SELF |
14 |
S.B. Karavashkin and O.N. Karavashkina |
|
6. Relativistic way to general covariance through the violation of conservation laws Speaking of especial mathematical way of relativistic
cognition, we cannot avoid the issue of term As J. Norton [32] correctly points, the problem arose when Einstein "came to conclusion that the equivalency principle which he advanced gives the clue to the relativistic theory of gravitation and to generalisation of relativity principle" [32, p. 57]. Later Einstein wrote on this issue the following: "Considerable achievement of general theory of relativity is, it saved the physics from the necessity to introduce the 'inertial frame' (either 'inertial frames'). This concept is not satisfactory, because it without any grounds (? - Authors) selects some systems from all ever minded coordinate systems. Then it is premised that the laws of physics are true only for such inertial systems (e.g., the law of inertia and the law of constant speed of light) (? - Authors). Thus, in the system of physics, the space as such is endowed with the role that selects it among all other elements of physical description. It takes a definitive part in all processes, being not affected by a feedback (? - Authors). Although such theory is logically possible, it seems to be not fully satisfying" [33, p. 854]. As we will see below, such neglect of the basic laws of nature discovered by Newton automatically caused a whole link of violations which Einstein had to admit, having unfoundedly rejected the inertial reference frame, and this rejection originated from the very beginning of his work on the project of Relativity. "The crucial results were obtained in 1912 and 1913, after Einstein arrived back to Zurich. Here Einstein founded the mathematical basis of general theory of relativity, using in mathematical issues the help of his friend Marcel Grossmann. But the theory which they built had a defect, as the authors did not succeed to involve to it the generally covariant equations of gravity field" [32, p. 57]. The cause was same. Einstein, having addressed to the generalisation of gravity laws, directed these efforts not to deepen the understanding of phenomenology of gravity field but to guess such geometric metric which by its outward parameters would seem like a motion of particle in the gravity field. As Norton evidences, in his Zurich notebook, "up to 3 006 14L, Einstein unsuccessfully runs over different simple expressions in a hope to find a suitable tensor; in the mentioned page he for first time explicitly writes down the Riemann curvature tensor. Most probably, it was Grossmann who told him the expression for the curvature tensor, since it is written near it: 'tensor of fourth rank by Grossmann' " [32, p. 62]. The features of such way of 'development' of physical ideas has been naturally reflected in the basic work by Einstein together with Grossmann "The project of generalised theory of relativity and theory of gravitation": "In the generalised vector analysis, the Christoffel four-sign symbols of second kind have a sense of components of mixed tensor contravariant in three signs and covariant in one sign. The main importance of these concepts for differential geometry of diversity given by its linear element (identical equality to zero of tensor Rikm is the necessary and sufficient condition in order the differential form could be reduced to |
![]() |
- Authors) allows us to premise a priori that these generalised differential tensors can appear useful also in composing the differential equations of gravity field. Actually, we can point at once the covariant tensor of the second rank and second order Gim , which could be involved in these equations, namely |
![]() |
(57) |
But in particular case of infinitely weak
static field of gravity, this tensor is unreducable to But factually Einstein understood, the difficulty was that
to the physical part of project the conservation laws were involved as Poisson equations: "Let the tensor |
![]() |
(58) |
To solve this problem, we could not find a method which would be so natural as in case of previous problem (seeking the expression for the tensor of energy-momentum - Authors), We had to introduce some far from being obvious, though probable admissions. The sought equation, most probably (italicised by us - Authors) has to be |
![]() |
(59) |
where It would seem, a clear discrepancy in the methodology of
Einstein's approach and reality would have to cause him to correct the methodology, but
only in case if Einstein did not profess Relativity. In accordance with it, in case of
discrepancy in relativistic conception and reality, it is reality what they think to be
incorrect. So immediately after Einstein introduced (59), he unfoundedly doubts the
Poisson conservation law and Newton law: "Accordingly to the
law of Newton - Poisson, it seems sensible to require, the equations (59) to be the
equations of the second order. But we have to object, this premise disables us to
find the differential equation being the generalisation of From this citation we clearly see that Einstein not for an
instant was falling to thinking, whether his wishes are consistent with reality. He is
ready to put off any proof, any conservation law, nothing to say of natural regularities,
striving to state his geometrical formalism true. And contradictions that arose in his
work are the direct consequence of neglecting the basic principles of methodology of
cognition that have been formed in classical physics. Well, what in essence intended
Einstein to fulfil? "With the help of absolute differential
calculus developed by Ricci and Levi-Civita, on the basis of mathematical study by
Christoffel, we succeed (? - Authors) to substitute the known
systems of equations of theoretical physics for such equivalent (in case of constant g |
![]() |
(60) |
We see from (60) that in its left part the mass m is involved as the measure of body's inertia determining its acceleration under affection of the force, and in the right part the mass m has been written as the weighting and providing the gravity interaction. But Einstein's task was to identify the laws in the inertial and non-inertial reference frames, reducing the body's motion in the gravity field to an usual accelerated motion. "An adherent of trivial modern theory of relativity with a definite degree of rightness calls the speed of material point 'seeming'. Namely, he can choose the reference frame so that the material point to have at the considered moment the zero speed. But if there is the system of material points that have different speeds, it already cannot introduce such reference frame relatively to which the speeds of all points will turn to zero. Similarly a physicists who stands for our view can call 'seeming' the gravity field, as through a definite choice of reference frame acceleration he can achieve, at a definite point of space-time the gravity field to turn to zero. But it is noteworthy that turning the gravity field to zero through the transformation is unachievable in general case for extensive gravity fields. For example, we cannot make zero the Earth's gravity field through choosing a suitable reference frame" [37, p. 285]. To the said Einstein adds an important stipulation: "If to describe the nature it might appear necessary to use an arbitrarily chosen coordinate system, the choice of the system has to be unlimited; physical laws have to be fully independent of this choice (general principle of relativity)" [38, p. 679]. And this really was Einstein to whom we have to attribute the identifying of the laws of nature in inertial and non-inertial reference frames. |
Contents: / 1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / 11 / 12 / 13 / 14 / 15 / 16 / 17 /