[Published in Philosophical Investigations, 33:4, 304-320]
Olli Lagerspetz, Philosophy, Åbo Academy
The aim of this paper is not to criticise Descartes for some error he may have committed but to look at what he says about the concept of error or mistake. The overall effect is to raise some doubts about the customary association of Descartes’ philosophy with Methodological Solipsism and Other-Minds scepticism.
Meditations on the First Philosophy1 belongs to the relatively few works of Western philosophy where the occurrence of cognitive or intellectual error2 is explicitly presented as a feature of our thinking that needs explanation.3 In particular, that discussion is included in the Third and Fourth Meditations. Given his general take on the philosophy of mind and perception, Descartes is facing two problems:
How is it possible that we commit errors, given the existence of a non-deceiving God? Descartes presents his answer in the Fourth Meditation, highlighting ’non-cognitive’ aspects of factual judgments (indeed, as we will see, much in the same way as Antonio Damasio does, in a purported objection against him, in his book Descartes’ Error).
How is the concept of error possible in the first place – given the fact that, according to Descartes, the concept of falsity is not applicable to ’ideas’, or mental contents narrowly defined? How could the thought of being in error possibly enter the mind of the subject of the Second and Third Meditations? As we will see, Descartes’ answer, invoking the subject’s ’idea of perfection’ bursts the initially solipsistic frame of the Meditations.
A discussion of these issues, I suggest, will lead to the conclusion that Descartes’ philosophy is less egocentric and more dialogical than is usually assumed. Here it indeed compares favourably with some recent philosophy of mind. Descartes may escape some criticisms applicable to later philosophers who have, consciously or unwittingly, incorporated Cartesian elements into their work. A form of Cartesianism has gained wide currency in twentieth-century debate because we see its perverse appeal and ascribe it to Descartes, in order then to criticise it, while often unconsciously retaining key elements of it. Gordon Baker and Katherine J. Morris make this point forcefully in their book Descartes’ Dualism.4
To quote a critical view, Norman Malcolm writes, in a juxtaposition of Descartes’ Meditations and Wittgenstein’s On Certainty:
Descartes believes that a single human being can, all by himself, arrive at many certainties. Wittgenstein’s view is that anyone’s certainty about anything presupposes a mass of knowledge and belief that is inherited from other human beings and taken on trust. Descartes thinks that one’s metaphysical certainties have a Divine backing and, therefore, must be true. Wittgenstein’s view is that when one’s certainty of something is ’objective’, one cannot understand how it would be possible that one is making a mistake […] – but that it does not follow that one is right.5
I hope to indicate there are reasons for a more favourable comparison. Descartes’ discussion of error in the Third Meditation may also be seen as a proto-private language argument, or a ’private error argument’,6 with at least the potential of opening up questions quite similar to those raised by Wittgenstein in the relevant sections of Philosophical Investigations. This is generally missed, I submit, in part because we tend to see Descartes’ contribution as more uniform and architecturally structured than we perhaps should.
As far as I can see, the present paper is largely compatible with, but not dependent on, the novel interpretation of Descartes’ philosophy of mind put forward by Baker and Morris. They distinguish between Descartes’ dualism and the kind of Cartesian dualism often ascribed to Descartes. They question the intelligibility of ’Cartesian’ dualism within the framework of Descartes’ own (partly Aristotelian) view of the mind.7 The present paper does not directly deal with dualism – or on how to demarcate the boundaries between ’the body’ and ’the mind’ – but rather with scepticism and solipsism, themes that do not lie at the centre of Baker and Morris’ attention. Moreover, Baker and Morris consciously want to present an internally consistent account of an overall position outlined in the corpus of his texts.8 My suggestion is instead that Descartes, in the Second Meditation, presents and, in the Third Meditation, abandons certain ideas that might be thought to support allegedly ’Cartesian’ dualism and Methodological Solipsism.
