What Happened to Dingle?

On Thu, 20 Jul 2000, Harold Ensle wrote:
> H.Dingle really saw what was going on. He also had a simple proof 
> that showed that SR was impossible. For some reason SRists can't 
> seem to understand it.

L Hoffman wrote: 
>Would you care to summarize Dingle's proof?  I haven't seen it yet.

You almost certainly HAVE seen it already.  It's the same "proof"
that every anti-relativity crackpot puts forward, in one form or
another.  In a nutshell, Dingle considers two systems of inertial 
coordinates x,t and x't' with a relative velocity of v, and then very
elaborately and clumsily constructs the partial derivative of t' with
respect to t, and the partial derivative of t with respect to t'.  He
notes that these partials are equal, and declares this to be logically
inconsistent.  Sad.

In a more familiar context, Dingle's "proof" can be applied to show
that Euclidean geometry is "impossible".  Consider two Cartesian
coordinate systems xy and x'y' with a common origin, but rotated
by an angle of q with respect to each other.  Given the xy coordinates
of any point, we can compute the x'y' coordinates by means of the
equations

   x' = cos(q) x + sin(q) y         y' = -sin(q) x + cos(q) y

Conversely, we can solve these equations for x and y in terms of
x' and y' to give the inverse transformation

   x = cos(q) x' - sin(q) y'       y = sin(q) x' + cos(q) y'

This is just elementary linear algebra.  Now, if we hold y constant
and vary x, how much does x' vary?  In other words, what is the 
partial derivative of x' with respect to x?  Letting Dx'/Dx denote
this partial, it's obvious that Dx'/Dx = cos(q).  Now we ask a
different question, namely, if we hold y' constant and vary x',
how much does x vary?  This is equivalent to asking for the partial
of x' with respect to x, and of course we have Dx/Dx' = cos(q).

Dingle's confusion is that (like some clueless freshman calculus 
students) he imagines Dx'/Dx and Dx/Dx' are algebraic reciprocals
of each other, which would imply that 1/[Dx'/Dx] = Dx/Dx', and
therefore cos(q) = 1/cos(q), which is impossible for any q other
than 0.  Must we conclude that Euclidean geometry (and linear 
algebra) is inconsistent?!  No, because Dingle's argument is
obviously specious.  Partial derivatives can't be algebraically
inverted.

The application of Dingle's argument to the Lorentz transformation
is exactly the same.  Two inertial coordinate systems xt and x't'
with a mutual relative velocity v are related according to the
equations

    t' = (1/g) t - (v/g) x        x' = (1/g) x - (v/g) t

where g = sqrt(1-v^2).  The inverse transformation is

    t = (1/g) t' + (v/g) x'        x = (1/g) x' + (v/g) t'

In this case we have the two partial derivatives Dt'/Dt = 1/g and
Dt/Dt' = 1/g.  Dingle erroneously assumes that 1/[Dt'/Dt] = Dt/Dt',
and so arrives at 1/g = g, which is impossible for any v other
than 0.  Again the fallacy is the assumption that partial derivatives
can be algebraicaly inverted.

Of course, we CAN invert total derivatives, so let's see what happens
if we take the absolute differentials (for any constant v) of the 
time transformation equations.  We have

                dt' = (1/g) dt  -  (v/g) dx

                dt  = (1/g) dt' +  (v/g) dx'

Dividing the first equation by dt and the second by dt' gives

              dt'/dt  =  1/g  -  (v/g) dx/dt
                                                            (1)
              dt/dt'  =  1/g  +  (v/g) dx'/dt'

Considering the first of these equations, notice that in terms of
the unprimed reference frame we have  dx/dt = 0 (by definition),
and so we have dt'/dt = 1/g.  However, in terms of the primed
coordinates we have dx/dt = v, which gives

              dt'/dt  =  (1/g)(1-v^2)  =  g

This shows how we arrive at either one of the partials, depending
on which direction in spacetime we are considering.  Likewise for
the second transformation equation we can consider the cases when
dx'/dt' = 0 or -v, giving the results 1/g and g respectively.

Also, since equations (1) are reciprocals of each other, we can
multiply them together to give unity, i.e.,

  [dt'/dt][dt/dt']  =  [1 - v dx/dt][1 + v dx'/dt']/g^2  =  1

Notice that dx/dt is the velocity of one worldline with respect to
some arbitrary reference frame, and dx'/dt' is the velocity of the
other worldline with respect to that same reference frame.  Let us
denote these velocities by  u  and  w  respectively.  Recalling
that g^2 = 1-v^2, the above equation becomes

                     [1-vu][1+vw] = 1-v^2

Solving for u gives the familiar formula

                            w + v
                      u =  -------
                           1 + wv
  
which is the relativistic speed composition formula.  Needless to
say, there is nothing inconsistent or self-contradictory here.
Dingle was simply making an elementary error.  Pictorially, his 
claim was that the segment ratios a/b and c/d in the figure below
must be equal, whereas special relativity claims they are reciprocal.

