Descartes to Kant: The Evolution of the Metaphysical Self

The question “Who am I?” is one of the oldest in philosophy-but it wasn’t until the early modern period that thinkers placed the individual self at the very centre of philosophical inquiry. Between the 17th and 18th centuries, three major philosophers-René Descartes, David Hume, and Immanuel Kant-offered radically different answers to this question. Together, their ideas trace an evolution from the confident, thinking subject to a self that is fragmented and uncertain, and finally to one that actively structures the world it experiences. Understanding how the concept of the metaphysical self developed across these three thinkers is essential for grasping many of the debates in philosophy of mind, personal identity, and consciousness that continue today.

Descartes’ cogito and the birth of the modern subject 🔗

René Descartes (1596-1650) is widely regarded as the founder of modern philosophy. His project began with a radical decision: to doubt everything that could possibly be doubted, and to see what, if anything, survived. This approach-known as methodic doubt-stripped away all assumptions about the senses, the external world, and even mathematics. Descartes even entertained the possibility that an all-powerful deceiver could be manipulating his every thought.

Yet at the very bottom of this process, one thing proved impossible to doubt: the act of thinking itself. If I am doubting, then I am thinking. And if I am thinking, then I must exist. This is the famous cogito ergo sum-“I think, therefore I am.” The very attempt to think away one’s own thinking is self-defeating. As Descartes wrote in the Second Replies, this is not a conclusion derived from a syllogism but something recognised as self-evident through a direct intuition of the mind.

The thinking self as foundation of knowledge 🔗

What made the cogito revolutionary was not just its logical elegance but the philosophical shift it represented. Before Descartes, medieval philosophy typically grounded knowledge in God, revelation, or the natural world. Descartes reversed this. He made the individual thinking mind the starting point-the one secure foundation from which all other knowledge could be rebuilt. This is what philosophers call the “modern subject.”

For Descartes, the self that the cogito reveals is essentially a thinking thing (res cogitans)-a substance whose entire nature consists in thought. The core of selfhood, as he saw it, lies in the mind’s capacity to question, understand, affirm, deny, and imagine. This marks a decisive move: the self no longer needed external validation for its existence. It validated itself through its own rational activity.

The mind-body problem 🔗

Descartes’ emphasis on the thinking self led directly to one of the most enduring puzzles in philosophy: the mind-body problem. He argued that the mind (res cogitans) and the body (res extensa, or extended substance) are fundamentally different kinds of things. The mind is non-physical, indivisible, and constitutes the essence of who we are. The body is physical, divisible, and operates as a kind of complex machine.

This raises an obvious difficulty: if mind and body are such different substances, how do they interact? When you decide to raise your hand-a mental event-how does that cause a physical movement? Descartes never fully resolved this puzzle. He proposed the pineal gland as the point of interaction, but this answer satisfied very few of his contemporaries or successors. The problem of Cartesian dualism remains a live issue in philosophy of mind to this day.

Criticisms of the Cartesian self 🔗

Despite its profound influence, the cogito attracted criticism from the beginning. Some philosophers argued that the move from “thinking is occurring” to “I am a thinking substance” assumes more than the argument actually establishes. Antoine Arnauld, a contemporary of Descartes, pointed out potential circularity in the reasoning. Others questioned whether the “I” in “I think” really refers to a substantial, unified entity-or whether it simply marks the occurrence of thought without telling us anything about what the thinker is.

These objections would find their most powerful voice in the work of David Hume, who directly challenged the idea of a stable, unified self.

Hume’s bundle theory: the self dissolved 🔗

David Hume (1711-1776) approached the question of the self from the opposite direction. Where Descartes used reason to arrive at certainty, Hume insisted that all our ideas must ultimately be traced back to sense experience. And when he turned his empiricist lens inward-looking for this supposed “self” through introspection-he came up empty.

In his A Treatise of Human Nature (1739), Hume reported that whenever he tried to catch himself, all he ever found were particular perceptions: a feeling of warmth, an image, a thought, an emotion. He never encountered anything that could be called a stable, underlying “self” behind or beneath these experiences. This led him to a striking conclusion: there is no such thing as a unified, enduring self. What we call the self is nothing more than a bundle of perceptions in constant flux.

What the bundle theory claims 🔗

Hume’s bundle theory holds that what we think of as a personal identity is really just a collection of different experiences-impressions and ideas-that follow one after another with remarkable speed. There is no deeper substance or soul holding them together. As Hume put it, looking for a single, unified self beyond those perceptions is like looking for a chain apart from the links that make it up.

So why do we believe in a continuous self at all? Hume’s answer is psychological. We have a natural tendency to attribute identity to collections of closely related things. We call a mature oak the “same” tree as the sapling, even though almost every part has changed. Similarly, the resemblance, succession, and causal relations between our perceptions create an illusion of a single, persisting subject. Memory plays a key role here-it links past experiences to present ones and creates the feeling of continuity. But for Hume, this feeling is precisely that: a feeling, not a fact about the nature of reality.

Consequences and difficulties 🔗

Hume’s position was radical and deeply unsettling-even to Hume himself. In the Appendix to the Treatise, he acknowledged that the relations binding together the bundle of perceptions are not “real” in a robust sense; each perception is a distinct existence that could, in principle, exist independently of every other. He struggled to explain what actually holds the bundle together and admitted he could not fully resolve the difficulty.

The bundle view also posed serious challenges for ethics and practical life. If there is no stable self, what grounds moral responsibility? Why should I care about my future if the “I” of tomorrow is a different bundle of perceptions? These questions foreshadowed debates in 20th-century philosophy, particularly in the work of Derek Parfit, who drew similar conclusions and explored their implications at length.