1. Descartes’ Method
First a few words about Descartes’ method. Descartes often describes knowledge as an architectural structure, hoping to base the necessary principles of scientific knowledge on the foundation of propositions that cannot be doubted by any thinking being.9 This certainly captures an important element of his contributions to science and philosophy. But Descartes insists he had a reason not to construct his Meditations ’more geometrico’, or as a deductive system based on the Cogito. In his second set of Replies he calls his mode of presentation ’analysis’, by which he means that his thoughts are presented in the order in which the meditating subject arrives at them.10
This is why I wrote ’Meditations’ and not ’Disputations’ as philosophers would do, nor ’Theorems and problems’, as geometricians would: namely, to point out that I will have no truck with others than those who do not refuse to consider and meditate the issue along with me.11
If we take him by his word we must not discuss the positions described in the different phases of the Meditations in abstraction from the process by which he arrives at his final views.12 The work tracks a meditative process where earlier positions are successively abandoned.
I am directing this warning specifically to philosophers of mind, who would typically present the predicament at which Descartes arrives at the end of the Second Meditation as his more or less final bid on the nature of the mental life. On that view, the Second Meditation presents us with a finished picture of that, while the existence of an external world remains to be demonstrated later on, as a separate issue. – Thus the relevant philosophers of mind would assume that the question how I am related to the world in experience can be compartmentalised into two questions: first, What are the contents of my mind? And secondly, How accurately do those contents represent goings-on in the external world? And they would take it that whatever answers Descartes gives to the first question would survive regardless of his answers to the second.13
That is what Baker and Morris describe as the ’two worlds view’ commonly ascribed to Descartes. According to that view, the mind contains mental entities (thoughts and sensations) understood as things, in analogy with things in an external, material world.14 The mind has privileged, indubitable access to its contents, which supposedly provide the basis for Descartes’ subsequent reconstruction of human knowledge on a secure basis. The construction of thinking as an internal world gives rise to familiar questions about how the internal and external worlds are related.
The view in the present paper is, in short, that Descartes sets out in the Second Meditation to explore how far he can get if he starts out from a purely egocentric conception of knowledge. However, he does not retain that conception. It seems to me, on the contrary, that an important break with it occurs in the Third Meditation; although there will be questions about how conscious Descartes was about the radical nature of that break.
2. The Predicament of the Second Meditation
In the end of the Second Meditation and in the beginning of the Third, the thinking subject (to which Descartes refers as ’I’ or ’ego’) is conscious of the fact that, among its thoughts, there are ’ideas’ (ideas), including those he calls ’feelings and images’ (sensus & imaginationes15). Regardless of the status of ideas in other respects, I can be sure that they ’are in me insofar as they are modes of thinking’.16 The subject knows that it thinks, and it has also knowledge of the specific qualities of its different modes of thinking, such as sensation (visual, tactile, acoustic, and so on), volition, doubt, and so forth.
In modern philosophy of mind, the initial position outlined here is known as Methodological Solipsism. It is described by saying that mental contents can be defined ’narrowly’. The contents of my thoughts – mental states, ideas or other thoughts – may be described independently of any judgments concerning their relation to facts in the external world; facts that may or may not correspond to them.17 Here, inside, everything may be the same regardless of whether I am really perceiving something out there.
Thus, according to Methodological Solipsism, a waking experience and a dream experience are essentially the same because they are the same experience: they present my mind with certain conscious states. The possibility of qualitative similarity of dreams with any waking experience is ensured by the fact that whatever the content of my experience, it is always possible to assume a dream where I enjoy the same experience.18 – Let me just say briefly that any semblance of plausibility that this view may have for us is largely a result of the thorough influence that Cartesian philosophy has exercised in the intervening centuries (largely via the Empiricists, who incorporated a certain reading of it into their concept of experience19). We certainly speak of ’seeing’ both in a dream and in a real situation. But countless things remain that can only be said in the one case but not in the other. For instance, if there is dispute about what really happened in a situation, the fact that I have seen something happen can be used as evidence. (In other words, there is an internal relation between my saying that I have – or: that someone has – seen an event and my claiming that the event took place.20 My accounts of dream experiences do not have this role.) Analogously, I swim both in my dreams and in real life, but this does not demonstrate that the real essence of swimming consists in my mental states as I swim. – Thus the presumption of complete qualitative interchangeability between waking and dreaming experiences is entirely due to the fact that ’experience’ has already been (re-)defined as a process in the mind; which is simply a requirement of Cartesian-cum-Empiricist theory. It is manifestly not what is meant by ’experience’ in a number of everyday life situations.
