Lockdown arrived in a fidget of activity. I spent the day before helping in my daughters’ school, hoping to feel of use. The teachers were busy photocopying worksheets, trying to figure out what would be required of them in the coming days and weeks. I talked philosophy with the children, class after class, in the large and draughty hall. We talked about aliens who wore sofas on their heads and whether sitting on your sister turned her into a chair. The space fizzed with laughter and chatter. The next day we all withdrew into our homes.
Descartes’s Meditations on First Philosophy begins with a withdrawal. The meditator closes the door to his study, clears his mind of all worries, and settles down to determine which of his opinions are stable and likely to last. There he stays for six days, reflecting on his beliefs, on God, on nature. There he comes to recognize his essence as a thinking thing. Descartes’s philosophy begins in confinement. It begins in isolation.
Many critics have worried about the privilege of this starting point, about the arrogance of assuming that philosophical progress can be achieved all on one’s own. “The life which I am constrained to lead,” Princess Elisabeth of Bohemia writes in a letter to Descartes, “does not leave enough time at my disposal to acquire a habit of meditation in accordance with your rules.” Philosophical reflection in a stove-heated room is all well and good but at some point the children need feeding.
This criticism of Descartes has it that the content of his philosophy is tainted by the isolationist methodology it takes up. The two connect because Descartes’s withdrawal from the world is not simply a reflection of the difficulty of doing philosophy when surrounded by others, by the “interests of my house, which I must not neglect, … the conversations and amusements that I cannot evade,” as Elisabeth has it in her letter to Descartes. It also sets the starting point for Descartes’s investigations.
What is this starting point? Descartes’s methodological isolation leaves him with a certain set of capacities as the tools for his investigations—no magnifying glass or telescopes needed here. These are the capacities that can still be exercised in isolation, most prominent among them the capacities involved in recognizing his opinions, in giving and withholding assent, and in reasoning on the basis of judgements he has made. These capacities are self-conscious capacities since they involve the meditator’s ability to recognize and assess some aspect of his perspective. Much of what we know and much of what we do is mediated by the exercise of these capacities. Think of the way we reflected on whether we were right to think that the schools would reopen in May, or whether it is appropriate to worry so much about a loved one’s health. Self-conscious capacities play a central role in our intellectual, ethical, and collaborative lives.
Self-conscious capacities are the tools of Descartes’s investigations. But they also give him the content for his inquiry. For the aim of the Meditations is to reflect on the way these capacities enable knowledge and to show how much knowledge of the nature of things can be gained simply by reflecting on the nature of these capacities and their role in inquiry itself. The project of the Meditations thus begins not only in the exercise of self-conscious capacities but also on the exercise of self-conscious capacities. It begins with reflection on reflection.
One continual source of exasperation with earlier philosophy is its insistence on starting somewhere where we no longer wish to start. This is all the more so when that starting point is motivated by considerations which now seem to us quaint or mistaken. It is common to present Descartes’s motivation in starting with self-consciousness as resting on an illegitimate demand for certainty. This can seem quaint in its requirement that stability be achieved only through building on firm foundations and mistaken in assuming that the meditator’s knowledge of his self-conscious life is somehow more secure than his knowledge of the world. Absent those commitments, why should we follow Descartes in taking self-consciousness to be the starting point for our investigations?
Well, philosophy has to start somewhere. And Descartes’s isolationist starting point need not be motivated by the thought that our self-conscious capacities are especially secure in delivering knowledge about our own minds. It can start instead from a more flat-footed recognition that these capacities are central to our human lives, both intellectual and practical. Understanding the nature of these capacities is thus a central part of understanding what it is to be human and so it is natural for a reflective person to alight upon those capacities that make reflection possible. Descartes’s isolationist starting point can be seen as a way of making vivid an inquiry which turns self-conscious capacities on themselves as a way to examine how much is built into the fact that we are creatures with such capacities.
Why think that our ways of thinking and experiencing the world are any less subject to historical contingencies than our ways of speaking and dressing?
Kant’s Critique of Pure Reason shares this construal of the philosophical project. All of the questions of philosophy, he tells us, can be found in the question: what is the human being? Kant does not think that we need to isolate ourselves in order to answer this question, but nor does he think it an empirical question to be answered solely through anthropology. Instead he is moved by the recognition that we are self-conscious creatures, capable of reflecting on our thoughts and actions. And at least part of his interest is the question of how much knowledge of the world is available to someone who has these capacities. On this way of understanding the Meditations and the first Critique, each starts with self-consciousness and aims to see how much knowledge of the world follows in its wake.
