>Stathis:  I don't know if you can make a sharp distinction between the 
> really weird universes where observers never evolve and the 
> slightly weird ones where talking white rabbits appear now 
> and then. Consider these two parallel arguments using a 
> version of the anthropic principle:
> 
> (a) In the multiverse, those worlds which have physical laws 
> and constants very different to what we are used to may 
> greatly predominate. However, it is no surprise that we live 
> in the world we do. For in those other worlds, conditions are 
> such that stars and planets could never form, and so 
> observers who are even remotely like us would never have 
> evolved. The mere fact that we are having this discussion 
> therefore necessitates that we live in a world where the 
> physical laws and constants are very close to their present 
> values, however unlikely such a world may at first seem. This 
> is the anthropic principle at work.
> 
> (b) In the multiverse, those worlds in which it is a frequent 
> occurence that the laws of physics are temporarily suspended 
> so that, for example, talking white rabbits materialise out 
> of thin air, may greatly predominate. However, it is no 
> surprise that we live in the orderly world that we do. For in 
> those other worlds, although observers very much like us may 
> evolve, they will certainly not spend their time puzzling 
> over the curious absence of white rabbit type phenomena. The 
> mere fact that we are having this discussion therefore 
> necessitates that we live in a world where physical laws are 
> never violated, however unlikely such a world may at first 
> seem. This is the
> *extreme* anthropic principle at work.
> 
> If there is something wrong with (b), why isn't there also 
> something wrong with (a)?

This is the problem of determining the appropriate "class" of observer we
should count ourselves as being a random selection on. There might indeed be
something wrong with (a); replace "The mere fact that we are having *this*
discussion" with, "The mere fact that we are having *a* discussion" to
obtain a dramatically different observer class. Your formulation of (a)
(*this* discussion) essentially restricts us to being a random selection on
the class of observers with access to internet and email, discoursing on the
"everything" list. Replacing "this" with "a" broadens the class to include
any intelligent entity capable of (and having) a discussion. 

The problem of determining the appropriate class seems a rather intractable
one. Choosing too broad a class can lead to unpleasant consequences such as
the doomsday argument; too narrow a class leads to (b). Mondays, wednesdays
and fridays, I believe that my appropriate reference class can be only one;
"Jonathan Colvin" in this particular branch of the MW, since "I" could not
have been anyone else. Weekends, tuesdays and thursdays I believe I'm a
random observer on the class of "observers".

Jonathan Colvin


Reply via email to