In passing, let me point out anyway that many philosophers of mind today see Methodological Solipsism (also known as Narrow Content and the principle of Neurobiological Sufficiency) as a valuable research strategy. One reason for this is that something like it must be assumed in any form of psychoneural identity theory. If a neurophysiological state is to be described as being identical with a mental state, it must be possible to individuate and describe that state independently of the external situation. Hence one must be able to define mental states ’narrowly’.21 At least in the present philosophy of mind there are, then, powerful incentives for taking Methodological Solipsism seriously despite its obvious problems.
The view just described is Cartesian in a received sense of the word, but Baker and Morris raise the question whether anything like it might conceivably have been Descartes’ considered view. Sensing, seeing, etc., were, for Descartes, modes of my thinking, but Baker and Morris argue that he did not claim they were independently existing entities that appeared before my mind. They were things that I do. Descartes certainly claimed that if I seem to see, or if I think (or judge) that I see22 that there is a torch in the room, then it always remains true that I think I have the experience of light. This is something I am entitled to assert regardless of whether there really is a torch in the room. Doubt is excluded.
This cannot be false, and this is what in me is properly called ’sensing’ [sentire]; and in this restricted sense [praecise sic sumptum] it is nothing else but thinking.23
But, as Baker and Morris point out, Descartes is not saying here that sensing or seeing is really thinking.24 He only says that in both sensing, seeing, and, for instance, walking, one can discern a bodily component and a mental component (sensing ’in this restricted sense’) – the latter being ’certain’.25
In any case, in the Second Meditation Descartes wants to focus on that ’thinking’ component of perception. For Descartes, this is a strategy for damage control in the face of the Dream argument. The subject described in the Second Meditation does not doubt, and cannot doubt, what he is thinking. However, the absence of doubt is bought at the price of at least provisional scepticism about external reality. There is no good reason, at this point of Descartes’ argument, to maintain that the thoughts that the subject is having constitute cases of seeing and hearing in the wider sense in which those words are usually understood: experiences of events and objects (’out there’).
3. Memory in the Second Meditation
The subject of the Second Meditation has ideas or thoughts that are neither true nor false in themselves. I may form mental images of many kinds, but as long as I have not passed a judgment about the relation of my images to anything outside them, the concepts of truth and falsity are not applicable to them.26 At the same time, according to Descartes, I will be certain that I have these thoughts. This kind of certainty pertains to my thoughts at the present moment. Now what about the subject’s knowledge of its past experiences?
Descartes discusses memory in two senses.27 On the one hand, he speaks of memories as judgments about events (mental or external) in the past. Those judgments involve truth claims and may be subject to doubt. For instance, extended chains of reasoning are vulnerable because they need to be retained in the memory.28 But in the present context of the Second Meditation, memory is simply the faculty that retrieves mental contents from the past without implying truth claims. In the Fourth Meditation, he includes memory, together with imagination, among the mental faculties whose exercise does not imply truth claims, and which are contrasted with the will.29
Pointing to a close analogy: the ’I’ now is in the same predicament as Wittgenstein’s keeper of a private journal, introduced in the Philosophical Investigations.30 Each time I have a certain sensation I jot down ’S’ in my journal. ’S’ is my private sign for the sensation, which I have named in an act of private ostensive definition – by concentrating my attention on the sensation and calling it ’S’. Now I recognise my present sensation as being sufficiently similar to the one I previously enjoyed; and I call it ’S’ once again. Wittgenstein questions the assumption that the ostensive definition makes me ’remember the connection right in the future’ between the sign and the sensation.
One way to understand Wittgenstein’s point would be to say that memory cannot be relied upon when applied to exclusively private states of mind. There is no way I can make sure the sensation I previously called ’S’ is sufficiently similar to the one I am enjoying now. This interpretation was put forward by A.J. Ayer and criticised by Rush Rhees.31 However, the real point is stronger and really not dependent on any limitation of my cognitive faculties. Should we try to imagine a being that always remembers everything right, it would still be caught in the same problem. Or rather, it would be unclear what the criteria are for remembering ’right’ or ’wrong’ in this case.
I have no ’criterion of correctness’32 for the different designations that I give to my sensations. Whether or not two sensations should be described as being sufficiently similar will, in this case, be dependent on just what I am inclined to see as sufficient, and only as long as I am so inclined. Anything or nothing may be enough. The subject ’has something – and that is all that can be said’.33 And even that would be to say too much (from the point of view of what the subject himself would be able to say34): ’”has” and ”something” also belong to our common language’35. I, the subject, feel – perhaps – that I have got hold of something peculiar, but all I can say about it boils down to ’an inarticulate sound’.
