# Kant on causation

I want to start off by saying that I'm not a philosophy major and haven't read any texts. All my understanding is from Wikipedia, Stanford website and Reddit. Please don't hold that against me.

I have a question on causation. I'll illustrate my question using the following example. I won't be talking about the noumenal world in this question.

Say my experience just consists of me, and two oppositely charged particles. That's the universe that I perceive. We see that they repel according to Coulomb's law and can predict what may happen. So we might be tempted to say that charge A CAUSES charge B to move. But it is possible that it's not charge A that caused B to move, but B spontaneously for no reason started to move as it is (no cause). Or some dark energy luckily pushed B for no reason at that exact time. So moving of charge B might not be attributed by A. Or motion of B might not even have a cause since causality is an illusion. I think this is what Hume says..?

Now I don't understand how Kant improved on this. Can someone explain why Kant still talks about causation (albeit in the phenomenal world)? Why is causation necessary for seeing the world (data obtained)?

To further illustrate the point, consider the universe which I experience just to contain one charged particle. I can conceive that the charge moves randomly/arbitrarily without any sense. I get it's position x(t) as a function of time. x(t) was a random path. It could've been anything else and there's no law that can predict it or even give a probability distribution. What does causation even mean here?

After some comments below, I wanted to also ask if Kant concept/framework of Causation is a theory (out of multiple theories) rather than absolute truth? It seems that all there is in the world is things happening, so concept of Causation doesn't make sense. So Hume's ideas seem like absolute truth whereas Kant's ideas seems like a theory.

Reminder that I'm always talking about phenomenal world and not things in themselves.

• It helps to keep two things in mind here: Firstly, Kant's thought is built upon our common everyday experience and tries to inquire the conditions of its possibility. Not some obscure thought experiments. Secondly, causation is one of the categories through which we perceive. Nowhere it is said that every object of experience has to be understood through all categories. Does this help your inquiry? – Philip Klöcking Oct 17 '20 at 18:33
• Kant doesn't deny that we can have sensations unbound in themselves by causality. But he says we would have no criteria of hallucinations vs. veridical perceptions sans the framework of substance, causality, etc. – Kristian Berry Oct 17 '20 at 18:47
• @PhilipKlöcking Though my example is a thought experiment for now, it may be possible to observe this randomness of motion of electron experimentally in the future. If such a process is indeed observed, it helps us say that lawlessness of the universe is not out of question, and if we believe in the idea that everything we perceive is just coincidental patterns due to arbitrary motion/effect of subatomic particles then is there any meaning to the idea of causation? – Razor Oct 18 '20 at 8:01
• @KristianBerry I think I need to read a bit more on what is Kant is saying. I'm not comfortable with the existence of the concept of causality. I don't see its necessity in first place as I've explained in a comment above – Razor Oct 18 '20 at 8:08
• Letting Kant beside, the scientific understanding of causation involves at least a statistically significant pattern and a falsifiable mechanism of causation between two observable events. Other than that, observations are simply independent events. I don't really know how one could possibly try to get rid of a notion that's the basis of modern science. What would you call "repeatable, lawful pattern between two observable events" instead? What is gained by getting rid of the term "causation" just because there are events where it is hardly applicable? – Philip Klöcking Oct 18 '20 at 16:37

## 2 Answers

A complete answer would need to explain just about all of Kant. First, I would say that Kant's ideas of mechanical causality were much disputed in his day and are outmoded now, in part due to quantum effects, as you note. Second, I should also note that Kant was doing epistemology, how we know things, what we can know. He concluded, reluctantly, that we cannot know "things in themselves" or noumena, only things as we apprehend them via phenomena. So it makes no sense in a Kantian context to talk about "absolutely true" or to describe noumenal things "behind" the phenomena.

So, causality. This was perhaps the pivotal issue that aroused Kant to contest Hume's account. Hume argued that we do not observe and cannot prove some essential relation called "causality." We only see A following B or "constant conjunction." How do we differentiate this from coincidence? Only probabilistically, by keeping count and by "custom." That is, by a kind of habituation. Probably most scientists today are more on Hume's side about this.

What was Kant's "improvement"? Since Hume was not introducing probability, his argument ended, in Kant's view, in a deep skepticism, casting doubt on the true advances of Galileo or Newton, a veritable catastrophe for an Enlightenment thinker. If our "sense" of causality itself is uncertain, how can we claim any true scientific knowledge or universal laws?

