Log in

No account? Create an account
Feb. 21st, 2011 @ 12:54 am Exploring Moralspace
I'm trying to give some logical foundation to the idea that there is one true moral system, and maybe provide some idea of how to find it.

A few things out of the way first.  A moral system maps a set of present circumstances and a proposed course of action to a rating of goodness and a level of confidence in that rating (both outputs scalar).  Some have suggested a moral system could take only the resulting circumstances as input, but such as system would be unable to distinguish between (for example) infanticide and birth control.  Most of the work of a moral system is done with standard descriptivist logic to determine how the action interacts with a set of terminal values.  Terminal values are things which are seen as atomically right or wrong -- no reason can be given.

Consider the act of contemplating one's own terminal values (as performed by an intelligent mind).  Can the mind conclude its own terminal values are flawed?  Certainly!  Most attempted altruists, confronted with evidence of their own scope insensitivity, will try to rectify it.  Once the new terminal values are in place (difficult in real life, trivial in a thought experiment), the act can be repeated.  We now have an iterated system -- something we can work with.

For example, the null value set may be consistent, but as soon as you care about something a little, you'll think it's good to care about it more.  Nullity is a repelling fixed point.

Monomaniacal obsession is probably attracting.  Caring a little about something else is bad because it distracts you from what's really important.  But only if the obsession-target is simple.  Understanding all the details of the most important thing in the world is clearly good, and if that renders the obsession inconsistent, the value-set isn't fixed.  So simple obsessions are a set of attracting fixed points.  Handwaving vigorously, I propose they form an uninteresting submanifold.

We probably won't see attracting cycles, because the intelligent mind performing the iteration will notice the cycle and do something else with it.

We can talk about the volume of valuesets which approach a given attractor in the limit, assuming sane shapes and a distance metric (is there a standard distance metric for functions which output R2?).  Fixed points of high volume are probably particularly interesting.

It's a long way from the Word of God, but at least this manages to draw some intrinsic landmarks on what was a rather trackless sea.
About this Entry
Date:April 9th, 2011 06:49 pm (UTC)
(Permanent Link)
Did you heard what Rob Matts said about that?