From: helbig@asclothestro.multivax.de   
      
   In article <1pfndq5.fqa8tgza6wspN%nospam@de-ster.demon.nl>,   
   nospam@de-ster.demon.nl (J. J. Lodder) writes:   
      
   > What Ole Romer really measured was the astronomical unit,   
   > aka the AU, in terms of light seconds.   
   > (that means 'really' in an operational sense)   
      
   Yes, but since that distance was known, he effectively measured the   
   speed of light.  (What he "really" measured was the position of the   
   satellites of Jupiter as a function of time.)   
      
   > You may have noted that he can no longer do that.   
   > The AU too has been given a defined value of 149597870700 m (2012)   
   > What a new Romer will be measuring nowadays   
   > is where the Earth really is, with respect to Jupiter.   
      
   Right.  But, again, Nature doesn't care how we define our units.  We   
   don't expect the distance from the Earth to the Sun to vary, but if it   
   did, we would notice it.  (Actually, the AU is now defined solely in   
   terms of the mass of the Sun and an orbital period, with no reference to   
   the Earth.)   
      
   > > Bringing back the other issues, we could of course still measure the   
   > > speed of light with those old methods, and if it did actually change   
   > > with time, we would notice it, despite the facts that it is now defined   
   > > to be constant and that we believe that it is constant.  I don't think   
   > > that likely; my point is merely that we cannot prevent the speed of   
   > > light from changing simply by defining it to be constant.  Rather, it is   
   > > defined to be constant as a practical matter because we have evidence   
   > > that it is.   
   >   
   > There you go again.   
   > You start with the (Platonic) assumption that there 'really'   
   > is some 'speed of light' in some absolute sense,   
   > independently of measurements of it.   
   > (and that we can then 'measure' it in some unproblematic way)   
      
   Yes.  There really is a speed of light in an absolute sense.  (That is   
   an important idea in special relativity.  I don't think that you are   
   claiming that special relativity is wrong.)  Of course we have to   
   measure it.  But how we decide to measure it can't change its speed.   
   How we define our units cannot affect Nature.   
      
   > And yes, of course we can define the speed of light to be constant,   
   > and then it really is constant because we defined it to be.   
   > That merely implies that we absorb the changes,   
   > if any, somewhere else.   
   > (so somewhere in our choices about units)   
      
   Right, but if a variation were really observed, we would probably change   
   the definition of the units and not define them in terms of the speed of   
   light.   
      
   > Let me take a conceptually simpler example to make the point clear.   
   > The last CGPM defined Boltzmann's constant, k, to have a fixed value.   
   > That means that the triple point of water now needs to be measured.   
      
   Right.   
      
   > There is no point in saying that we cannot define the value of k,   
   > because it 'really' might be changing.   
   > Conversely there is no point in saying that the triple point of water   
   > 'really' is at 273.16 kelvin in some absolute unchangeable sense.   
   > There is no physical reality involved,   
   > it is just definitions versus definitions.   
      
   But this case is different.  Boltzmann's constant is essentially a   
   relation between temperature and energy.  Temperature has long been   
   understood to be related to the energy per molecule.  It's not   
   completely analogous to seeing ct as a fourth dimension, since that   
   makes sense only if there is some special speed.  That is similar to the   
   "mechanical equivalent of heat", since heat and mechanical energy used   
   to be measured in different units.  That is merely an accident of   
   history, like measuring liquids in gallons and solids in bushels.  It is   
   qualitatively different than defining a basic unit via a constant of   
   nature.   
      
   > Digressing into philosophy: your 'real speed of light'   
   > is an example of a Kantian 'Ding an sich' ,   
   > and as such unknowable.   
   > It has no place in physics.   
      
   My claim is much less: the concepts of speed, and the speed of light,   
   existed before the current SI definitions.  People measured the speed of   
   light.  One can still do those experiments today.  They still make   
   sense.  (Many people have investigated cosmological models with a   
   varying speed of light, for example.)  We don't expect the measured   
   speed of light to change AND THAT IS WHY THE METRE IS NOW DEFINED IN   
   TERMS OF THE SPEED OF LIGHT.  But if it did, I doubt that we would   
   change all quantities which involved the metre.  Rather, we would   
   re-define the metre so that the speed of light could be seen to vary.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)