ISSN 1063-7796, Physics of Particles and Nuclei, 2006, Vol. 37, No. 1, pp. 104–134. © Pleiades Publishing, Inc.

, 2006.
Original Russian Text © A.F. Zakharov, V.A. Zinchuk, V.N. Pervushin, 2006, published in Fizika Elementarnykh Chastits i Atomnogo Yadra, 2006, Vol. 37, No. 1.

Tetrad Formalism and Reference Frames

in General Relativity
A. F. Zakharova–d, V. A. Zinchukd, and V. N. Pervushind
a National Astronomical Observatories, Chinese Academy of Sciences, Beijing, 100012 People’s Republic of China
b Institute of Theoretical and Experimental Physics, ul. B. Cheremushkinskaya 25, Moscow, 117259 Russia
c Astro Space Center, Lebedev Physical Institute, ul. Profsoyuznaya 84/12, Moscow, 117810 Russia
d Bogoliubov Laboratory of Theoretical Physics, Joint Institute for Nuclear Research, Dubna, Moscow oblast, 141980 Russia

Abstract—This review is devoted to problems of defining the reference frames in the tetrad formalism of General
Relativity. Tetrads are the expansion coefficients of components of an orthogonal basis over the differentials of a
coordinate space. The Hamiltonian cosmological perturbation theory is presented in terms of these invariant dif-
ferential forms. This theory does not contain the double counting of the spatial metric determinant in contrast to
the conventional Lifshits–Bardeen perturbation theory. We explicitly write out the Lorentz transformations of the
orthogonal-basis components from the cosmic microwave background (CMB) reference frame to the laboratory
frame, moving with a constant velocity relative to the CMB frame. Possible observational consequences of the
Hamiltonian cosmological perturbation theory are discussed, in particular, the quantum anomaly of geometric
interval and the shift of the origin, as a source of the CMB anisotropy, in the course of the universe evolution.
DOI: 10.1134/S1063779606010035

CONTENTS 3.5.1. Primary quantization 113

1. INTRODUCTION 105 3.5.2. Secondary quantization 113
2. REFERENCE FRAMES IN MECHANICS 106 3.5.3. Bogoliubov transformation
2.1. Newtonian Mechanics 106 and universe creation 113
2.2. Special Relativity As a Model 3.5.4. Quantum anomaly of conformal time 114
of General Relativity 107 3.6. Levi-Civita Transformations
2.2.1. Relativistic mechanics as a consequence and Hubble Law 114
of electrodynamics 107 3.7. Quantization Results 115
2.2.2. Poincaré–Einstein dynamics 4. REFERENCE FRAMES IN GENERAL
of a relativistic particle 107 RELATIVITY IN METRIC FORMALISM 115
2.2.3. Geometrodynamics 4.1. Dynamics and Geometry
of a relativistic particle 108 of General Relativity 115
2.2.4. Reduction of geometrodynamics 4.2. Dirac Hamiltonian Approach 115
to the Einstein theory (1905) 109 4.3. Gauge Symmetry of Reference Frames
2.2.5. What could Einstein not know in 1905? 109 in General Relativity 117
2.2.6. Quantum anomaly of geometric interval 110 4.4. Class of Functions of Dynamical Variables 117
2.2.7. How does the invariant reduction differ 4.5. Universe Evolution As a Zero Mode Solution
from the choice of gauge? 110 of the Constraint Equation 118
2.2.8. Relativity as a principle 4.6. Separation of Zero Mode
of gauge symmetry 111 and Reduction of Action 118
3. RELATIVISTIC QUANTUM
4.7. Problem of Initial Data in General Relativity 119
MINI-UNIVERSE 111
3.1. Homogeneous Approximation 4.8. Low-Energy Dynamics
of General Relativity 111 and Correspondence Principle 119
3.2. Hilbert Variational Principle 111 4.9. Relative Units of Measurement 120
3.3. Reduction of Action 112 4.10. Universe Creation 121
3.4. Friedmann Equation 4.11. Matter Creation from the Vacuum 122
and Classical Cosmology 112 4.12. Results 123
3.5. Quantum Cosmology 5. REFERENCE FRAMES
as a Quantization of Constraints 113 IN TETRAD FORMALISM 123

