Paul A. M. Dirac, Interview by Friedrich Hund (1982)

"Physics & Astronomy"
21 followers
Yesterday, you gave us a short account of a very important development in theoretical
physics that started with Einstein's special relativity, with the symmetry of space and
time, and then led to a very large development, which ended with the other symmetry, the symmetry
with the particle and antiparticle.
That was the main subject, and the key word was symmetry.
Yes, that is so.
I wanted to emphasize that Einstein was the first to realize the importance of symmetry.
Lorentz had worked out the mathematics of the transformations.
The transformations are called Lorentz transformations, but Lorentz did not realize the symmetry which
was implied by these transformations.
He thought that there was one frame of reference which was the good frame, the physical frame,
and that other frames of reference were just mathematical fictions.
I remember the words.
He introduced local time as a concept, but he said local time different from the real
time.
There he has emphasized that there is one real absolute system, the ether.
And there he was quite wrong.
And it took him a long time, I believe until 1909, before he accepted the Einstein view.
Yes, he gave a lecture in Göttingen in 1909 or 10, and he said that perhaps we must give
up the concept of equal time, of the absolute equal time, but he, yeah, I can say in German,
Yes, Einstein started off with a different point of view, and he showed how he was an
independent thinker.
And with the word symmetry, in this connection you mean symmetry between time and space.
Correct, yes.
What symmetry was in the center of the origin has changed in time.
At first it was symmetry of space and time, then came other symmetries, symmetry between
electron and anti-electron, positive and negative charge.
And then more recently, other symmetries have appeared in physics, symmetry between proton
and neutron.
Yes.
And this is only an approximate symmetry, what physicists call a broken symmetry.
But if we, I would ask you something about the form of symmetry in your work, that symmetry
space-time showed, it's important if you constructed your, what we call direct equation
of the electron.
In the days before, physicists preferred the so-called Schrödinger-Klein-Gordon equation,
but this equation has the time in a second derivative, and that was not acceptable.
For you, you had constructed what we call transformation theory, and there you needed
the concept of probability and the first derivative.
That is correct, yes.
And that forced me into a different kind of equation, and this different equation brings
in the spin of the electron.
It was very unexpected to me to see the spin appearing in this way.
A spin, that means moment of momentum and magnetic moment, both together.
They both come together.
I thought one would first get a satisfactory theory of a particle without spin, and then
one would proceed to a more complicated theory of a particle with spin.
But it turned out differently.
The simplest kind of particle does have a spin.
Yes, you mean perhaps Pauli Weisskopf with the scalar theory, and yeah, physicists saw
that they can incorporate any spin into a consequent theory, but it is quite more complicated.
Yes.
I agree.
The simplest particle that could exist was the fermion with a spin one-half, the electron.
That is so, yes, yes.
It was a great revelation when this appeared, and then with the solution of these difficulties,
one was able to concentrate on the one remaining difficulty, how one should deal with the negative
energy states.
It took quite a while, perhaps a year to understand that, and to bring in the concept of a vacuum
in which all the negative energy states are occupied.
Is it possible to express it in the following manner?
In this theory of the electron, let me say the Dirac theory, you can treat pairs of particles
or positrons, electrons, in the form of a one-particle theory.
But if you have scalar spin, scalar particles, a Pauli Weisskopf theory, you can treat pairs
and only in the form of a many-body theory.
Yes, yes, that is so, yes.
Should we speak a little about general relativity?
Yes.
General relativity was considered by Einstein as his most important discovery.
He was working entirely alone on new mathematical ideas, and it was a big triumph for him to
be able to get his ideas straightened out and to introduce a new theory of gravitation,
bringing in a very powerful kind of symmetry.
This symmetry is of importance in physics only where gravitational fields occur, while
the symmetry previously, the symmetry of special relativity, is of importance in all physics,
so that this further symmetry introduced by general relativity, although it is such a
wonderful mathematical theory, does not have the big effect on physics.
And I might say that my recent work has been very much concerned with Einstein's general
relativity, and I believe that the times and distances which are to be used in Einstein's
general relativity are not the same as the times and distances which would be provided
by atomic clocks.
