Home   Bellotti 1   Bellotti 2   Bellotti 3   Bellotti 4

Tony Bellotti

Descartes' Method of Doubt

In this essay I will assess Descartes's employment of his Method of Doubt, as presented in his Meditations on the First Philosophy [Descartes 1641]. I will argue that by implicitly accepting a causal model of perception, Descartes did not apply the Method of Doubt as fully as he could have.

The Method of Doubt

Descartes's principal task in the Meditations was to devise a system that would bring him to the truth. He wanted to build a foundational philosophy; a basic edifice from which all further intellectual enquiry could be built. It was essential that his foundational beliefs were sound. If any one of them were at all in doubt, then it put the credibility of the whole structure of knowledge in jeopardy. Thus, Descartes utilised a method of systematic doubt to weed out those beliefs of which he could not be entirely certain. This approach is called the Method of Doubt, and Descartes likened it to 'that of a man who takes all the apples out of a barrel one by one, inspects them, and then puts the sound ones back' [Williams; p.59].

Descartes's overall method is twofold. Firstly he used the sceptical Method of Doubt. Secondly, he intended a constructive phase whereby he would rebuild the edifice of knowledge based upon the core truths that remain after the employment of the Method of Doubt. However, as Russell points out, 'the constructive part of Descartes's theory of knowledge is much less interesting than the earlier destructive part' [Russell, p.550]. This is because in reconstructing his body of knowledge, Descartes made use of many assumptions that he had not shown to have escaped the fires of the Method of Doubt. 'It uses all sorts of scholastic maxims... No reason is given for accepting the maxims' [Russell, p.550]. This essay will only deal with the sceptical side of Descartes's philosophy.

The underlying principle behind Descartes's sceptical approach is that there is a distinction between belief and truth. For example, having made a pot of tea five minutes ago, I may well believe that it is now full and ready to pour. But in truth, perhaps, someone else may already have drunk the tea and emptied the pot while I was out of the kitchen waiting for it to brew. Although I think this is unlikely, and I continue to believe the pot is full of tea, I cannot be sure of it. Thus it is possible that I may believe something, but to my surprise find that it is not true. This situation is not inconsistent. The Method of Doubt ultimately involves the task of removing all uncertain beliefs, ensuring that only beliefs that are certainly true beliefs remain in one's philosophy. Descartes states in the first paragraph of Meditation 1 that 'I had accepted, even from my youth, many false opinions for true, and that consequently what I afterwards based on such principles was highly doubtful; and ... I was convinced of the necessity of undertaking ... to rid myself of all the opinions I had adopted'.

Descartes saw that the Method of Doubt could be applied, generally, to a whole class of beliefs. Thus he would not have to indulge in the laborious endeavour of checking each and every one of his beliefs, separately. Instead, he could deal with them in groups by doubting any common characteristic that they may share. 'Nor for this purpose will it be necessary even to deal with each belief individually, which would be truly an endless labour; but, as the removal from below of the foundation necessarily involves the downfall of the whole edifice, I will at once approach the criticism of the principles on which all my former beliefs rest' (from Meditations 1). With this approach Descartes has an opportunity to bring doubt to bare on the whole class of beliefs that are based on sensory perception. Descartes best elaborates this argument using his malicious demon thought experiment.

The Malicious Demon

Descartes imagines that there is an all-powerful malicious demon (malin genie) who is controlling his perceptions, and through that control is making him hallucinate or dream that he is in the world. As Descartes states, 'I will suppose that the sky, the air, the earth, colours, figures, sounds, and all external things are nothing better than the illusions of dreams, by means of which this being has laid snares for my credulity' (Meditations 1). It is not Descartes's intention to seriously suggest that this is the case. Rather, Descartes is suggesting that from his first-person perspective, he has no grounds to suppose that it is not. Thus, he cannot be certain that his beliefs about the external world, full of sky, trees, houses, animals, other people, and so on, are true. There is always the conceivable possibility that all these things are just illusions being constructed by the malicious demon.