His response was, in effect, to point out that we certainly seem to observe causality! His critique then painstakingly argues that in addition to the "sense data" of science and empiricism, our observations also have a certain order, a matrix of relations that is always there, and through which we always view the world. People often liken this to a pair of spectacles we can never take off, the "lenses" being our sensory apparatus.

He called these necessary relations "categories," after Aristotle and "causality" is one of these. These categories are a priori in that we do not perceive them as we do phenomena, yet they are "in" experience and we could not even have coherent experiences without them. He carefully infers these necessary categories by his "transcendental deduction." These are relations that must be there if we are able to experience things as we do. Needless to say, this is a very difficult, ingenious, and debatable demonstration.

It may not have been an "improvement" on Hume's causality, but it was great and very influential merging of the positions of empiricism and idealism. It reasons that human experience will always contain a certain a priori structure, not unlike Chomsky's "deep grammar" underlying language or other types of structuralism. We may not be able to grasp "absolute" metaphysical truths in the manner of the older idealists, but we can be certain of universal structures for all human beings or rational beings.

Causality is just something we all embed into our experience if we are to have coherent experiences at all, like the rules of language. Causality is necessary to any "sense" of time, in which things move in one direction but not backwards. His aim, or one of many, was to preserve the necessity of validating physics empirically by means of experience, yet secure it against skepticism by means of universal a priori categories, such as causality.

This was a major advance amidst the arguments of his day, a basis for universal scientific certainties with no appeal to God or metaphysics, though not one most would adhere to today, at least in the description of causality. I'm sure others can improve on this simplified answer. I have left your examples out of it, because I didn't think they were helpful. But, yes, the subatomic world has pretty much swept away these older ideas of "efficient causality," though not the Kantian approach.

• Very concisely efficient and adequately replicated rendering of Kant's Copernican revolution. Thank you Nelson. The SEP should trumpet this answer. Would enjoy hearing your thoughts on Yirmiyahu Yovel' s book; "Kant's Philosophical Revolution". It's a parsing or unbundling of the CPR. Cheers, – user37981 Oct 20 '20 at 12:42
• Thanks, I'll check out the book, though my unread stack is getting tall as Kansas corn. – Nelson Alexander Oct 20 '20 at 15:01
• I would add that casualty is a necessary part of a more profound assumption we must make in order to have freedom and the very concept of truth. Namely, we must assume that we a) share (and are a part of) the same objective reality, and b) that reality is explainable through logic and reason <== and causality enables explainability – Yuri Alexandrovich Oct 21 '20 at 17:09
• @Yuri. I suspect Kant, as a preeminent Enlightenment thinker, would agree with that. Like others of his day, he was deeply troubled by the secular, skeptical direction of someone like Hume's thinking and wanted to give a grounding for universal morality quite apart from religion. His "causality" is actually dualistic, for in addition to "natural causal laws" he describes human freedom as a second type of "causality," enabling moral choice alongside the mechanical causality of physics. – Nelson Alexander Oct 21 '20 at 17:26
• Yes, accepting the dualistic nature of casualty, or even of the truth itself (moral vs scientific) is fairly common among modern philosophers... I don't think it accurate though. It appears this way because though deriving moral truths from scientific ones is certainly possible, the process is not very straightforward. OK, eventually it is, but only after one achieves a pretty deep understanding of oneself, others, their life, and the world around. Basically one has to know what they are doing... And that's why knowledge is the only true virtue. – Yuri Alexandrovich Oct 21 '20 at 18:22

This question is multi-faceted and I'll try to answer the aspects as shortly and succinct as possible. I think the main problem you are facing is that you in some sense did not quite fully understand and follow the Humean insights here.

### 1. Hume and Kant

Hume put the position of empiricism - that knowledge can only be acquired via experience - to the extreme. He consequently got rid of all metaphysical and alleged "certain" knowledge of old and told us that there is nothing the like and that we will never be able to know anything about the real world for certain. All there is are habitual patterns of perception and how things "really are" is simply beyond us.

Kant did not exactly like that outcome. He studied Leibnizian metaphysics which basically were the pinnacle of the attempts to make classical metaphysics coherent (without losing God, that is) for most of his life. Thus, he strived to secure a place for certain metaphysical knowledge, ie. knowledge about the structure of ultimate Reality.

