[PREV - LEARNING] [TOP]
FAITH_CYCLE
June 22-27, 2007
In general, I'm a
hyper-critical person --
I don't like to believe
anything I'm told, even
when it's by obvious
experts.
I like to understand the evidence and
reasoning behind every conclusion... and
even then I keep looking for ways the
conventional interpretation might be wrong.
Does that sound like a pack
of self-flattering, pretentious,
noble sentiments?
But really being a person
like this -- as opposed to
being a person who (when
convenient) claims to be
one -- is not at all an
un-mixed blessing.
It's a near impossibilty
to really examine
everything so closely -- Feynman (see "Surely You're
a genius-level intellect Joking") tells a story about
like a Feynman might be being led-astray temporarily
able to get away with it by what was widely regarded as
(maybe), but in practice common knowledge among the
it's ridiculous for physicists in the field, and
someone like myself -- only after learning they'd
relatively bright, but changed their minds did he go
hardly a genius -- to try back to the original paper,
to live this way. and then he remembered how
unimpressive he thought it was
At the very least, the first time he had read it.
I think it makes
my progress in He resolves at
learning things that point to
much slower than understand But then: Murray
it might be everything for Gell-man
otherwise. himself, and complained that
take no one's the point of all
For example, to this day, word for of Feynman's
I'm not *really* sure I anything. stories was to
believe in all that stuff show how much
about electron orbitals smarter he was
and the shell structure of than everyone
the the atom. Doesn't it else.
seem kind of clumsy and
arbitrary? I mean, the It's amazing
"Von Pauli Exclusion how well a
Principle"? *What* Actually: little self-
principle? It's totally thinking of deprecating
ad hoc! the atom as a humor can
badly written conceal that
Well okay, so maybe software sort of thing.
it helps explain module would
the semiconductor probably work Not that I
band gap phenomena for me. would know
that transistors anything about
are based on... Okay, so the such tactics,
interface is myself.
And, yeah, okay there's clumsy. What
other supporting data... else is new?
like spectroscopic lines Just deal.
(and little things, like
the fact that lasers work)
and there's the periodic
table of the elements and
all that...
That might convince
*some* people, but This sort of radical
I want *more* myself. skepticism might be
useful for someone
doing theoretical
work, but it's an
impediment to just
learning the material
well enough to use it.
So, should I try the
opposite approach?
Maybe the right thing
to do is to be a "true
believer" at least at
the outset.
Make the assumption that
the experts know exactly
what they're up to,
emerse youself in that
worldview uncritically...
then later, step back and
see if you can conjure up
some heresy. Ride the
faith cycle...
I continue to be
hyperskeptical in the Another variant of
world of software the rough/fine
engineering -- certainly strategy, but in
it's a field that reverse.
deserves it, it's far TWO_LEVEL
less intellectually
rigorous than something Instead of refusing
like physics and admittance to anything
chemistry. that doesn't meet
your standards, you
Snake oil abounds. throw the doors open
and hold an open house,
(Not to mention then selectively eject
some other animal the bad elements later.
products.)
SHANDY_TOWN_MIND
And that sounds a lot
I stick with perl rather than like a standard strategy
play the language-of-the- for designing "social
month game, but even as a network" web sites.
perl programmer, I'm
reluctant to take up the perl ACCESS_RAMP
techniques that are touted as
the latest thing...
The hot topic right now
would be "Class::InsideOut"
vs. "Class::Std", but
myself, I'm only just now
thinking about giving the
old "Class::MethodMaker" a
try.
And I still haven't
started using
"Class::Exception".
It's been standard
practice for over
five years, though,
maybe I should
just go for it,
and see what I think
of it later...
I find it very
difficult to
evaluate a software
technique in the
abstract, without
actually trying it.
There are always
gotchas that don't
turn up until you
apply it to a task.
--------
[NEXT - THINKING_HOWTO]