The conclusion that Descartes draws from the malicious demon thought experiment is that all beliefs based on sensory perception are doubtful because we can never be sure that they are not illusions or hallucinations. Thus, we cannot accept any such beliefs as true beliefs within our foundational philosophy. This means excluding the vast majority of our ordinary beliefs. I can no longer simply accept that the teapot is full. Not only that, but I also cannot accept that the teapot exists at all, nor a kitchen that the teapot is in. Neither can I believe in the living room that I am in and can see all around me now. Indeed, I cannot accept any beliefs about the world I live in. All of these beliefs are based on perceptual observation; therefore all of them are doubtful, following Descartes's argument. This doubt about the world even extends to our own bodies, because we only know about our bodies through our sensory perception of them. This is, of course, a strongly solipsistic, and therefore unsatisfactory, conclusion.

A counter-argument to the malicious demon thought experiment, that may spring to mind, is that the malicious demon is simply a completely fantastic and unlikely possibility. Our belief in the world of external objects is sound because it remains the most probable scenario. The difficulty with this counter-argument is that there is no evidence to verify such a claim. To make a probability statement about something requires some weight of evidence. For example, I know that it is very probable that I will be able to catch a train to work tomorrow at exactly 09:03 because that is what it says in the timetable, and for as long as I can remember there has almost always been a train running at exactly that time (exceptions being due to unusual train breakdowns, or the wrong leaves on the line, etc.). Without such evidence, I could not credibly claim that the train should arrive at that time. In the case of the malicious demon, there is no evidence to suppose that it is any less likely than the existence of external objects, because the evidence is precisely the same in both cases; i.e. both could cause the same perceptions. In any case, even if I could say that the malicious demon was unlikely, this would not help because its unlikelihood still allows for its possibility, so there would still be doubt cast on the existence of external objects.

The Causal Model of Perception

The malicious demon thought experiment works only if we accept the causal model of perception. Very simply, this model asserts that our perception only comes about through a causal process beginning with an external object. For example, when I look at a tree, what happens is that there is an external thing — the tree — from which light rays are reflected. Some of these light rays enter my eye and hit the retina that lines the back of the eye. On being hit, cells on the retina send electrical pulses along the optical nerve to the brain. In the brain, these signals are processed and somehow the image of the tree is formed in my mind. Thus, between the external object of my perception, and the representation (or image) of it, there are several layers of causality. This causal model of perception is illustrated in the following diagram:

External Object -- Medium -- Sensory Apparatus -- Representation

tree -- light rays -- eye; optic nerves; brain -- image of tree

Within this model, the only thing we are immediately aware of is the representation. We are divorced from the external object itself by the chain of causality. With the malicious demon thought experiment, Descartes simply changes the causal model to the following.

Malicious Demon -- Representation

malicious demon -- image of the tree

Thus, both models lead to the same representation, and the representation of objects is all that any particular person is aware of. Hence, from that person's perspective, there is no way of knowing which causal model is the correct one. What Descartes shows is that if one believes any one particular causal chain then any other will do just as well, because from the point of view of any individual at the end of the causal process there is no way to tell which is the correct conception of the external world and of how it is perceived.

Doubting the Causal Model of Perception

As expressed by Williams [p.58], 'Descartes regards it as self-evident that if I have veridical perceptions, then I have experiences which are caused by things outside myself'. For Descartes, perception of objects 'is known only through the medium of ideas' [Williams, p.59]. He considered that causality is part of the concept of perception. However, it is not at all clear that this is self-evident. It could be argued, in support of Descartes, that it is straightforwardly obvious that when we perceive something, there is a separate something that is perceived. However, what is not so clear is that this something is a thing external to the perception. Whenever I perceive something, what is evident is the appearance of the object; there is no evidence of an external cause of that appearance within the perception itself. For example, when I look at a tree, I simply see the tree. This perception does not necessarily imply an unseen cause for the appearance of the tree. Indeed, common sense tells me that what I perceive is the tree directly, not mediated by ideas as Descartes would have it. This is an argument that may be put by a follower of G.E.Moore [Williams, p.58]: since Descartes's notion of perception is contrary to common sense, this is grounds for doubting the causal model of perception, that Descartes supposed was self-evident. Other philosophers have also argued against his causal model of perception. In his Refutation of Idealism, in answer to Descartes, Kant (B274) argues that objects of perception must be apprehended immediately in consciousness.

