Immanuel Kant – What can we know? (opens in new tab)

> I think when Einstein or Newton make statements they are making statements about reality and not just sense impressions.

Einstein in particular made huge strides by thinking in terms of sense impressions, e.g. by using his imagination to take on the perspective of an observer falling off a roof:

> I got the happiest thought of my life in the following form:…for an observer in free-fall from the roof of a house there is…no gravitational field. Namely, if the observer lets go of any bodies, they remain relative to him, in a state of rest or uniform motion, independent of their special chemical or physical nature. The observer, therefore, is justified in interpreting his state as being "at rest."

The emphasis on sense impressions over an appeal to an external reality is something he owes to Mach: https://blog.superb-owl.link/p/church-of-reality-ernst-mach-...

You know that how you experience the world is not the world as it is.

Your eyes move about constantly but you dont see it, and the brain removes things, and the brain adds things, you smell, you listen, you feel, you end up with a construct you perceive, that you know is inaccurate.

You will only ever know the inside of your own head, your own consciousness.

To the extent that everything could be all inside your head, everything.

Some people like to think about this world being a simulation being run by some sort of beings for fun. They could again be simulations, recursively, until perhaps there could be a non-simulation at the top.

and it might be that the "simulation" they are running is only calculating and modelling a single brain and they plant things inside of it.

From that perspective nothing that you will ever perceive is real.

> there is an underlying reality apart from that which we can never know.

I'm surprised this is so unsettling for many. I feel like this life beats intellectual humility into anyone smart enough to notice it pretty much nonstop from the age of reason.

> I think when [...] Newton make statements they are making statements about reality and not just sense impressions.

That's not what Newton thought. Hypothesis non fingo[1].

> Maybe time isn't real in the way we perceive it but I don't think the underlying nature of time is something undiscoverable.

Time is an extension through something, but through what?

[1] https://en.wikipedia.org//wiki/Hypotheses_non_fingo

> Maybe time isn't real in the way we perceive it but I don't think the underlying nature of time is something undiscoverable.

My take is: humans are just another living thing. Most (all?) living things don’t know everything about the underlaying “reality“ (e.g., my dog doesn’t know about electromagnetism). Hence, most likely, humans cannot be sure that what they know IS the underlaying reality (unless, of course, we think we are somehow special).

Here is my thought that helped me intuit it.

Every image is taken from a specific point. Every image that you ever see, every image that will be, is all taken from a specific point. There is currently no image that is taken *not* from a specific point. Even 3d images.

The question of the thing in itself is like asking, "How would this pencil look like if not seen from a specific point? There should be some way to answer this, after all, why would a description of something needs something else in its specification?"

But then you start to realize how utterly impossible it is to imagine how something truly looks, when its not viewed from any specific point.

Cause we just never ever seen such an object ( a true image ).

> there is a sort of human perspective that shapes our understanding of reality and there is an underlying reality apart from that which we can never know.

This is quite the religious notion if you think about it.

> "there is an underlying reality apart from that which we can never know"

If you find this interesting, I'd very highly recommend reading Hegel. In my mind, he expands upon Kant to a new level. With Hegel, there isn't a world "out there, out of human reach". Rather, reality is always shaped by interactions of my impressions. There is no "real rock" that I am barred from experiencing in itself. There is no essence behind the form.

"Essence accordingly is not something beyond or behind appearance, but — just because it is the essence which exists — the existence is Appearance (Forth-shining). [...] Essence is the Concept as posited Concept. In Essence the determinations are only relational, not yet as reflected strictly within themselves; that is why the Concept is not yet for-itself. Essence– as Being that mediates itself with itself through its own negativity [relation to otherness]– is relation to itself only by being relation to another" https://larvalsubjects.wordpress.com/2006/12/13/hegels-conce...

Since you bring up Einstein, it's worth noting that he felt like he had disproved Kant, yet he claimed to be hugely influenced by Hume, an alleged empiricist (whom Kant even explicitly called out), though Hume's "empiricism" is largely misunderstood. He overtly understood that we don't directly perceive the real world, for instance, as have many philosophers since and including Plato.

Maybe you would find it less unsettling - and I believe you would be better understanding Kant's ideas - if, instead of thinking about it as "an underlying reality ... which we can never know", you think of it as "something (or nothing!) underlying reality ... which we can never know". So, yes, Einstein and Newton made statements about reality. And, yes, all of reality is (theoretically) knowable to humans (including the nature of time). But maybe there's something beyond reality itself? (Or maybe not!) Newton proved that we can understand gravity. Einstein proved that we can understand electrodynamics. Kant proved that we cannot understand god(s) (including whether or not god(s) exist(s)), because god(s) is by definition beyond reality, beyond space / time, beyond the observable universe.