A lot, according to Descartes. The first edition of the Meditations takes as its subtitle: “in which are demonstrated the existence of God and the immortality of the soul.” The second edition softens the claim to the more accurate: “in which are demonstrated the existence of God and the distinction between the human soul and body.” They make clear that Descartes thinks great mileage can be wrung out of the nature of self-consciousness. Simply by reflecting on the nature of self-conscious capacities, we can come to recognize our own existence, our nature as a thinking thing, the existence of God, and, eventually, the existence and nature of the material world. Kant is similarly optimistic: it follows from the very fact that we are self-conscious that we are situated in a world of causally interacting substances, located in space and time, with properties that change over time according to general laws. Reflection on self-consciousness can get us to knowledge of the world.
There can seem something magical about these projects. For how could reflection on the capacities involved in self-conscious thought lead us to the distinction between soul and body or to the existence of a world of causally interacting substances? In Descartes’s case, the magic is underwritten by a proof of God’s existence, since God’s status precludes our being deceived about the material causes of our beliefs. For Kant, it is underwritten by his claim that the nature of the world depends, in some sense, on our ability to cognize it. It is God’s grace and beneficence and Kant’s transcendental idealism which allow us to get substantive knowledge of the world from reflections on the nature of self-conscious capacities.
Neither of these options seemed tenable to philosophers in the twentieth century, and few today are brave or foolhardy enough to have God or idealism play the role they did in Descartes’s and Kant’s own arguments. One option is to give up the starting points. Another is to keep the starting points but weaken the conclusions. Instead of trying to uncover the nature and existence of the world, we aim only to show how self-conscious subjects must think or experience the world. Perhaps self-conscious subjects must think of or experience themselves as situated in a world of objects. This gives us a link between self-consciousness and objects in some broad sense. But the retreat from worldly conclusion obviates the need for God or idealism.
This option played out in two ways in the twentieth century. The first is found throughout the phenomenological tradition. It starts with self-conscious subjects and aims to show that they must experience themselves as embodied subjects, situated in the world of objects. This is an argument from self-consciousness to the world in some sense but the project is now understood as aiming only to show how we must experience the world. It was a self-conscious reaction to Kant’s own arguments and aimed to avoid transcendental idealism by confining itself to conclusions about the way things are presented as being rather than how things actually are.
The second route is now most associated with P.F. Strawson but it threads throughout mid-twentieth-century analytic philosophy. It too starts with self-conscious subjects but the aim instead is to show that they must think of themselves as embodied subjects, situated in a world of objects. Strawson describes this, in the subtitle and introduction to his book Individuals, as a project of descriptive metaphysics: that of showing how self-conscious creatures such as ourselves must think about the world. And in his commentary on the Critique of Pure Reason, it is explicitly presented as a version of Kant’s project, shorn of its commitment to transcendental idealism.
These retreats can feel unsatisfying to those who share the ambitions of the Cartesian and Kantian projects. (“When the children ask for bread,” Michael Dummett once asked, “will you give them a stone?”) But even on their own terms, they betray a striking confidence about the immutability of our perspective on the world, one which contrasts with the way R.G. Collingwood, Stuart Hampshire, or Bernard Williams understood their own work. Our fundamental ways of thinking, Strawson says, are ones which “[have] no history—or none recorded in histories of thought; there are categories and concepts which, in their most fundamental character, change not at all.” But why think that our ways of thinking and experiencing the world are any less subject to historical contingencies than our ways of speaking and dressing? And why should it be easier to show how self-conscious subjects must think or experience the world than it is to show how the world really must be?
What are the practices that sustain our faith in ourselves as the agents of our thinking?
The alternative is to reject the Cartesian and Kantian starting points. Why should we hamstring our intellectual activities by abjuring the deliverances of our senses? Descartes’s isolation goes too far. At the tail end of the eighteenth century, the philosopher, physicist, and aphorist George Christoph Lichtenberg makes the opposing complaint. Descartes’s starting point is not isolationist enough, he says, for it involves claims to which he is not entitled. “One should say it is thinking, just as one says, it is lightning. To say cogito is already too much as soon as one translates it as I am thinking. To assume the I, to postulate it, is a practical requirement.”