A word of caution. Wittgenstein is not presenting a knock-down argument turning on the premiss that words can only be applied if definite criteria of correctness can be applied in their use. Rather he says that the distinctions between meaning and not meaning, remembering and not remembering, experiencing and not experiencing disintegrate in the complete privacy that he is proposing to imagine. – The situation initially appears clear when we look at it: a person trying to keep track of his sensations, sensations that we identify and imagine him as having. Thus the perspective from which we first approach the situation is not that of the private diarist but that of an omniscient narrator. But that initial frame of reference disappears when we try to put ourselves into the shoes of the diarist. We no longer have two descriptions: the situation of the private diarist as we know it to be, and the correlated description that we suppose he will be able to give of that well-defined situation. The scenario disintegrates when only the diarist’s perspective is available. Its presumed intelligibility depended on the tacit assumption of an objective point of view from the outside; thus, on the fact that we did not really imagine the subject’s sensations as private after all.
If this interpretation is valid, it also follows that the subject of the Second and the beginning of the Third of Descartes’ Meditations cannot remember – in the sense of ’recalling past events, either mental or material, either inside or outside its own mind’. The subject has the experience of certain psychological stuff – ideas – being present, but what that is is quite undecided at this stage. Nor can the subject really be said to know anything about its states of mind. It has them; that is all. It cannot reliably compare them with earlier or later material, nor do any other things normally associated with ’knowing’. It also follows that the meditating subject’s radical uncertainty at this stage is not simply the outcome of the suspicion that an evil demon may be deceiving him also about his own states of mind; nor would it be remedied by the mere removal of that particular suspicion.36 As long as the subject only has access to his own states of mind he does not know anything about them.37
Perhaps surprisingly, Descartes indeed says so much. In the early part of the Third Meditation he says he is certain of nothing – not even of his own existence, nor that 2 + 3 make 5. In his present state (i.e., as long as he is not assured of the existence of a non-deceiving God), he cannot even exclude possibilities ’in which I see a palpable contradiction’.38
Norman Malcolm comments, ’I do not see how one can avoid the conclusion that Descartes worked himself into a dreadful confusion in the Meditations’. Echoing Arnauld’s famous accusation of circularity39, he adds, ’Descartes is in the impossible predicament of trying to hoist himself by his own bootstraps’.40 – However, it seems possible to respond that Descartes is, at least to an extent, conscious of his confusion at this stage of the Meditations. It does not represent his last bid but rather a transitional position that he hopes to overcome.
I said Descartes was conscious of his situation ’at least to an extent’. There is some debate on his exact position. In his second and fourth set of Replies, he seems again to be understating the radical nature of the confusion he has reported.41 Now he seems to be saying that the difficulties only arise because of the unreliability of memory and hence do not affect my knowledge of such truths as I can intuit directly; for instance, of my own existence.42 This, however, is in direct conflict with Descartes’ earlier express words. Perhaps this shift in Descartes’ apparent avowed position is in analogy with the difference between Ayer and Rhees on private language? – On the other hand, Timo Kajamies argues43 that, in the passage most frequently invoked in support of the more conservative interpretation of the range of Descartes’ doubt,44 Descartes is referring to a passage in the Fifth Meditation45 where (unlike the Third Meditation) the truth of his first principles is already established. To read that argumentative setting back to the Third Meditation is to disregard the dynamic character of the Meditations.46
4. Descartes’ Private Error Argument
The overcoming of Methodological Solipsism in the Third Meditation is achieved by mobilising an element from the outside. Alternatively – Descartes demonstrates that an element from the outside was present already at that earlier stage. The inclusion of unacknowledged material is what Arnauld criticised as an illegitimate move.47 But it can also be seen as Descartes’ recognition of the limitations of the position he has previously outlined. In order for his earlier description of the subject’s mental capacities to make sense at all – describing the subject as doubting, willing, and so on – one must avow the presence of some element not included in the original description.