His ingenious approach was saying that to some extent, Hume was right to question metaphysical knowledge - insofar it is acquired directly via (sensual) experience. But if we take a look at the fact that we humans in fact are capable of sharing experience of a common world, there has to be something enabling us to do so. Some common mechanism of perception and thought, if you like. To uncover these patterns is what his philosophy is all about.

### 2. Kantian causality

As you might surmise by now, the thinking in the question has one major flaw: It assumes the metaphysical being and actuality of scientific insight, ie. that science would tell us something about metaphysics. That is not quite in line with Hume (or more elaborate recent philosophies), nor the methodological foundation of science itself: Whatever the status of scientific knowledge is, no matter how intricate and deep the insights are, it can and, as a matter of fact, and by all means, will be replaced by new, more intricate and deep insights and theories.

In other words: Our current theories about the random movement of quantum scale objects may be what fits best our past observations, but this does not mean that we can actually rule our that they follow what we call "causality" or that they are only a placeholder for underlying structures yet to be unveiled.

Now, Kant's causality is nothing but the way we perceive the world. But at the same time, it is supposed to be more than a mere "habitual pattern" or contingent theory. It is so basic that if it wasn't for causality and the other categories, we weren't even able to coherently form an experience of things in the world (his transcendental deduction tries to justify this claim).

As much as this approach was revolutionary, there is still ongoing research trying to entangle hard-wired perceptual patterns from culturally ingrained ones. Most of philosophy agrees that while Kant himself thought his theory to be penultimate, there is cultural and historical plasticity to a priori concepts.

### 3. Conclusion

Randomness is not at odds with Kantian causality. It actually is a placeholder for "we cannot make sense of it since it does not (seem) to follow causality, which we need to form a coherent experience and make sense of things". That is why we seek to find patterns in all this randomness, to make a "thing" out of objects which may be better described as fields or waves or strings or something entirely different which eludes our ability of intuition. Realistically speaking, the idea to visualise the path of an electron as singular object moving around actually runs against some basic principles of quantum mechanics.

So we basically have a twofold misunderstanding here: Firstly, you mistook our perceptions, or, more correctly, visualizations and rationalisations, of the quantum objects to be exactly how things actually are. And then, secondly, you mistook causality to be a universal property of Reality itself, which it actually isn't according to Kant, and followed that since things at the quantum level "really are" random, Kant must be wrong.

The question that really has to be asked here is whether we even can get "rid" of causality since a) it is so imminent in and vital to our lives as predictability is a cornerstone of all our actions, and b) as far as we know, quantum scale laws do not at all apply to our macroscopic world and cannot have any direct bearing in it, thus even if causality may become somewhat fuzzy there, it simply does not make any practical difference.

• @Nelson Alexander- "I would say that Kant's ideas of mechanical causality were much disputed in his day and are outmoded now, in part due to quantum effects, as you note." Because I'm fairly confident that there is no reason to accept any random effects at the quantum atomic level of activity to affect, let alone have any effect whatsoever on either human understanding of the reality of our experience, or any any of the theoretical underpinnings of the shape of the known universe, it puzzles me, what possible explanation for QM to 'disrupt' any concept concerning reality. What's up with that? – user37981 Oct 22 '20 at 14:56
• @CharlesMSaunders I do not know whether this stands below the wrong answer, but I couldn't agree more: Since Kantian philosophy is all about our manifest world and how we are able to perceive it through our senses (including our inner sense), there is no bearing of quantum-level laws or facts on Kantian categories as long as they do not directly effect or reach into the macroscopic universe. In some sense, the Kantian explanation ends at being able to read machine-mediated measurements off a screen (or sheet of paper). Everything beyond that is speculative analysis of phenomena. – Philip Klöcking Oct 22 '20 at 18:54
• I guess these remarks are aimed at me. Actually, I'll have to think about this. But I did toss that QM stuff in pretty carelessly, so thanks for the pointer. Is it true that quantum observations would have no impact on Kant's or a Neo-Kantian paradigm? I really don't know. And, come to think of it, I don't know what Kant included under "perception" via increasingly mediated instrumentation. I guess you are right, it would make no difference, but I'll have to sort it out. – Nelson Alexander Oct 22 '20 at 19:53