Lichtenberg’s remarks were enormously influential on those who read them: Nietzsche, Schopenhauer, the logical positivists, all drew inspiration from them. But their point is easily misunderstood. Lichtenberg is not making a claim about the possibility of a subjectless episode of thinking. The idea is not that Descartes needs to rule out the possibility of thoughts without a subject in order to move from self-consciousness to the world. Lichtenberg is rather raising a question about our entitlement to think of ourselves as the agent of our thinking. Sometimes thoughts strike us like lightning. When this happens, we are their patient. It is central to our self-conscious lives that this is the exception: we are first and foremost the instigating agents of our thinking. The question raised by Lichtenberg is: with what right do we think of ourselves as such?
Lichtenberg’s remarks raise a puzzle that has surprising force. The self-conscious thinking undertaken in isolation is an agential activity: giving and withdrawing assent is something we do. What entitles us to start with this agential activity? Neither experience nor conceptual mastery nor the machinery of Kant’s transcendental philosophy put us in a position to know that we are the agents of our thinking. Lichtenberg’s alternative is that to assume the I, to postulate it, is a practical requirement. Making sense of this suggestion requires us to draw a distinction between two different ways in which we can take a stand on things. One is theoretical, of the sort involved in believing that things are thus and so. The other is a distinctive sort of practical assent which Kant sometimes calls faith.
Kant thought we had to have faith in God’s existence. He meant by this that we are required to assent to God’s existence as a condition on being subject to the moral law. And since this assent is practically required of us, in virtue of being subject to the moral law no less, it possesses a distinctive sort of practical justification which contrasts with the ordinary theoretical justification conferred by experience or a priori reasoning. Faith is thus not to be contrasted with justified assent. It is a form of it. Reflection on Lichtenberg’s remarks show that self-conscious subjects do not have theoretical grounds for believing that they are the agents of their thinking. Rather, it is a condition on being self-conscious that we practically assent to the claim that we are the agents of our self-conscious thinking. We are required to have faith.
What relevance does this have for the Cartesian and Kantian projects of moving from self-consciousness to the world? Lichtenberg’s remarks show that we must have faith in ourselves as the agents of our thinking. But faith can be undermined and it can be sustained. This is a central theme of the final part of Iris Murdoch’s Metaphysics as a Guide to Morals, a work which has no interest in the Cartesian and Kantian projects. Murdoch instead draws attention to the beliefs, experiences, practices, and rituals which religion once provided for us and which she thinks need to be co-opted to support our search for the good. Kant, too, recognizes the way that specific human institutions such as the historical church can sustain articles of faith. And this provides us with a connection to the world, since our faith in ourselves as the agents of our thinking can be and is sustained by beliefs, experiences, and practices that relate us to the world. Self-conscious subjects must have faith in themselves as thinking agents, and it is our relations to the world that sustain this faith.
What are the practices that sustain our faith in ourselves as the agents of our thinking? Prime among them is the practice of holding and being held accountable for one’s judgements. Consider philosophical discussion. At its best, it involves holding people accountable for their views. When you tell me that there are no such things as chairs, I take you at your word. I may ask for reasons. I will criticize your view. (“What is this I am sitting on?!”) It would not make sense to engage in these practices if you were programmed to hold your views or if your thoughts struck you like Lichtenberg’s lightning. The questions and challenges involved in philosophical discussion are part of a practice in which we hold each other accountable for our thinking.
This can sound overly intellectual. But it is part of ordinary life. Sitting with a group of friends in a pub in Manchester one Christmas Eve, I confessed my love of the Christmas song playing on the radio. To say that I was held accountable for my thinking would be to put it politely. (To feel the force of this story, you either have to know my friends or have been in The Trevor on a Christmas Eve.) Mockery, ridicule, teasing: these are also ways of being held accountable for one’s thinking.
Our sense of ourselves as the agent of our thinking is an article of faith. That faith is sustained by the practice of holding one another accountable for our thinking.