The concepts of error and doubt are central here. Descartes recognises it is inconceivable that the subject, as presented in the Second Meditation, could ever be in doubt about anything. The concept of (’formal’48) falsity cannot be applied to ’ideas considered merely in themselves’.49 However, the meditating subject does have doubts. Descartes argues that in order for the subject to do so, it must possess ’the idea of perfection’. This gives conceptual space to the subject’s idea of its own lack of perfection.
Descartes establishes that, among his different ideas, there is one that cannot have its origin within himself. This is the idea of God, the perfect being. That idea cannot come from him because he is imperfect and since ’there must be at least as much in the entire efficient cause as there is in the effect of that cause’.50 Since he is finite, the idea of an infinite substance cannot originate in him.
The importance of an idea of absolute perfection lies in the fact that in its light I can recognise my own shortcomings. My recognition of the possibility of a perfection greater than my own is, in other words, internally related to my awareness of the fact that I am prone to error. We may compare this with the use of geometrical figures in reasoning. Geometrically speaking, a circle on paper is never entirely correct. But its shortcomings can only be seen by someone who treats it as an approximation of a perfect figure.
For what reason should I understand that I am in doubt, that I hope for something, that is, that I am in want of something and not perfect in every respect, unless there were an idea of perfection in me, in comparison with which I saw my own shortcomings?51
This is a transcendental argument of a kind. A solipsist that doubts is impossible.52 Since it is a fact that I doubt, it follows that I am not a solipsist: I am already counting on someone or something outside me, a source of truth independent of my own mind. – The possibilities of both knowledge and doubt are thus internally related to the idea of someone being in a position to correct me.
Descartes calls God a substance, which is the term he also uses for the thinking and extended substances. But, as Emmanuel Lévinas has emphasised, he is not simply completing his metaphysics by adding one more substance to it. In Lévinas’ terms, his reasoning ’breaks up the unity of the ”I think”’.53 Descartes says explicitly that he cannot ’comprehend’ the infinite. There may be aspects of God, he says, that he cannot understand in any way nor even ’touch with his thinking’ (attingere cogitatione54). As Lévinas puts it, the idea of God ’signifies the non-contained par excellence’.55
It overflows every capacity; the ’objective reality’ of the cogitatum breaks up the ’formal reality’ of the cogitatio. […] We will say that the idea of God breaks up the thought which is an investment, a synopsis and a synthesis, and [which] can only enclose in a presence, re-present, reduce to presence or let be.56
The important thing for this argument is not the mere existence of a perfect being, nor indeed the fact that this being is perfect. The point is that it is a being to which I stand in a certain relation. It is not one more ’thing’ of which I am conscious (and which could still be construed solipsistically, in terms of my ideas) but it is a being that, somehow, has a claim on me. I stand in a relation to it, where that being is in a position to correct me. The importance of the fact that the other is God and not a human being lies, perhaps, in the fact that a solipsistic subject might still construe other humans as bunches of ’ideas’. God, by ’overflowing every capacity’ of the subject’s mind, obviously resists such construals. Descartes’ appeal to God thus implies de-centralisation or de-subjectivation of thinking, a moving away from the position of the Second Meditation. The presence of someone else introduces the possibility of dissent. The subject acquires the concept of being right or wrong, or a logical space for disagreement. Judgment is internally related to the idea of a discussion where our interlocutor may teach us something new and unexpected.
Thus the meditating subject’s progression in the Third Meditation may be described as a ’private doubt argument’ or a ’private error argument’. As Lévinas emphasises, the idea of being in error, and hence also of objectivity, presupposes recognition of the other, of someone else than the thinking subject, as a possible source of truth.
Like Wittgenstein’s ladder57, the egocentric position of the Second Meditation is overcome and thrown away in the Third Meditation.
5. Error According to the Fourth Meditation
The Third Meditation has established a non-deceiving God who then serves as a quality control of the subject’s clear and distinct intuition. But this faces Descartes with the opposite horn of the dilemma: how is human error possible at all? Ideas, as we saw, cannot be false in themselves. His famous answer in the Fourth Meditation is that error is the result of two interplaying factors: my intellect (facultas cognoscendi, intellectus) and my will (facultas eligendi, voluntas). Intellect in a narrow sense does not involve judgment.
