Re: GR or QED?

David Bowman (
Wed, 29 Apr 1998 13:21:11 EDT

Concerning Glenn's comments about his source (H. Ross) for the claim that
GR is the most precisely confirmed theory:

> I was corrected by both George Murphy and David Bowman when I said that GR
> was the best confirmed theory. I have thought about where I got that idea
> all day and figured out that it was from Hugh Ross. I found something that
> is close to that statement. he Says, " A recent experiment has confirmed
> the accuracy of Einstein's equation to at least twenty-one places of the
> decimal (within 0.0000000000000000001 percent.)" Creation and Time, p. 99.
> He cites S. K. Lamoreaux et al, "New limits on Spatial Anisotrophy from
> Optically Pumped 201Hg and 199Hg", Physical Review Letters 57(1986) pp
> 3125-3128.

George Murphy then surmised:

> I don't have this book of Hugh's but will check the PR article
>today. From the title it sounds like a special relativity confirmation
>- in GRT space _isn't_ strictly isotropic.

I don't have Hugh Ross' book either, but I did get a chance to check out
the PRL paper to which he refers. It seems that George's surmise from
the title is quite correct. The paper *does* only report on a particular
test of one aspect of local Lorentz invariance, (i.e. *special*
relativity, SR). That aspect is the local isotropy of space regarding
the possibility of anisotropic quadrupolar electromagnetic couplings to
the nuclear energy levels in Hg arising from a local anisotropy of space.
The paper claims an upper bound of about 10^(-21) for the fractional size
of the magnitude of such couplings. The results also provide a de facto
confirmation for the Equivalence Principle, EP, since it is via the EP
that we are allowed to ignore gravitational (curvature) effects when
considering phenomena in freely falling frames at the most local scale
sizes in spacetime. Crudely speaking, the EP (the strong form of it)
states that for all physical phenomena observed in a sufficiently small-
scale region in spacetime there exist local inertial frames, i.e. the
freely falling ones, such that *special* relativity holds in them. Thus,
if a laboratory-scale experiment demonstrated an invalidity of an aspect
of SR it would automatically be a demonstration of the invalidity of the
EP. It is true that if the EP is invalidated then GR is invalidated as
well since the EP is the basis for how GR embeds the local Lorentz
structure of spacetime into a theory that treats gravitational effects in
terms of spacetime curvature.

The main problem with Ross's claim regarding this paper is that it is not
a confirmation of the *specific distinctive* features of the predictions
that GR makes to the exclusion of all other theories of gravitation.
Bare SR (along with GR, and all other gravitation theories that
incorporate the EP) predict that space is locally isotropic (in a freely
falling frame). Even Newton's theory holds that space is isotropic
(if one sticks to mechanics and doesn't try to embed electromagnetism
into it via a luminiferous ether moving relative to the observer).

Since QED also incorporates SR in its formulation, an anisotropy of space
would would invalidate it as well. This paper could also be taken as a
confirmation of QED as much as a confirmation of GR. Thus the results of
this paper do nothing to discriminate among a host of potential theories.
The only theories that could possibly be disconfirmed/falsified by these
results are theories that specifically predicted that space *was* locally
*an*isotropic by a relative amount significantly greater than 1 part in
10^21. I don't know of any serious physicists seriously entertaining any
such theories (although this does not mean, by any means, there are
none--just that I don't know of them.)

Glenn asked:

> My question: Is Hugh correct about the applicability of this level of
> accuracy?

It seems so. But all that the results indicate is that, to this
high precision, the energy levels (or at least their differences) of Hg
nuclei (as seen by the precession of their spins) in a magnetic field
remain the same for all orientations (in the plane perpendicular to the
Earth's spin axis) of the magnetic field w.r.t. the fixed stars.

> Is the referenced article relevant to the issue?

It is relevant to the question of the an/isotropy of space. Any
relevance it has to *specifically* GR is misleading.

> Has any part of QED been verified to that many decimal places?

Since this result confirms QED just as much as it confirms GR, (since
both presume the validity of SR) then, I guess, the answer is yes. But
such confirmations *are* misleading as to their significance or their

David Bowman