Hume’s denial of a substantial self created what many saw as a philosophical crisis. If neither the senses nor reason can establish that there is a unified subject of experience, how is coherent experience possible at all? This was the challenge that Immanuel Kant set out to address.

Kant’s transcendental self: a new synthesis 🔗

Immanuel Kant (1724-1804) famously said that it was Hume who woke him from his “dogmatic slumber.” Kant recognised the force of Hume’s arguments but refused to accept their sceptical conclusions. His response, developed most fully in the Critique of Pure Reason (1781), was to rethink the relationship between the mind and experience from the ground up.

Kant agreed with the empiricists that all knowledge begins with experience. But he insisted that experience itself has a structure that cannot come from the senses alone. The mind does not passively receive a stream of impressions; it actively organises sensory data according to built-in frameworks-the forms of space and time, and a set of fundamental concepts called the categories (such as causation, substance, and unity). Without these organising principles, raw sensory data would be nothing more than what Kant called a chaotic “manifold”-a formless mass of impressions with no coherence.

The transcendental unity of apperception 🔗

At the heart of Kant’s theory of the self is the concept of the transcendental unity of apperception. This is a technical term, but the idea behind it is essential: for any experience to be my experience, it must be possible for me to attach the thought “I think” to it. All of my representations-every perception, memory, and thought-must be capable of being recognised as belonging to a single, identical subject. This self-consciousness is not something I learn from experience; it is the very condition that makes experience possible in the first place.

Kant calls this the “I think” that must be able to accompany all representations. It is “pure” or “original” apperception because it does not come from the senses-it is an act of the understanding, a spontaneous activity of the mind. The unity it provides is what allows us to combine scattered sensory impressions into the unified experience of an object, a situation, or a world.

How Kant differs from both Descartes and Hume 🔗

Kant’s transcendental self represents a careful middle path between the Cartesian self and the Humean dissolution of the self:

Against Descartes: Kant denied that we can know the self as a substance. The “I think” is not a window into the soul. Descartes believed that through the cogito we could gain knowledge of the mind as an indivisible, immortal substance. Kant argued this was a mistake-a paralogism, or faulty inference. The unity of the “I think” is purely formal and logical; it tells us nothing about the nature of the thinking subject. We cannot conclude from the fact that consciousness is unified that there exists a simple, permanent soul. As Kant put it, the “I” is the background of the field of experience, not an object of knowledge.

Against Hume: Kant argued that Hume was right to say that introspection reveals no impression of a stable self, but wrong to conclude that no such unity exists. The unity of the self is not something found in experience-it is the precondition of experience. Without a unified subject, the stream of perceptions would never coalesce into coherent experience. Hume was looking for the self in the wrong place: among the contents of experience, rather than in the structure that makes experience possible.

The self as an active organiser 🔗

One of Kant’s most important contributions is the idea that the self is not a passive recipient of data but an active agent in the construction of experience. The process by which the mind combines the manifold of intuition into unified representations is what Kant calls synthesis. Synthesis operates at multiple levels: we apprehend sensory data in space and time, reproduce past impressions through imagination, and recognise patterns through the application of concepts.

This activity of synthesis is governed by the categories of the understanding-concepts like causation, substance, and quantity that structure every possible experience. For Kant, these categories are not learned from experience; they are brought to experience by the mind itself. The transcendental self is thus the ultimate source of the order and lawfulness we find in the natural world-not because it creates objects, but because it determines the form in which objects can appear to us.

Limits of self-knowledge 🔗

Kant drew a sharp distinction between the empirical self-the self as we experience it through introspection, in time, subject to change-and the transcendental self, which is the formal condition of all experience but is itself never an object of direct awareness. We can know what we appear to be (empirical self) but never what we are in ourselves (transcendental self). This echoes, in a transformed way, Hume’s insight that the self eludes direct observation-but Kant turns this limitation into a philosophical principle rather than a source of scepticism.

From Descartes to Kant: what changed and why it matters 🔗

The journey from Descartes to Kant represents one of the most consequential arcs in Western philosophy. Descartes gave us the confident, thinking subject: a self that knows itself through reason and stands as the foundation of all knowledge. Hume dismantled this subject, showing that introspection reveals only a stream of disconnected perceptions with no underlying unity. Kant then reconstructed the self-not as a knowable substance, but as a necessary formal condition of having any experience at all.

Each position carries practical implications. If Descartes is right, the self is fundamentally stable, rational, and separable from the body-a view that has influenced everything from religious conceptions of the soul to legal notions of individual responsibility. If Hume is right, personal identity is far more fluid and constructed than we ordinarily assume-an insight that resonates with contemporary findings in psychology and neuroscience about the fragmented, narrative nature of the self. And if Kant is right, the self is neither a discoverable object nor an illusion, but the indispensable framework through which we engage with reality.

Modern cognitive science has found echoes of all three views. Like Descartes, researchers recognise the special status of first-person consciousness. Like Hume, they observe that the self is more constructed and context-dependent than common sense suggests. And like Kant, they understand that some form of unified processing seems necessary for coherent experience. The question of the metaphysical self, far from being settled, remains at the frontier of philosophy and science alike.

What do you think? Is the self a stable entity that we discover through reflection, or a construction our minds generate to make sense of a constantly shifting stream of experience? And if Kant is right that the self is a necessary condition of experience rather than something we can directly know-does that make the question “Who am I?” ultimately unanswerable?

References

1. https://plato.stanford.edu/entries/descartes-epistemology/
2. https://en.wikipedia.org/wiki/Cartesian_Self
3. https://www.britannica.com/topic/bundle-theory
4. https://academic.oup.com/book/3488/chapter/144677509
5. https://plato.stanford.edu/entries/kant-mind/
6. https://plato.stanford.edu/entries/kant-transcendental/
7. https://link.springer.com/article/10.1007/s11466-007-0016-0