There are good theoretical reasons for believing that that is so, and for believing that gravitational
forces are getting weaker compared to electric forces as the world gets older.
There is some observational evidence for that, observations of the moon, which have
been made accurately for centuries with respect to the time provided by the Einstein theory,
and which have been made since 1955 with atomic clocks, and there is some evidence of a difference
between the two times.
The evidence is not as complete as one would like to have.
People are still working on the subject, in particular with the Viking lander which was
put onto Mars in 1976.
One is able to send radar waves to Mars and get back the reflected waves, and then one
can measure in atomic time how long it takes these waves to go to Mars and to come back.
The results that one gets are unfortunately very complicated because there are many disturbances.
There are disturbances caused even by meteors.
There are many more meteors passing close to Mars than there are passing close by the
Earth, and these disturbances all have to be taken into account.
People are still working on this subject, and I hope that they will get a definite answer
pretty soon about the question of whether there are these two times, the Einstein time
and the atomic time, with a difference between them.
What is a natural clock then?
Today, physicists used to say the vibration of a molecule or an atom is a natural clock,
and the extension of an undisturbed crystal is a natural length.
What would be the natural time and natural length then?
Those are the atomic times and the atomic lengths, and that is the best way to define
these quantities, but it may be that quantities defined in this way are not the ones to be
used in the Einstein theory for astronomical observations.
Should we say Einstein theory then must be changed or completed, or can it remain with
another interpretation?
The equations remain, but they are to be applied differently.
But in Einstein's equations, gravitation constant is assumed to be constant, and in
these modern developments it is not.
Well, whether the gravitational constant really is a constant or not will depend on what units
you use, because it's a matter of interpretation.
Yes, yes.
There is no suggestion that Einstein is wrong, but just on how his equations are to be used.
Einstein's theory has turned out to be a wonderful theory, so successful in all its
applications, but there is this question of the interpretation which is not yet settled.
But now then it can happen that Einstein's general relativistic theory becomes more important,
nearly so general as the special.
It is not so important for...
Yes, of course, it had only to do with cosmology and gravitation.
Yes.
But perhaps gravitation one day will be connected with the other forces, weak forces, strong
forces, electromagnetic forces.
That is possible, but that lies far ahead.
Yes, the factor 10 to the minus 14 is very little.
I think you were the first who laid the attention to these extreme constants, 10 to the minus
4, 10 to the minus 80, but we have no certain results.
The question is open.
There is some evidence from the observations of the moon, but one would like to have...
Yes.
...some confirmation coming from the radar observations from Mars.
Yes.
Yes, but can we say that quantum mechanics also had a great consequence, the whole theory
of matter, the whole theory of the properties of matter chemistry and so on is a consequence
of quantum mechanics.
In the atom, the motion of the electron is slower than that of light.
Yes.
And the chief features of quantum mechanics are non-relativistic.
That I would agree, and I think that the correct connection between quantum theory and relativity
has not yet been discovered.
Yes.
Can I ask you something of the modern development?
Now we hope that perhaps in a short time we get a unified theory of the particles.
Now we make differences between strong interaction, electromagnetic interaction, weak interaction,
gravitation, these four.
What are your hopes?
I think that the present methods which theoretical physicists are using are not the correct methods.
They use what they call a renormalization technique which involves handling infinite
quantities and this is not really mathematically a logical process.
I would say that it is just a set of working rules rather than a correct mathematical theory
and I don't like this whole development at all.
I think that some other important discoveries will have to be made before these questions
are put into order.
In particular, there is the problem of explaining the fine structure constant, the number 137
which plays an important role in physics and the question is why should it be 137 instead
of some other number?
That has not been explained at all and I feel that it is necessary to get an explanation
of that before one will make an important advance in understanding atomic theory.
The fine structure constant is one of them, also the relation between weak and strong
interaction.
For strong interaction perhaps you can say the constant is nearly one, it is really perhaps
one.
Then the mass, the proportion of the masses, electron mass, proton, neutron mass, I think
that is the other problem.
These two problems will be connected I think.
Probably yes.
There is quite a different problem of the ratio of the mass of the proton to the mass
of the electron.
The question is whether the ratio of these masses remains constant or whether it develops
slowly with time.
People are now looking into this problem.