By means of the intellect I merely intuit such ideas concerning which I can pass judgment, and no error properly speaking can be found in [the intellect] considered in this restricted sense.58
To will consists in ’asserting or denying, pursuing or avoiding’ (affirmare vel negare, prosequi vel fugere59). Error comes about when I use my will, pass judgments (judicia60) and assert or deny the ideas with which my intellect provides me.61 Error is then not ’pure negation’ or ignorance. It involves actively pursuing the wrong direction.
Due to the fact that my will has a greater range than my intellect I am tempted to extend my judgment to areas of which I have no knowledge.62 Bodily sensations are, in this sense, temptations: they may give me ideas – e.g., of secondary qualities – that I incorrectly ascribe to objects. The central thesis is that error comes about when existing cognitive material is confronted with an impulse to apply the material. Descartes thus construes the thinking subject as active, in an anticipation of Kantian approaches.
Incidentally, Descartes and Antonio Damasio seem to have a similarity of approach here, although Damasio is obviously unaware of it. In Descartes’ Error63, Damasio argues that all rational thinking in the normal sense includes judgments concerning plausibility and importance. These are not simply derived from considerations of logical coherence or factual correctness. A kind of emotional element must be presupposed.64 Like Descartes, he assumes a dynamic faculty that sifts in the cognitive material. Damasio supports his views with a neurophysiological argument. However, in the unacknowledged background there is a conceptual point, which has been independently spelled out by D.W. Hamlyn.
Hamlyn argues that belief cannot be merely described in terms of likenesses or representations. To have a belief is to assert what is believed.65
And if this means anything, at least it means this: if someone realises that things stand thus and not so, without caring for the fact that they do so stand, i.e., if truths and untruths concerned him equally much in the sense that he was concerned about neither, we could, on good grounds, claim that he has no beliefs.66
[L]ack of interest in the truth would exclude the possibility of believing. This means that a person who has no emotions cannot believe and cannot know either. And a machine that in some sense can identify what is fed into it in the form of information and carry out computations on the basis of it, does not have beliefs for that.67
Hamlyn, however, takes the argument further than Damasio, emphasising the importance of ’the concept of action’ and ’the social relations that in part are grounded on emotion’.68
6. The Demand of Coherence
What was previously said about the Third Meditation may in part be obscured by the fact that Descartes, in the subsequent Meditations, proceeds in a way that seems to indicate he can detect illusions all by himself by identifying incoherence.69 But it should be kept in mind that Descartes, in the later parts of the Meditations, has already passed the critical needle’s eye of the Third Meditation. The position from which he is now speaking is that of a kind of rationally adjusted everyday understanding. Life exhibits a certain normality or regularity that helps us recognise plausible assumptions about our environment.
In the last Meditation Descartes returns to the problem that initiated his search for certainty. He was sitting at the fire, as he thought, dressed, but suddenly woke up realising he was lying naked in bed.70 Now he says he can tell dreams from daytime experience because –
dreams are never united by memory with all the other events that occur in life […] for if someone actually showed up for me while I am awake and then disappeared at once, as in a dream, […] then I would be right to judge him to be a ghost or an illusion created by my brain rather than a real man.71
The suggestion is that I can discover illusions because of the lack of coherence between my different senses, or between my senses and my memory and understanding. Daytime experience exhibits a degree of regularity, which is absent in dreams. But this presupposes that the subject is already conversant with the general distinction between reality and dream. This is why the criterion could not be applied right away in the First Meditation.
Our ability to form an understanding of ’reality’ as something opposed to fantasy is part of our involvement in ’normal life’; something that, it seems to me, comes naturally as we are born and welcomed into a life with others. We pursue the truth with others, naturally accepting the possibility also of being corrected by someone else. In his Meditations, despite appearances, Descartes finally did not dispose of the role of the Other in favour of an egocentric conception of knowledge. But by substituting God for the human other, he hoped to re-create the foundations of certainty in a way that would make them independent of any specific social milieu.
Cartesian approaches within subsequent philosophy of mind (acknowledged and unacknowledged alike) have largely assumed that the egocentric position outlined in the Second Meditation is Descartes’ last bid, not a position that he abandons in the further course of his philosophical therapy. The present paper has argued for an alternative reading. The Meditations include an implicit, and partly explicit, critique of Methodological Solipsism at its very inception.72
Philosophy, Åbo Academy
Fabriksgatan 2, 20500 Åbo, FINLAND