                  A     N  H     B
                   \    /  \    /
                    \  /    \  /
                     \/      \/ 
                     /\      /\
                    /  \b  c/  \
                   /    \  /    \
                 a/      \/      \d
                 /    . '/\' .    \
                / . '   /  \   ' . \
              ./'      /    \      '\.
     t'=0 . ' /       /      \       \ ' . t=0
        '    /       /        \       \    ' .


The more interesting question is what happenned to Dingle.  For 
most of his life he wrote approvingly about relativity, beginning
with an article in 1922.  He even wrote a text book on special 
relativity (1940) in which he carefully explained the relativity 
of simultaneity, and so on.  But then in his later years (apparently
beginning around 1959) he embarked on a passionate anti-relativity 
crusade, and exhibited a complete inability to grasp the relativity
of simultaneity.  This culminated in his 1972 book entitled "Science
at the Crossroads".  The obvious question is, how could he explain 
a concept for years, and then later demonstrate a complete failure 
to grasp that same concept?  

I think the answer can be found through a careful reading of 
Dingle's essay for the Encyclopedia Britannica on the philosophical
consequences of relativity (written during his pro-relativity 
days).  This article, in retrospect, shows that his acceptance 
of special relativity (he never claimed to understand general 
relativity) was based firmly on the notion that there is no 
external objective reality.  He believed that the essential 
significance of relativity was that, in his words,

  ...the idea of something existing objectively, which physical
  measurements revealed, had to be given up...  The philosopher
  must henceforth interpret physics in terms of operations and
  their results alone, leaving external existences out of
  account...  Physics was thus thrown back on the unadorned
  description of itself as the discovery of relations between
  the results of chosen operations of measurements.

It's clear that Dingle accepted relativity (prior to old age) as 
simply a collection of brute facts that need not yield any coherent
picture of an objective external reality.  His view was similar to
the modern acceptance of quantum mechanics with the "measurement 
problem" unresolved, i.e., we can't think of a realistic model 
of an external reality that always yields the results of our 
measurements; we can only describe the patterns in those results
as abstract brute facts that must be accepted.  In other words,
Dingle's attitude (in his early days) was that "one does not 
understand relativity, one merely gets used to it".  This of
course is a paraphrase of a famous remark concerning quantum 
mechanics, but the point is that Dingle was entirely mistaken
in applying this "shut-up-and-calculate" approach to relativity,
because in fact relativity (unlike quantum mechanics) is an 
entirely classical theory, and is firmly based on a perfectly
coherent model of objective external reality.  In retrospect we
can see that the young Dingle never grasped this model - indeed
his whole philosophy of science (in those years) was that 
relativity had rendered all such models unviable.

If he had applied this line of reasoning to quantum mechanics, he
would have been in the mainstream of scientific thought, which 
still today has been unable to reconcile the full range of 
demonstrated quantum phenomena with any classically realistic 
objective model.  However, Dingle was very mistaken in applying 
this line of reasoning to special relativity, which is a purely 
classical theory with a perfectly sound objective model (Minkowski
spacetime).  Looking back at Dingle's early explanations of the 
twins paradox, we can see that he had NEVER grasped this simple 
model.  Instead he had simply told himself (like someone thinking 
about Schrodinger's cat) that when we make certain measurements 
we get certain results, despite the fact that he himself did not 
understand it.  He accepted this because he believed that NO ONE 
understood it, and in fact he believed that that was the whole 
point of relativity, i.e., that we must now believe things 
that cannot be "understood" in the classical sense of being 
manifestations of an external objective reality.

Then in his later years he rejected this approach (as did most of
the formerly enthusiastic circle of logical positivists), and decided
that we cannot reasonably dispense with the idea of an objective 
reality.  (Ironically, this was also Einstein's mature view.)  The
problem was that once he made this change, it exposed the fact that
he had never grasped the simple objective model of special relativity.
In fact, he had spent much of his life trying to convince himself and 
others that no such model was possible, and that this impossibility
was the whole message of relativity.  He could not, in his old age,
accept the idea that in fact relativity had a perfectly simple 
objective model, and that his views on this subject, to which he 
had devoted much of his life, had always been fundamentally flawed 
and misguided.  To contemplate such a possibility is not within the 
capacity of old men (as is demonstrated daily in internet newsgroups).

On a related point, I note that Professor McCrea (with whom Dingle 
had his famous "flame war" in the pages of Nature in 1962) just 
passed away in April of last year (1999) at the age of 94.  Obviously 
McCrea was on the right side of the debate, but in reading his 
rebuttals of Dingle's position I think they weren't as clear and as 
direct as they could have been.  Of course, any false premise can be
refuted in infinitely many ways, and there is always the temptation 
to pile refutations on top of each other, even though the effect of 
this is often to blunt rather than sharpen the refutation.

Return to Albro's Menu
Сайт управляется системой uCoz