This practice is central to our interpersonal relations. It is part of what Strawson once characterized as the participatory stance, that view on others that we take up when we see them as accountable for their actions, as appropriate targets of attitudes such as gratitude and resentment. To hold someone accountable for their judgement is take that judgement to be imputable to them. It is to take them as the agent of their thinking. The practice of being held accountable provides us with a framework with which to sustain the idea that we are the agents of our thinking. It is not necessary for this faith. But nor is it irrelevant. Like the stories, images, and rituals that make up a religious tradition, it is a way of shoring up that assent we make on practical grounds alone.
And so to the world. For the practice of holding and being held accountable necessarily involves someone other than myself, someone to whom I am answerable and whom I am in a position to call to account. This other thing is not the world of bodies that Descartes leads us to in the Sixth Meditation. Nor is it Kant’s world of causally interacting substances, located in space and time. It is another person, distinct from me and my activities. And that allows us one kind of resolution of the Cartesian and Kantian projects. For the faith in myself as the agent of my thinking which is bound up in the exercise of my self-conscious capacities is itself sustained by a practice centred on thinkers independent of me and my activities. Isolation ends not through God or idealism but through faith and practice. Self-consciousness is sustained by others in the world.
Does this complete the Cartesian and Kantian projects? It falls short of all that Descartes and Kant hoped to achieve, and in two ways. First, in the strength of the connection: both Descartes and Kant wanted to show that self-consciousness requires a relation to an objective world. But the practices which sustain our faith, central as they are to creatures such as us, are not strictly necessary for practical assent. Second, in the nature of the objective world to which we relate: Descartes’s world is one of material substances; Kant’s is one of causally interacting unities, substances which exist in a single public space, persist through time, and continue to exist beyond our perception of them. The world we find in the practice of holding one another accountable is a world of other thinkers, distinct from us and our activities. This is a social, not material, world.
Both shortcomings might be a cause for regret. When the children ask for bread, Michael Dummett asked above, will you give them a stone? No doubt not, John McDowell replies, but what are we to do if the children ask for the moon? The failure of Descartes’s and Kant’s arguments, together with the failure of their twentieth-century cousins, might ground a pessimistic meta-induction about future success. If we can go no further, there is nothing to mourn about falling short. It would be like the power of flight: something which would be nice to have but whose absence should not be a cause for regret.
The difference in world remains. Descartes and Kant sought a world of material bodies. We have found only a social world of other thinkers. Is this a shortcoming? Although Descartes’s argument ends with the existence of material things, the text itself ends with the Objections and Replies. Do not call them “solutions,” Descartes insists in a letter to Mersenne. “It is a good idea to call them ‘First Objections’, ‘Second Objections’, and so on, and then to put ‘Replies to the Objections’ rather than ‘Solutions’ so as to leave the reader to judge whether my replies contain solutions or not.” Descartes’s stated aim is the world of material bodies. But it is in disputation with others that he finally rests.
Why should this be a less attractive destination than a material world of objects? Mutual attention is at the root of child development. And we care more about recognition from another than we do for the rigid fidelity of things. We should not confuse the objectivity of the social world with the objectivity of the material world—that much is true. But nor should we take one as its sole exemplar. The world of other people, like the world of material objects, is not of our making.
One should say it is thinking, just as one says, it is lightning, Lichtenberg writes. To say I am thinking is already too much. Lichtenberg is right that the Cartesian and Kantian projects assume the existence of a subject who recognizes herself as the agent of her thinking. And right, too, that this is unsupported. To assume the I, to postulate it, he tells us, is a practical requirement. I have suggested that our sense of ourselves as the agent of our thinking is an article of faith. That faith is sustained by the practice of holding one another accountable for our thinking. Conversation, argument, gossip: these are the spaces where we say “I think.” They are the spaces that relate us to other thinking beings, independent of me and my activities. They are the spaces that relate us to an objective world.
“I shall be glad if people make me as many objections as possible,” Descartes writes to Mersenne, “and the strongest ones they can find.” The objections collected by Mersenne were bound together with the Meditations in a single volume, two parts of a continuous project. Descartes’s meditations end in a world of extended substances. But the Meditations itself ends with other thinkers, criticizing, cavilling, and holding to account. This is how we end our isolation: in argument, in conversation, in draughty halls filled with laughter and chatter. In philosophy.
Editor’s note: This essay is an edited extract from his book The Practical Self, forthcoming in 2024 with Oxford University Press.