🚀 go-pugleaf

crb7q@kelvin.seas.Virginia.EDU (Cameron Randale Bass) writes:

>In article <1992Aug10.184700.3204@sbcs.sunysb.edu> rscott@csws1.ic.sunysb.edu (Robert Scott) writes:
>> [claim that SR follows from signal delays] 

>      It is not this simple.  It is not just that there are 'speed of 
>      light delays'.  That was perfectly understandible in terms of 
>      a fixed propagational ether.  It is that the speed of light
>      appears to be the same no matter what your motion (or relative motion
>      with respect).

The interesting thing about this idea is that even though the 
space as modeled in the program has an "ether" (coordinates are
presumably stored internally in some absolute frame), objects
made entirely of formations of robots exchanging signals behave
in a Lorentz-invariant way, and the speed of light appears to
be the same in all frames, provided that you use measuring
equipment available within the program's physics: measure time
with a canonical pair of robots bouncing a signal, etcetera.

>     Restrict signals to the speed of light all you want, but you need
>     the Lorentz covariance to actually model relativity.  How do
>     you suggest programming this without using the SR framework?

Reread his post:  he's putting it in by denying his robots, which
are the only particles in the model world's physics, any 
knowledge of quantities that aren't Lorentz invariant.  The
*kinematics* of SR will indeed emerge if you only use systems of
robots as clocks, meter sticks, etcetera.  The only real problem
is that this is a model of insufficient generality to show how
SR intrudes into real-world phenomena, without the added
specification that c should look the same in all frames no matter
what measuring devices you use-- but that's more an issue of
interpreting the program than of writing it.  The model also
doesn't really address the full complexity of SR *dynamics*, but
one can model such things as collisions by giving the robots
appropriate instructions.

>>4.  SELECT A NEW VELOCITY, SPECIFIED AS THE VELOCITY THAT WOULD BE
>>NEEDED TO TRAVEL FROM EVENT X TO EVENT Y, WHERE X AND Y ARE AFFINE
>>COMBINATIONS OF THE POINTS OF ORIGIN OF PREVIOUSLY RECEIVED MESSAGES.  
>>IF X=Y OR IF THE SPECIFIED VELOCITY IS LIGHT-SPEED OR GREATER, THEN THE
>>COMMAND IS IGNORED.

>     Event Y already occurred?  How would one reach an event that
>     had already happened?

The robot isn't going to go *to* event Y, it's going to have the
*velocity* corresponding to the interval between X and Y.  This
is a means of specifying velocities that is completely Lorentz
invariant.

>      What you have here sounds very much (though in a confused
>     fashion) like galilean relativity with lightspeed delays.  I think
>     that you will find, to your misfortune, that you either have
>     a preferred frame or light (or messages) does not travel
>     at a single speed with respect to all observers.

Look again, keeping in mind that the only clocks and meter
sticks you can use are formations of robots.  Don't be fooled
by the all-caps (note to Robert Scott: on Usenet people will
read your posts more carefully and take you more seriously if 
you use lower case).

>>IT HAS ONE EXTRA DIMENSION, CONSISTING OF THE GROUP OF
>>UNIFORM EXPANSIONS OF SPACE-TIME ABOUT AN ARBITRARY REFERENCE POINT. 
>>NEVERTHELESS, I STILL THINK THAT THE EXPERIENCE OF LIVING IN THIS SORT
>>OF VIRTUAL WORLD CONVEYS ALMOST ALL OF THE FLAVOR OF WHAT SPECIAL
>>RELATIVITY IS LIKE.  THE EXTRA EXPANSION SYMMETRY IS VERY EASY TO
>>UNDERSTAND (DESPITE THE RARITY OF TRUE EXPANSION SYMMETRY IN NATURE, AS
>>ILLUSTRATED FOR EXAMPLE BY GALILEO'S ANALYSIS OF WHY IT IS UNHEALTHY
>>FOR AN ANIMAL FOR IT TO GROW TOO BIG), AND ONCE YOU'VE GOTTEN USED TO
>   
>     You are confusing the existence of a symmetry of the underlying
>     equations (space, etc.) with a symmetry of objects that exist 
>     in the space.  

He's not.  He's talking about a symmetry that physics may or may
not possess, and such things as scaling relations in nature do
rely intimately on the symmetry.  Suppose dilation symmetry really
did exist in nature: then a physical evolution enlarged in space
and time by a constant factor would still be an allowable physical
evolution, and lots of things having to do with objects would
be different.  Matter of a given composition could possess any
density, and objects could be structurally sound at any size
provided that the matter were scaled appropriately.  There
could be tiny civilizations made of incredibly dense versions of
ordinary matter.  The nonexistence of dilation symmetry in
real physics is what prohibits all this.

Personally I'd be all for giving the robots proper-time clocks.
That would both get rid of the extra symmetry and also imply
some Lorentz-invariant physics independent of the exchange of
light-speed signals, such as we have in real life.

Many SR-kinematic phenomena can indeed be made to follow from
the kind of model Robert Scott is talking about.  It's a little
misleading to say that it follows entirely from signal delays,
since you additionally have to base all physics upon the 
transmission of the signals.  It's better to say that it
follows from the intrinsically Lorentz-invariant nature of
the model he's constructed.

I would also argue that the treatment of SR from a geometric
viewpoint is ultimately simpler than thinking in terms of
a model where all physics consists of signal transmission.
Nevertheless, a program like the one he's described would
illustrate relativistic kinematics in a novel way.

-- 
Matt McIrvin, professional gradgrind, amateur Usenet drifter