Further, if we analyse what is meant by causality, following Hume or Kant, we find that any application of it must be empirical, hence contingent. This is because a causal relation involves linking a cause with an effect, both of which are empirical events. So it is not possible for any particular perception to necessarily have an external cause. Thus the causal model of perception must be contingent, therefore not self-evident.

It may be argued that we could justify a belief in the causal model of perception empirically. For example, I have empirical evidence of the causes of visual perception by knowing about light rays, eyes, optic nerves, and so on. And I know about these things empirically, by observing them in the world or being told about them or reading about them in books: i.e. through perception. But beliefs based on perception are the very things that are doubted in the malicious demon argument (which follows from the causal model of perception). Therefore, an empirically justified causal model of perception is inherently mistaken because by implication it undermines its own premise.

Ultimately then, Descartes offers no argument to support his causal model of perception, and as we have seen, the claim that it is self-evident can be challenged. Further, it is not possible to support his assertion by empirical evidence. Thus, it would seem that the causal model of perception is at least doubtful. Therefore, if we wish to follow Descartes's Method of Doubt completely, we must also relinquish the causal model of perception. Once it is doubted, the malicious demon argument can no longer be put.

External Reality

By assuming a causal model of perception, Descartes was implicitly assuming the existence of an external reality: that is, a world that exists quite independently of his mind and his thoughts about that world. Thus, although Descartes says that his malicious demon thought experiment implies that he must be doubtful of the existence of external things, it does not imply that at all. What it implies is that we must be doubtful about the nature of external reality. After all, what Descartes argues is that he cannot be sure whether the external causes for his perceptions are external objects corresponding one to one with his perceptions (so, for example, if he sees a chair then there correspondingly exists an external chair that is the ultimate cause of that perception), or whether the external cause is a malicious demon. In both cases, Descartes is postulating an external reality, even if it does turn out to be the malicious demon.

It is only by doubting the causal model of perception that we are in a position to doubt the existence of such an external reality, rather than just the nature of external reality. This is a stronger thesis than the one Descartes derives from his malicious demon argument. It has the merit that it does not involve Descartes's presupposition regarding external objects, internal ideas and their relationship. It is this presupposition which leaves Descartes in a solipsistic position where he must doubt the existence of external objects. Free from this notion, we can investigate what is actually meant by saying something is external or internal from the solid foundation that is promised once we have completed the Method of Doubt. Consequently, for example, ignoring Descartes's scepticism, Kant put forward the doctrine that externality is in the phenomena of perception itself, through the form of space. So external objects are immediately perceived.


Descartes's foundational philosophy, postulated in the Meditations, is in principle a sound method. Arguably, it is the only method with which to begin an investigation of epistemology and metaphysics. It rests fundamentally on the Method of Doubt that requires a first-person investigation of the relationship between the thinking individual and the surrounding world he or she perceives. This approach to philosophy, shifting the analysis to a first-person foundational investigation, has had great influence upon later philosophers, including Kant and Husserl.

However, by uncritically holding the causal model of perception as true, Descartes did not apply his Method of Doubt as fully as he could have done. By accepting the causal model of perception, Descartes held on to an unhelpful presupposition of what is meant by the external and internal worlds. This presupposition leads, through the malicious demon argument, to solipsism. By casting doubt on the causal model of perception it is possible to complete the Method of Doubt free of this presupposition.


Descartes, R. (1641) Meditations on First Philosophy translated by Veitch, J. (Everyman 1912)

Russell, B. (1946) A History of Western Philosophy (Unwin)

Williams, B. (1978) Descartes: The Project of Pure Enquiry (Penguin)

Kant, I. (1787) Critique of Pure Reason translated by Norman Kemp Smith (Macmillan 1929)