Saturday, 4 July 2009

Kant for dummies – the “green sunglasses” explanation

Apologies to everyone for the self-indulgence, but i wrote this for my course and thought I’d share it with the internets. I promise it’s back to Hello Kitty vibrators and suchlike asap.

 

I looked around to find a nice summary of kant's metaphysics but couldn't, so I thought I'd summarise them for ya's:

Kant in a nutshell:

Here's my favourite way of explaining his insight:  imagine if everyone on earth wore green-tinted sunglasses, all the time.  To them, everything would look green.  In fact, if they didn't know they were wearing sunglasses, and if they had absolutely no way of taking them off, then as far as they were concerned, the world would be, objectively, green.

As it might be with green sunglasses, says Kant, so it is that our brains impose certain things on the way we perceive the world.  Things like space, time, logic and mathematics...  And in fact, it's impossible to conceive of an outside world that's truly independent of the way we perceive it - just as no-one in the green sunglasses world would be able to conceive of a non-green colour.

Without getting into too much techie detail, the question he was seeking to answer with this was "how is it that we know some things about the world to be true, without requiring evidence?" These things, what K. called "synthetic a priori-propositions", are non-trivial bits of knowledge like:

  • all of maths & geometry
  • "every event has a cause"
  • the rules of logic

and so on... Nowadays, we can question some of the above - it's far from clear, for example, that every event has a cause, or exactly what "cause" means...

But the fundamental insight that our objective reality is partially constructed by the ways that our brain perceives the world is a strong one. There's no such thing as colour, except insofar as we perceive it, for example... what else might be like this, we wonder?  Might we one day meet aliens that don't agree with us about Maths?

Here's where it might get relevant to AI - an artificial intelligence has a different type of brain to ours.  They will therefore perceive the world in a different way to us... Could it be radically different?  Or does the fact that AI's will have been (presumably) have been in some way constructed using our human assumptions such as maths and logic, mean that they will share most of the fundamental basis of perception with us?

 

Here's where it might get relevant to AI - an artificial intelligence has a different type of brain to ours.  They will therefore perceive the world in a fundamentally different way to us... Could it be radically different?  Or does the fact that AI's will have been (presumably) have been in some way constructed using our human assumptions such as maths and logic, mean that they will share most of the fundamental basis of perception with us?

 

Suggestions for improvement on the above are welcome, as are any challenges that I’m misrepresenting the big K. My memory if undegraduate days is pretty hazy…

6 comments:

  1. I realised I'd be amiss not to also link to this, from the same source.

    ReplyDelete
  2. This is not a criticism but a discussion:

    The green sunglasses analogy assumes that the content of our sensory experience comes directly from the "thing in itself" of objective reality and that our sensory apparatus is merely a filter, limiting the range of our perceptions but not validity of their content. My understanding of Kant was that he was saying something much deeper. The content of our sensory experience is determined more by how we perceive things (i.e. the categories) than what it is we are perceiving in the first place. Kind of like as if the green sunglasses transformed the images into sound. As you say, it's impossible to conceive of an outside world that's truly independent of the way we perceive it.

    Of course Kant then turned it into more of a headfuck by saying basic concepts like "causation" were categories. He uses these to construct a absurd gigantic philosophical system of synthetic a priori "knowledge" allowing him to make judgements on a broad range topics from aesthetics to politics, which in my view is a bit bonkers. However I know you are a big fan of Hegel, who took it to the outer limits of the ludicrous.

    On the AI point I think the answer will surely be the latter. We could program a different concept of space and time but ideas like "quantity", "causation", "infinity", etc. would surely have to be the same.

    This is all based on my hazy memory of Kant so may in fact all be wrong. Still, interesting stuff.


    Ben love the T shirt BTW.

    ReplyDelete
  3. ok so maybe the green sunglasses isn't so much of a summary analogy as an introductory analogy, a stepping stone on the way to Kant's full position. I didn't want to get too far into the whole "thing-in-itself" concept cos i thought it would scare beginners too much...

    also, i don't think Kant really questions the validity of our perceptions... in fact, he almost does the opposite when he says that we all have the same categories (space time etc) so that we can call them *objective* features of the outside world - as true as anything is. we can't know things in themselves, the only reality we have is that mediated by the categories, this is the reason that we can all agree about maths being true - as true as anything we'll ever be able to know.

    was kant a bit pessimistic? I think the whole point of science is to get to know "things in themselves" as much as we can. Already the bonkers implications of QM are letting us call into question some of kant's unshakeable categories, like causation. relativity gives us bendy space & time...

    perhaps we can't know things-in-themselves directly, or in our phenomenology, but we can get to know them intellectually.

    logic and maths seem like pretty solid bets for the time being though. but just you wait until we get some proper taoist mathematicians and scientists...

    i would like to see some taoist engineering, actually. "welcome to buddha airways. you have reached your destination, without traveling. the destination that can be reached is not the destination." etc etc.

    ReplyDelete
  4. Ben, that webcomic is awesome. theres another philosophy rpg strip here: http://dresdencodak.com/2006/12/03/dungeons-and-discourse/

    ReplyDelete