective and in the subjectivity of the subjective.

5o Yet, should the subject confuse these two domains, believing that he supposed what was perceived and that he perceived what he supposed, he would have lost the continuity of conscience and of memory, which is, as we have seen, a condition for the possibility of doubt. Therefore, the doubt about the reality of the world cannot present itself as a mere choice between two possibilities of equal value sprang from the same origin, but always as a choice between something given and something supposed, between the perceived and the invented.

6o It is not possible to doubt about the reality of the world without knowing first that this doubt, and the supposition which serves as its basis, are but pure inventions of the subject himself, and this invention is formally and temporally distinct from the act of perceiving and from the perceived content. Doubt is a supposition that an invented world is more valid than the received world, a supposition based, in its turn, in the conscience to invent, to suppose and to pretend. The doubt about the reality of the world is always and necessarily an act of pretending, and the more the pretender works to take this doubt seriously, to make it more and more verisimilar, the more the glow of the performance will attest to the difference between the verisimilar and the true, as in a play, we applaud the actor exactly because we know that he is not the character.

7o But this conscience of pretending would be impossible if it were not founded, in its turn, on the conscience of the difference between to think and to be, to imagine and to act. For, once it is implied the conscience of the difference between to supposing and perceiving, parallel to the conscience the ego has of its own actions, there wouldn’t be a way to deny that the thinking ego is conscious of the difference between the supposed action and the effective action, once the effective action is not just thought, but physically perceived, exactly like the beings from the physical world. I cannot, therefore, doubt about the beings of the physical world without, in the same act, doubt about the physical acts I see myself performing, like the movements of my hands and legs. But, at the same time, I cannot doubt about them without questioning, in the same instant, the continuity and the oneness of the ego, which is, in spite of it, a premise for the act of doubting about just any thing. Here is another reason for which doubt, being dubious in its own nature, would not be able to establish itself if not by doubting about itself, that is, being aware that it is founded on a supposition and on deliberate pretending. Here is also why doubt is something so rare and demanding: it implies a movements that contradicts itself, that questions the very conditions which make it possible.5

8o Finally, doubt is only possible when it is known that something, either in what is perceived or in what is supposed, is not adequate, when it does not meet a fundamental requirement of truthfulness. But how could the doubting subject demand truthfulness of his suppositions if he did not have any idea about truthfulness? This demand would be inconceivable without an idea of truth, even as a mere imaginary object of desire. The desire for bases presumes in the subject at least the possibility of imagining that his knowledge can be even more reliable than he actually feels at a given moment, that is, truth as an ideal and the option for truth. But, at the same time, we saw that the subject did not know this truth just as an abstract ideal, but that he already was aware of at least one actual difference between truth and falsehood: the difference between what is given and what is supposed, followed by the true awareness that what was received was not supposed and what was supposed was not received.

Thus, doubt erects itself upon a whole building of perceptions and presumptions: far from being first logically, it is a very elaborate and sophisticated product of a knowledge machine. Far from having a founding power, it is no more than a more or less accidental and secondary manifestation of a system of certainties.

However, if things are as such, if the primacy of methodical doubt is only the primacy of a mistake, then are under suspicion, similarly, the Kantian primacy of the critique problem, the positivist dogma of the impossibility to have a valid metaphysical certainty, and many other belief which today’s man takes, in spite of his own will, as obvious and blatant truths. But this is a subject for future addresses, which will be presented in other opportunities. Thank you.


  1. First part – abridged – of the essay “To Doubt about Doubt and to Criticise Criticism: Preliminaries for a Return to Dogmatic Metaphysics”, distributed among the students of the Permanent Seminar on Philosophy and the Humanities in March 1996. 
  2. When I talk about “succession and coexistence”, it seems that I utter a monumental nonsense. But the yes and the no of which doubt is made are coexistent in one aspect, and successive in another. Logically coexistent as terms of a contradiction, they are psychologically successive, that is, they enter the stage of conscience in a cyclical, alternating mode: one enters, the other leaves, as day and night coexist in the sky and succeed each other in a certain point on Earth. 
  3. A first version of this analysis of Cartesian doubt can be found in my book Universality and Abstraction and Other Studies (São Paulo: Speculum, 1983), under the title “The Cartesian cogito on the light of spiritual psychology.” 
  4. “Certain and indubitable” or “uncertain and doubtful” are predicates which do not apply to a fact as such, but to the judgements we make about it. 
  5. It is a deviation of the human mental apparatus, a painful movement that suppresses itself, and which rare men are able to endure for much time without great risk to their psychological integrity. The possibility of taking this risk and overcome it rests on the existence of a so solid, so deeply rooted body of beliefs, that a man may grant himself the luxury of leaving it for a mental trip, sure to find it again when he comes back. This possibility, in its turn, can only be accomplished in the highly differentiated urban societies and cultures, where the thinking individual is given space for flights of imagination which will affect in nothing his conduct as a citizen or as an honourable subject attends to his duties; where he is given, more than that, free space to think one thing and do another, to cultivate the defensive hypocrisy which is notably absent among the primitives, and that, for good or bad, is a solid protection of the individual conscience against the tyranny of collective discourse. Hence the peaceful coexistence between the revolutionary audacity of Cartesian doubt and the conservatism of the “provisory morals” that make it possible.