I don't mind this notion in the sense of 'our sensory instruments can never/we can never know if they truly reflect reality', but the platonic/neoplatonic version where there's an idealised real world everything we see is a poor imitation of always bugged me

Electrons, quarks, gluons and muons are more probability fields than they are objects. Their mass is more of a fluidic slurry storm than actual matter.

Our world may be knowable but only within sloppy boundaries.

I don’t just find it unsettling or just an idea I don’t like - I hate this idea.

Why? Because it actually renders everything, anywhere, ultimately arbitrary.

Can you say there isn’t an underlying truth in the perspective of the anti-vaxxer?

Can you say there isn’t an underlying truth in the perspective of the pedophile?

Can you say there isn’t an underlying truth in the perspective of the flat earther?

According to Kant… well, you can’t completely. The logical consequences of this, if fully realized, would annihilate society.

It would seem in Kant’s view that Nobody, ultimately, has any right to pronounce objective truths on anything. Not even basic ones like “you shouldn’t kill the innocent.” That’s just a reflection of your own personal reality that has no inherent value over someone else’s reality.

Why learn algorithms when you could eventually invent a hash table?

I don't know if every brain can come to the realization that its inherently limited in what it can perceive of truth, especially since these limitations themselves are obscured from us.

You think you could?

Anyway, the work of others is a good way to get higher-level concepts into your head to toy around with.

Let's just grant you would come to the same conclusions eventually — people that specialised in thinking about metaphysics, ethics, epistemology etc. and spent years writing their arguments in (to a first approximation) well-reasoned prose, which was then subjected to the critiques of their colleagues that followed in subsequent generations who instead wrote their own arguments.. at the very least seems like a huge time saver.

I suppose ultimately you should read philosophy if you find it interesting. I like talking about philosophy, but find it hard to talk with people who have formally studied it, because they often rigidly cling to sets of ideas that get attached to the people who came up with them, rather than thinking more freely.

I often think it's a little unfair that history and knowledge are (obviously) biased towards people who thought enough of themselves to write their thoughts down.

I'm sure there are a lot of people who could make strong claims on the invention of certain thoughts or theories, but the bar happened to be too noisy for anyone to hear them.

Famous philosophers spend their whole lives to end up with a few unique conclusions. Do you believe you could come to the same conclusions as every single philosopher combined, in one lifetime?

I don't think that's a stupid question.

It's useful to read philosophy as we're exposed to the thought process of that specific author/philosopher. A lot of times, the conclusion you arrive at is influenced heavily by the lens through which you view the presuppositions.

It's useful to be exposed to different viewpoints and conclusions. You often read something that gives you pause, and you think, "I never thought of it that way."

I think this is akin to asking "why study computer science when you can just think it up yourself". The answer being you won't even have anywhere to start and you certainly will write bad software if someone did give you a computer to type away on. The field of computer science is much younger and smaller than philosophy.

Why stand on the ground when there are a multitude of giant's shoulders to stand on?

Philosophy is a good tool to find very poetic and often quite succinct descriptions of what you might otherwise come to on your own. It also provides a shared language among people, although often at the expense of inclusivity

It's like anything else. If you're interested in the subject matter, the most productive is a combination of the two.

Also analytical philosophy is the direct precursor to computer science and a huge influence on math as we know it.

You should try reading it, then, to see if it will be of any value to you.

You gain intellectual efficiency.

> "We note in particular that in the sequence induction IQ tasks, our system achieves human-level performance. This is notable because the Apperception Engine was not designed to solve these IQ tasks; it is not a bespoke hand-engineered solution to this particular domain. – Rather, it is a general purpose system that attempts to make sense of any sensory sequence, that just happens to be able to solve these IQ tasks “out of the box”.

Worth repeating:

> just happens to be able to solve these IQ tasks “out of the box”.

Will read further...

Honest question; what could anyone get out of this? What is the point of interpreting kant if the interpreter of kant can't even perceive what humans perceive? Kant doesn't seem to have much value outside of the context of human perception.

Just started reading this thesis. I am absolutely blown away by the clarity and beauty of its writing. This is rare.

This is the most staggering kind of serendipity to discover that such a thing exists. Thank you :)

Fascinating. Thank you.

> He takes it as axiomatic that we cannot use empiricism to determine universal rules, which is something that seemed pretty well established as false by the time between when it was written and when I read it.

I don't think this is true, at least with respect to the last ~80 years of the philosophy of science: the current "central dogma" of scientific inquiry is that all inferences must be falsifiable in order to be scientific.

In other words: good science generates inductive and parsimonious explanations, not universal rules. To be universal would be to discharge falsifiability.

Edit: For what it's worth, I think Kant's contemporaries also found him ponderous and difficult to read.

For what it's worth, the distance of time isn't the only thing that makes Kant difficult to read. There's also the distance of language - translation is almost inevitably going to blur the original lines of the original (I've heard it said, not of Kant but of Heidegger, that it's easier to learn German and read Sein und Zeit than it is to read Being and Time). And to make matters worse, Kant has always been criticized as a notoriously poor writer, even by his contemporaries.

(And there's the underlying complexity of his thought, of course.)

Science doesn't determine the rules of the universe, it generates models with predictive power within a domain of validity

> that we cannot use empiricism to determine universal rules,

we indeed cannot do this and Feynman illustrated this frequently using the analogy of chess (https://www.youtube.com/watch?v=o1dgrvlWML4).

As Kant argues we can never observe the rules of the physical world as things in themselves, just as we can't directly look at the rules of chess by observing games. We can only get individual hints that we piece together into models of the world in ways we can understand.

There are more options than “progresses in linear orderly fashion” and “totally arbitrary.” And if you read someone like Kuhn [1] you may question the progressing narrative of science.

I’d also be wary of separating scientific insights from philosophical/religious culture. Most famous scientists historically believed themselves to be looking for a divinely-organized system, and it’s unclear to me if they would have even looked in the first place without that underlying belief structure. You could go back even further and say all scientific inquiry basically rests on a conceptual distinction between Self and World, which is very much a trait of Monotheism.

I think philosophical insights could roughly be described as a mix of human biological realities (which haven’t really changed, ever) interacting with historical and technological events - and progressing over time in response to those events. Kant, for example, is mostly talking about things that are essentially fundamental to the human experience, even if he was in dialogue with the ideas that were popular in his time. People have argued against Kant, but I don’t think we can really “surpass” the kind of problem he’s investigating in the way you might think we have surpassed an outdated theory of physics.

1. https://en.wikipedia.org/wiki/The_Structure_of_Scientific_Re...

Flippantly: science accretes ideas, while philosophy sheds them. In other words: the progression you perceive in science (and mathematics, etc.) has historically occurred because of parallel progressions in philosophy. That has been less true of science for the past century, but it's still very true of mathematics.

> Is that also a progression of insights in a similar way?

Yes: Western philosophy is a "tradition" in the sense that it has a clear (branching and twisting) lineage from pre-Socratic thought to contemporary philosophy. There are lots of commonly recommended "history of philosophy" books that can provide that overview; Russel's "A History of Western Philosophy" is a commonly recommended one.

I would think of it more like software development if hardware remained the same rather than science. Yes there are shifts in paradigms based on what the use cases of the day are (and what the perceived shortcomings of the previous paradigm were), and yes there are some things that will be widely agreed to be advancements, also dead ends.

Once the language interface exists new generations of ideas transmit and evolve. The emergence of individualism from societies that were collectivist is an example of the magnitude of the significance.

Every system has arbitrary hard-coded constants deep down in its bowels somewhere. Why should the foundation of all knowledge be any different?

He tried to be cutesy and show how an enumeration of the criteria for these categories form "judgments" which define the categories themselves, essentially suggesting that this is likely to work since the analytical domain fundamentally involves self-consistency.

You want an article that takes 5 minutes to read to define "truth"? The author even says they're just trying to get you hyped and in my opinion I think they did a great job.

I don't think this was meant to be a scholarly explanation, and the author hints as much at the end of the post.

Kant was my main focus (now some time ago), and I thought this was a reasonable explainer for an audience that probably hasn't read much metaphysics.

I like Hume.

I like Kant.

I recommend Wittgenstein.

And Hobbes.

For Kant, space is the a priori condition of external experience. Time, the a priori condition of internal experience.

Space and time are conditions of human experience and hence human knowledge is limited to only phenomena in time and space.

And morals. We get a priori knowledge of morals. And that’s the important thing for Kant.

I leave you with an anecdote about him: once he was teaching a lecture and a student’s button was missing. He struggled to teach/cancelled the course and admitted the irritation was intense because of the button. My other philology friend told me.

Strange and interesting, makes me humanize and think about him in a different context.

Perhaps you should work on completing your incompleteness theorem.

Similar ideas can be found in https://en.wikipedia.org/wiki/Theory_of_constructed_emotion

Except instead of kaleidoscope of 6 emotions, you have a circle of affect - how pleasant/unpleasant and how aroused/tranquil do you feel in the moment affects your perception of the world and what kind of thoughts do you think.

Think of a sparse mixture of experts generating a space of solutions. A different mixture of experts may generate the same space. That the space can be factorized in multiple ways grants you knowledge of reality.

Even worse: it is not difficult at all to not even try (or have the idea cross one's mind) to model such things, but rather dismiss it as nonsense, pseudoscience, conspiracy theory, woo woo, etc. We might even live in a reality like this.

The oldest article about Kant on HN is from nearly sixteen years ago:

https://news.ycombinator.com/item?id=96994