Could someone explain Kant's view on free will to me?

As I understand it, Kant says that it is necessary that we have free will, because otherwise it would be unreasonable to hold people responsible. We hold people responsible for things, even though their decisions appear to us to be deterministic and we do not hold them responsible for the circumstances that led to their decision, but we hold them responsible for having made this decision. This would only be reasonable if their decision was not determined by the circumstances (and the laws of physics etc.) but rather by human reason, uninfluenced by any empirical conditions. Therefore, the reason must be part of the noumenal world, from where it is able to cause our behaviour.

However, I think I am understanding this wrong, because I don't see how this implies free will or why we would hold people responsible. Kant's reasoning seems to be:

We hold people responsible->We should hold people responsible->Holding people responsible would not make sense without free will (free from laws of physics/empirical influences)->There must be free will in the noumenal world.

However, I think I am misinterpreting him, because that first implication does not make sense to me.

Could someone help me?


2 Answers 2


First off, I want to say this is a really good question that reflects real thought on an interpretative issue in Kant studies. Second, I think you're grasping some major things but also thinking backwards (by which I mean imposing contemporary categories on what Kant is doing).

In terms of your question, one major issue is going to be where in Kant you are reading. The account you are giving sounds closest to the Critique of Pure Reason (Kritik der reinen Vernunft). There, I take Kant's position to be as follows:

  1. As far as empirical science is concerned, our actions are determined
  2. Moral responsibility requires our actions to be free

This is what Kant calls the "Third Antinomy" (or perhaps that is just what contemporary Kant scholars call it).

But there are some specific details where I think you may be deviating rather sharply from Kant (or at least wording them in some significantly different ways). The background for this is that Kant's account of human knowledge is predicated on the belief that our knowledge is limited to objects (those things we place under the forms of sensibility and categories of the understanding). To put it another way, Kant is a skeptic about certain types of knowledge claims, because he thinks we never encounter things in an unencumbered way. But the manner in which we relate to objects is called "understanding."

It turns out that our actions are renderable both through our understanding which places things under the categories of cause and necessity (to name the most relevant two). But our actions take place for ourselves under the auspices of reason which is a different faculty. The easiest way to say this is that I understand your actions, but I reason to my actions on Kant's account.

So to return to the details in your question, you state:

Therefore, the reason must be part of the noumenal world, from where it is able to cause our behavior.

but at least on my reading of Kant, there's a minor error here. Viz., Kant does not see this as a conclusion but rather as a premise that relates to what reasoning is. But if you're just reading the Critique of Pure Reason as your source, then you might come up with this conclusion. The problem is that Kant changes how this works between the 1st Critique, the Groundwork, the 2nd critique, the Metaphysics of Moral, and Religion. Each of them gives a slightly different explanation of how we are morally/noumenally free yet empirically determined.

You're correct to think this progression is wrong:

We hold people responsible->We should hold people responsible->Holding people responsible would not make sense without free will (free from laws of physics/empirical influences)->There must be free will in the noumenal world.

No Kant work states this.

There's two important features in Kant's ethics that do seem similar. I take it that Kant asserts (at least outside the Groundwork this becomes a faktum of reason) we have reason (Vernunft) and this means that we can choose our own actions -- in fact, we can choose them despite our empirical natures. Thus, we are responsible for our actions, and thus, we are not determined.

A better but related argument has to do with Kant's belief that the world is ultimately just / what is called in the literature Kant's proportionality principle/thesis. Kant, for reasons that are nearly opaque, asserts in his moral philosophy that the world will ultimately turn out to be just. But this is contrary to empirical evidence about this world (evil people die happy; good people die sad; etc). Thus, Kant asserts a God brings about just desserts in an afterlife.

Depending on which work you are looking at in Kant's moral philosophy, the basis of the claim that we are free is different.

Critique of Pure Reason -> An antinomy shows that we cannot know we are free, but we live that we are free. Reason is free and distinct from understanding. Groundwork -> Part III is an argument that we are somehow free (not well-accepted or understood) Critique of [Pure] Practical Reason -> we are understood to be free as a fact of reason. Metaphysics of Moral: Doctrine of Virtue -> we are free as a condition for a just world. Religion within the Bounds of Reason Alone -> we are free at least insofar as we make a sort of root decision that impacts our later ability to be free.

Notably, it moves from an argument in the first two works to an assertion in the latter three.

  • Thank you for the detailed response! I am reading a translated extract from the Critique of Practical Reason. If I understand your response correctly, Kant begins with the assumption that the reason exists in the noumenal world and from there he reasons that we must be free (so the opposite of what I thought). So, basically the reason controls us and is free from cause and effect/any influences at all and therefore we are free as well? And our freedom would then imply complete responsibility, right?
    – Dasherman
    Mar 10, 2015 at 14:00
  • I don't quite understand the expression "reason controls us". We have both a faculty of understanding limited by the apparatus (i.e., forms of sensibility and categories of understanding) and a faculty of reason that basically does logic regardless of anything else. Choosing our actions is a function of logic. / We are free insofar as our actions arise from reason -- which it turns out is noumenal rather than phenomenological (and importantly does not suffer from the limitations of understanding).
    – virmaior
    Mar 10, 2015 at 14:39
  • Would it be better to say that reason determines what we do or choose to do? I assume the latter, right?
    – Dasherman
    Mar 10, 2015 at 14:49
  • We are the beings that reason according to Kant ... ideally, we choose our actions in accord with reason. When we do so, we act autonomously and morally. When we don't we let something else determine our morality.
    – virmaior
    Mar 10, 2015 at 15:04
  • So according to Kant we can either follow what reason dictates or something else? Is that a choice we have, or is that simply something determined by circumstances?
    – Dasherman
    Mar 10, 2015 at 15:21

Good question. Maybe if Kant would had access to our knowledge today would he would update his thought. He was a genius, no doubt. It seems to me that noumena [(in the sense they are the (supposed/.../presumed) things themselves, which we in lack of better concept so far, we say: reality] impose certain constraints to life which has got thinking ability and life with thinking ability responds to such constraints with a set of heuristics which moral/ethics/.../"what works" are examples and maybe are just byproducts of thinking ability.

So, in the end if we eliminate free will we have to start to find another heuristic to deal with it and moral & cia are not enough for it.

Now, with AI (artificial intelligence, which I prefer to name artificial integration) we just lost ANOTHER degree of freedom... so it is about time to think about the reduction of freedom implies in terms of the "next" heuristic...

After all we are just algorithms producing biochemistry and bio-electrical signs. See for example the book of the picture... ☺

enter image description here

  • This does not answer the question and follows exactly the reductionistic reasoning Kant opposed in his writings "to make room for faith [and free will]".
    – Philip Klöcking
    May 2, 2018 at 7:50
  • Thank you for the comment. Can you be more incisive, please, dear @PhilipKlöcking?
    – KwanzaKymi
    May 23, 2018 at 4:22

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .