The soul, God, and the limits of proof
⏱ 8 min read · ⏳ active
Two of the oldest promises philosophy ever made: to prove that your soul is immortal, and to prove that God exists. Kant admired both projects, took them apart bolt by bolt — and concluded that pure reason can deliver neither. Not because the conclusions are false, but because the proofs don’t work. Here is the anatomy of two failures.
Two ideas remain from our set of three. Last lesson handled the world-whole (the antinomies). Now the SOUL — the object of rational psychology — and GOD — the object of rational theology. In both, traditional metaphysics claims to reach a real super-sensible object by pure reasoning alone. In both, Kant shows the reasoning quietly cheats.
Begin with the soul. Rational psychology has exactly one starting point: the “I think” that can accompany all my representations. Every thought I have is mine; there is a single “I” running through them. From this thin thread, the metaphysician spins a whole doctrine: the soul is a substance, it is simple (has no parts), it is permanent — and therefore it is immortal Prolegomena to Any Future Metaphysics §46.
Kant calls this reasoning a paralogism — a fallacy, a piece of reasoning that looks valid but is not (from the Greek for “false reasoning”). Where is the cheat? In treating the “I” as if it were an object I intuit, when it is only the logical subject of thought — the formal “who” that must be presupposed whenever anything is thought at all. The “I think” is a condition of thinking, not a thing that is thought about Prolegomena to Any Future Metaphysics §47.
Look closely at the word substance. To apply the category of substance and get knowledge, Kant has argued all along, you need a corresponding intuition — something given in experience that persists. But the “I” is given in no intuition at all; it is the bare form of self-consciousness. So calling it a “substance” borrows a category and pays nothing for it. The soul’s substantiality and permanence — hence its immortality — cannot be proved by pure reason Prolegomena to Any Future Metaphysics §48.
Why does Kant call rational psychology's inference from 'I think' to 'the soul is an immortal substance' a PARALOGISM?
Now God, and rational theology. Here reason reaches for the third idea: the unconditioned ground of all being — the most real being, the being that grounds everything and depends on nothing Prolegomena to Any Future Metaphysics §55. The tradition offered three great proofs, and Kant sorts them into three named types.
The ontological proof argues from the mere concept of the most perfect being to its existence (a perfect being that lacked existence would not be perfect). The cosmological proof argues from the bare fact that something contingent exists to a necessary being as its ground. The physico-theological proof (the argument from design) argues from the order and purposiveness of nature to a wise author of it. Three routes, one destination: a super-sensible God.
Kant’s objection is a single, structural one, and it is the same lesson as the paralogism and the antinomies. Every one of these proofs tries to use the categories — existence, cause, necessity — to reach an object lying entirely outside possible experience. But the categories yield knowledge only where an intuition can be given, i.e. within experience. Applied beyond it, they generate no cognition, only the illusion of one. So all three proofs fail as proofs — they overstep the one domain where their tools work Prolegomena to Any Future Metaphysics §55. Kant works the three through in detail in the The Critique of Pure Reason Transcendental Dialectic.
Collect the three verdicts and you have Kant’s headline negative result. Metaphysics as a would-be science of super-sensible objects — a body of knowledge about the soul, the world-whole, or God as they are in themselves — is impossible. Rational psychology, cosmology, and theology, the three departments we mapped in lesson 1, each promise cognition of an object that no possible experience can supply. On that promise, all three default.
And notice what Kant is not saying about the thing-in-itself, because students trip here constantly. He does not say we cannot know that a super-sensible reality exists — the very notion of an appearance implies something that appears. What he denies is that we can know what it is — its inner nature, whether it is one soul or many, a personal God or not. Existence-as-the-correlate-of-appearance: admissible. Determinate knowledge of the super-sensible: impossible.
What is Kant's core objection, shared by the ontological, cosmological, and physico-theological proofs of God?