Third Antinomy (On Freedom)
Background: The Antinomies in General
(Terminological note: Kant doesn't speak of "antinomies," just of "the antinomy of pure reason," which divides into four "conflicts." But most commentators describe the four "conflicts" as "antinomies," and I will follow this usage.)
The antinomies were what first led Kant to formulate the doctrine of transcendental idealism. (This is the idea that the world we know is "empirically real" but "transcendentally ideal": we know only "appearances," which are "objective" in the sense of being the same for all humans, but nevertheless are not things in themselves. Only things in themselves could be "transcendentally real.")
Kant thinks that humans are compelled by the nature of our reason to reach contradictory conclusions about such things as the size and temporal extent of the universe (i.e. the empirical world).
The reason is that we are compelled always to seek the "conditions" of what we perceive.
In the antinomies, Kant distinguishes between "condition" and "conditioned." This is meant to be a very general distinction. Examples include:
causes are conditions of their effects (which are therefore "conditioned")
earlier times are conditions of later times
larger spaces are conditions of their smaller parts
The idea is that, beginning from what we directly experience, we are driven to seek its "conditions": to find the causes of the things we experience, to learn about the history that led to the present, and to investigate further and further into the reaches of space.
No matter how far we have managed to extend our knowledge, we always want to push it further. We want, in fact, to push it back all the way, and this raises the question: how far is all the way? For example, does time stretch back infinitely far, so that there is an infinity of past times, or is there a first time?
Example: First Antinomy
Kant's strategy in the antinomies is to offer what he takes to be genuine proofs of two apparently contradictory propositions, then to show that both proofs make the same illegitimate assumption.
For example, the first antinomy has to do with whether the world has a beginning in time. (Actually it concerns both time and space, but let's focus on time.)
Thesis: the world has a beginning in time.
Antithesis: the world is infinite with regard to time.
The proofs of both the thesis and the antithesis are by reductio ad absurdum. The proof that there is a beginning starts by assuming that past history extends infinitely far, and argues that this leads to a contradiction (since an infinite series supposedly cannot be completed). The proof that the world is infinite starts by assuming that there is a first time, and tries to show that this leads to a contradiction (since it seems that we can coherently ask what happened before the first time).
The mistake, in Kant's view, lies in thinking that "past history is infinite" and "there is a first time" exhaust the possibilities. Kant thinks that there is also a third possibility:
1. The world is a thing-in-itself and is infinite
2. The world is a thing-in-itself and has a first time
3. The world is only appearance
If the world were a thing in itself, we would be led into a contradiction by the equally good arguments for 1 and 2. But this only shows that 3 is the correct option: the world is not a thing in itself, but merely an appearance. Thought of this way, it is consistent to say both (1) there is no first time, since no matter how far back we push our knowledge, we can always push it back further; (2) past time is not infinite, because no matter how far we push back our knowledge, we will have pushed it only finitely far.
Third Antinomy: Kant on Freedom
Kant seems to have three different but related conceptions of freedom.
I am free if and only if:
1. My actions are brought about by reason rather than by desires or urges ("impulses of sensibility": A534/B562). This is what he calls the "practical sense" of freedom (A534/B562).
iff 2. I originate my actions (I am a "first beginning"), rather than just being a middle link (a "subordinate beginning"). This is the "cosmological sense" of freedom (A533/B561).
iff 3. I do things not because of what is but because of what ought to be (A547/B575).
These three accounts of freedom are linked by two theses.
1 is connected with 2 by the thesis that reason is noumenal, not phenomenal. Thus if reason gets me to do something, the action is noumenally produced; but time doesn't apply to the noumenal realm, so there is no earlier event which causes this. Kant is explicit about this thesis at A553/B581: "For since reason itself is not an appearance and is not subject at all to any conditions of sensibility, no temporal sequence takes place in it even as to its causality, and thus the dynamical law of nature, which determines the temporal sequence according to rules, cannot be applied to it."
1 is connected with 3 by the thesis that reason tells us what ought to be the case (rather than what is actually the case). Sensibility and understanding combine to give us knowledge of the empirical world, i.e. to tell us what actually is the case. But reason (in its practical rather than its theoretical use) gives us imperatives, tells us what we ought to do (A547/B575).
It is definition 2, the idea of the agent as a "first beginning," that seems to create a conflict with the law of causality that Kant defends in the Second Analogy. The world of appearances is completely causally determined, on his view: every event has a sufficient cause. To the extent that the agent is part of the empirical world (and the empirical self is part of the empirical world, an appearance), agents are just as subject to the law of causality as any other part of nature: "all the actions of the human being in appearance are determined in accord with the order of nature by his empirical character and the other cooperating causes" (A549/B577).
Kant's solution appears to be that we are (or may be) not merely appearances but also noumena. Although the empirical self is completely causally determined, the noumenal self which underlies it is not subject to time or causality, so it can produce the action without itself being caused by anything (and thus it's a "first beginning" of action).
So we have a strange kind of compatibilism. The empirical self is just one more part of the causal stream, and all its actions are completely determined. But it is (or may be) a manifestation of the noumenal self, which isn't bound by the law of causality. "In its empirical character, this subject, as appearance, would thus be subject to the causal connection, in accordance with all the laws of determination . . . but in its intelligible character . . . this subject would have to be declared free of all influences of sensibility and determination by appearances. . . . Thus freedom and nature, each in its full significance, would both be found in the same actions, simultaneously and without any contradiction, according to whether one compares them with their intelligible or their sensible cause" (A540-540/B568-569).