I am a little further on in it, almost half way, but I wanted comment on Fen's comments.
Human "understanding" in my opinion only exists in the narrative. Also it's a necessary human trait to filter information to generate narrative.I think this is what Taleb argues. We naturally seek explanations and story in order to comfort ourselves that we understand. I guess the question then boils down to, is the quest to be comfortable or is the quest to understand? In a number of cases I think Taleb is right is that it is very very dangerous to think you understand something. I think if you combine trying to find explanations with a deep mistrust of any conclusions you may reach you are probably on the right past.
Dogma and overconfidence are very dangerous creatures.
That description changes depending on our audience.... take a narrative line, follow it through until the narrative no longer fits and then adjust it until it does.I agree that we all have very different minds, and different narratives help get different people closer to the truth. I would be wary of finding a narrative that `fits' purely because it makes people feel more comfortable. I think a certain level of discomfort with your beliefs is important.
Not everyone wants to stare into the abyss of human understanding of randomness over breakfast.This speaks to the concern that not everyone actually cares. True. And maybe not starting the journey is better for people who don't want to continue along that journey. Maybe. I don't have an answer for that. For me... I am keen to get on the road.
You can be nihilist and dismiss everything that attempts to interpret anything, but I don't think that's helpfulI agree 100%. That's why I am intrigued by Taleb's arguments which strike a cord with me at the same time that I follow Garr Reynolds, Seth Godin and Malcolm Gladwell so closely. I think stories and narratives are the only way we understand something. I like Taleb's reference to Karl Popper and coming up with `Bold Conjectures' but then being your own beliefs strongest critics and spending more time looking for why you might be wrong then why you might be right.
I look forward to reading your blog when you start it Fen.
Exciting Times.
5 comments:
I think this exchange misses something pretty fundamental. We do actually know a lot of stuff for sure, so we can tell stories that are true. The trouble is that there are many things that our tiny brains are completely unable to understand, stories that we tell about these things are likely to be shit, like Thor being responsible for lightning.
As our brains swell and we invent technology and theoretical models we really do understand more that we caan attach narrative to, but that doesn't make telling stories about anything ok.
If the unwashed masses don't want to deal with the fact that people don't understand lots of stuff, well there are lots of other things to do, like watching the simpsons or whatever.
Stuart- good point re: things we can't understand, but you undermine your point by citing technology and models-- they only ever exist as a way of interpreting data into a visual or summarized form from which a human being can create... a narrative.
That narrative largely retains the same properties as something blagged off the top of your head, but with the added problem that the human now believes it to be of greater value than the blagged narrative.
When we boil for example physics right down to the lowest element, we get to a conjecture, or something axiomatic that we can't prove or disprove. i.e. we know nothing about it.
My point about the audience is that it is the meeting point of narrative and audience that gives it value. If you can save the life of a Nordic tribesmen by telling them not to stand on a hill waving metal rods at the sky in a thunderstorm because a very grumpy and somewhat warlike God might smack them with an enormous invisible hammer, then you've probably done some good, and the narrative has value. Who cares if it's shit?
Trev don't get me started on Taleb's views! He loves Monte Carlo models and yet as ever, any model requires assumptions that one can simply not be critical enough of. It's his biggest undermining point- that Black Swans, if they exist, defy ANY attempt to predict the nature of any future outcome. Only God should trade it seems.
I didn't mean to imply that these models weren't narratives of a sort, just that some are in fact better than others. We can find ways of understanding some stuff but not other stuff (yet).
So, of course I do think it is of greater value. And it's worth trying to know the difference between stories that look like F=ma and "assume people are rational".
I know about Hume's problem of induction, but I have a Bayesian attitude to episemology, I'm not bothered by the fact that my naturalism is ultimately just faith.
I've often thought of examples similar to your nodic chap. I understand the impulse and subscribe to it in the case of children. But I am a libertarian and I'm not comfortable making decisions for other people even if I judge that it would be for their own good. I think it's disrespectful and often dangerous. Lots of people feel passionately about their pet issue and are convinced that people Just Don't Get It! and then conclude that it's okay to exagerate their case in order to generate the "correct" response from people. This happens with terrrorism, global warming, nuclear power etc etc etc.
Nicely put. I think I'm on the same page here- although I'd go further and point out that a model is exactly that kind of foisting of superior knowledge. The "Don't Get It" bit is hidden in the document called "assumptions" and written in maths geek.
Models are just narratives in maths clothing.
PS, the exaggeration in models, to extend the analogy are often called "parameters" :-)
I do recall a certain Rating Agency who shall not be named looking at the output of a certain model we worked on, saying "please adjust this parameter, that parameter, remove this bit, and adjust that until the outcome is closer to our assumed profile for AAA losses."
I accept, for point of argument, that model of, say how an atom may work is potentially a little less subjective than the average financial "model", but maybe not.
Post a Comment