d. the Gerard wrote:
** A lot of people think making things hard policy
means people
will actually follow them. This means editorial guidelines get phrased
as didactic policy.
** Policy is harsh stuff, and there's a limit to how much people
will hold in their heads.
Here's another slant:
It's easy to imagine we need lots of reasonably hard rules,
because everyone knows that people are imperfect, and that
asking them to "be reasonable" or to "follow the spirit of the
guidelines" doesn't always work. So the goal is to remove
imperfect, subjective human judgement from the equation, and
replace it with a complete, cohesive, consistent set of nicely
objective rules, which will ensure magnificently coordinated,
correct behavior.
And that might work, *if* the rules were in fact complete,
cohesive, and consistent.
But they rarely are, even in a real society where there are
armies of paid legislators and civil servants poring over the
legalese. In an ad-hoc project where anyone can edit anything
(including the guidelines and policy documents), it's a virtual
certainty that the rules will never be complete, cohesive, or
consistent.
But when the rules are incomplete or inconsistent, fallible human
judgment is right back in the center of the equation again, as
various people (both on the doing-things side and the enforcing-
or cleaning-up-things side) try valiantly to decide *which* rule
ought to apply in each situation.
In the worst case, you end up with a lose-lose situation,
in which the rules exact all the costs that rules often do
(in terms of stifling initiative or creativity, taking on a life
of their own, and requiring significant resources to maintain,
thus parasitizing the rest of the enterprise), while delivering
none of the benefits, meaning that the costs of living without
rules -- the dependency on fallible human judgement, and the
eventual chaotic anarchy that may result when people do what they
want and fight over whether they're allowed to -- are also borne.
As always, it's important to take a step back and assess whether
a system -- in this case, a set of rules at a given level of
attempted completeness -- is actually achieving its goals,
and is doing so commensurate with its costs.