I like boba tea. Every season is boba season. Every Saturday is boba day. However, every Saturday I have to make a difficult decision. I like lychee, and I also like tapioca pearls. Good news—they are almost always available in my boba shop. Bad news—they don’t go well together (in my opinion). So I have to decide what to take. How come it is so hard to decide to take lychee or tapioca if everything is predetermined, if my choice is predetermined? This time I read a book that tries to give my choice a chance to matter. It’s “The Problem of Free Will and Naturalism: Paradoxes and Kantian Solutions” by Christian Onof.

Whenever you hear physicalists say that everything is predetermined and free will doesn’t exist, they are referring to the consequence argument. The argument goes like this:

(Premise 1) If determinism is true, then our actions follow from previous events and the laws of nature.

(Premise 2) We have no control over what had happened before we were born, nor do we have control over the laws of nature.

(Premise 3) Determinism is true.

(Conclusion) Our actions are not up to us.

Accepting the conclusion will bring serious consequences, in particular, to the jurisdictional system and, in general, to how we live our lives. The Problem is therefore finding a way to accommodate free will in the realm of natural causality.

The author determines two facets of a Problem:

- The psychological account (third-person view) must explain how it would be possible to hold a person accountable for actions that are causally explained by psychology.

- The volitional account (first-person view) must show how I can regard causally determined acts as something that I can control with my will.

The book covers one resolution of the Problem that was developed by Immanuel Kant (German philosopher, 1724-1804). Here's a fun fact about Immanuel Kant: he lived a super regimented lifestyle—so much so that people in Königsberg (his hometown) supposedly set their watches by the precise times of his daily walks. He might have also drunk boba every Saturday. Or he might not have.

Kant’s theory’s main pillar is the adoption of transcendental idealism. “Transcendental” here doesn’t mean mystical. Rather, it should be understood as this: the structures of the human mind “transcend” or shape how we experience the world. Furthermore, it is “idealism” because, as Kant suggests, we can only know the world around us through these mental structures. Therefore, things and events only appear to us in a certain way shaped by our senses, while they exist as things-in-themselves in some objective reality. So the world as we experience it is, in part, a product of our mind—hence, “idealism”.

However, appearances and things in-themselves are two sides of the same coin. The relation is that the thing in-itself (or ultimate reality) grounds appearances (what we can see, measure, etc.). In other words, the thing-in-itself is what makes the appearance possible—it grounds or underlies it.

This is exactly the loophole for free will. If the nature that we perceive and observe is not the complete reality, there is a possibility for freedom outside nature but in reality in-itself. It could be that when I see the causal chain of events in another person's actions, it only appears to me so, while in complete reality (which I cannot measure) the person's actions are influenced by her freedom. Because all causality is grounded in reality in-itself, the agent may have the capacity to play a part in this grounding.

Transcendental idealism is Kant’s general framework. When concerned with free will, Kant (and the author) provides a detailed account of how it might work. Three constituents are introduced: Willkür, Wille, and Gesinnung.

- Willkür = our freedom to adopt a maxim (like, “I’ll be honest”). It’s a power of choice.

- Wille = the rational will that sets in motion our choices after we have adopted a maxim.

- Gesinnung = “supreme maxim”, whole moral configuration of a person.

The agent is free in her adoption of maxims of action. Namely, she is free to choose the principles that govern her own actions and derail from external forces, not letting impulses decide for her. Transcendental freedom underlies these choices, even if her actions appear deterministic from the outside.

It was my first encounter with transcendental idealism, and it is bittersweet. Transcendental idealism does sound appealing and plausible. Of course, we don’t see reality as it is. Dogs perceive the world apparently very differently from us, and bats perceive the world in a way we cannot even imagine. Also, I can agree with the fact that we are limited in our knowledge due to the way we are (but I would prefer to be optimistic). However, transcendental idealism doesn’t provide a satisfying answer. It denies that we can even in principle discover what reality in-itself is. From that it follows that we cannot know whether we have free will. The absence of acceptable empirical assessment renders transcendental idealism as mystical and not scientific (in my opinion). Perhaps it will change one day.

The book includes many formal assessments of critique and solutions concerning freedom in transcendental idealism. Even if you don’t believe in transcendental idealism, there is good coverage of alternative approaches to the Problem. Maybe you may even find a point of view you can agree with. I didn’t find. Moreover, after finishing the book, my hard decision to pick lychee or tapioca didn’t become any easier.

Interesting argument against observables:

“(Premise 1) Any observation of the totality of objects/events requires that the observer (with instruments) be located beyond this totality in space/time.

(Premise 2) There are no possible locations for a putative observer of the totality of a finite world. [what is observed is no longer the totality; a further observer would need to observe the first observer]

(Premise 3) There are no possible locations for a putative observer of the totality of an infinite world.

(Premise 4) The world as totality is not observable.

(Premise 5) There could be non-natural objects/events.

(Conclusion) Naturalism is false.”

June, 2025