104
 TETRAD FORMALISM AND REFERENCE FRAMES
5.1. Action, Linear Forms, and Symmetry 123
5.2. Reference Frame 124
5.3. Invariant Evolution Parameter 124
5.4. Hamiltonian and Constraints 126
5.5. Equations of Exact Theory 126
5.6. Resolution of Constraint Equation
and Reduction of Action 127
5.7. Geometric Sector 127
5.8. Hamiltonian Cosmological
Perturbation Theory 128
5.9. Central Gravitational Fields 130
5.10. Cosmological Generalization
of Newtonian Potential 130
5.11. Cosmological Generalization
of Schwarzschild Solution 131
5.12. Transformation of Reference Frames 131
6. CONCLUSIONS 131
1. INTRODUCTION
Since Newtonian times, theoretical physics has been
dealing with the laws of nature referred to as equations
of motion and some initial data needed to solve these
equations unambiguously. The initial data are to be
measured by a set of devices identified with a reference
frame. The guiding principle of modern physical theo-
ries is to define the transformation groups of reference
frames (treated as manifolds of initial data) that pre-
serve the equations of motion. The Galilei group in
Newtonian mechanics and the Lorentz–Poincaré group
in Special Relativity (SR) can serve as such examples.
Thus, it is important to establish a sense of the group
of general coordinate transformations in General Rela-
tivity (GR) [1]. The group of general coordinate trans-
formations was assumed by A. Einstein to be a direct
generalization of the Poincaré group, in which inertial
reference frames are substituted by “nonuniformly
moving reference frames” [2].
However, such an interpretation of general coordi-
nate transformations was then revised within the frame-
work of a new geometric principle (developed by
H. Weyl and V.A. Fock) of constructing physical theo-
ries, namely, the gauge symmetry principle. Whereupon,
GR stimulated the discovery of gauge symmetry.1
Fock noted that the group of general coordinate
transformations in GR took quite a different physical
role than the Poincaré group in SR [6]. Fock’s idea was
then developed further by Yang and Mills [7, 8],
Utiyama [9], Kibble [10], and others. They substanti-
1 Since1918, Weyl [3] had been trying to find geometric principles
to formulate electrodynamics. Weyl’s attempts were crowned
with success [4] only after the discovery of quantum mechanics.
The redefinition of quantum-mechanical momentum operator in
the presence of an electromagnetic field was the starting point of
Weyl’s geometric principle of electrodynamics (presently known
as a local gauge symmetry principle). Such a redefinition was
given by Fock in 1926 [5].
PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1
105
ated the modern understanding of the role of gauge
transformations, in particular, general coordinate trans-
formations in GR. According to this understanding, the
local gauge symmetry significantly differs from the
symmetry of reference frames, which is associated with
a set of initial data and integrals of motion. While addi-
tional symmetry of reference frames leads to new initial
data; i.e., to an increase in its number, the observation
of local gauge symmetry results in new conditions
imposed on the initial data, thereby decreasing its num-
ber.2 As was shown in [14], these consequences, fol-
lowing the invariance of field theory with respect to
gauge transformations of coordinates and field func-
tions, were formulated in Hilbert’s report “Foundations
of Physics” [15, 16], presented November 20, 1915, to
the Göttingen Mathematical Society. The GR action
was first written out in this report, and the GR equations
were derived by variation of the action. In addition, Hil-
bert first formulated the theorem that was later referred
to as the second Noether theorem [17–19]. This theo-
rem leads to the interpretation of general coordinate
transformations as gauge ones and, therefore, to all the
consequences concerning both a decrease in the num-
ber of independent degrees of freedom [20] and the
appearance of constraints imposed on initial data [14].
It can be said that principles of modern physical the-
ories were first formulated in Hilbert’s paper “Founda-
tions of Physics” [15, 16]. According to these princi-
ples, the basic notions of classical physics, namely,
action functional, symmetry of initial data, and equa-
tions of motion, should be completed with “geometric
notions” concerning invariant interval, gauge symme-
try, and constraint equations for initial data, respec-
tively. These principles include the Dirac’s definition of
observables [11], in particular, initial data varying
under transformations of reference frames, as invari-
ants of gauge transformations. For gauge-invariant
observables to be defined and, therefore, the gauge arbi-
trariness to be removed in solutions of equations, gauge
(i.e., general coordinate) transformations must be sepa-
rated from transformations of reference frames.
The problem of separating general coordinate trans-
formations from relativistic transformations of refer-
ence frames in GR was solved by Fock [21], by intro-
ducing an orthogonal basis and physical quantities
forming a Lorentz group representation. The Maurer–
Cartan linear differential forms serve as such quantities
and describe the motion of orthogonal bases in the
physical space of events. The components of a tetrad
defined as the square root of a metric are expansion
coefficients of the Maurer–Cartan forms in coordinate
differentials. The Maurer–Cartan differential forms,
being invariants of general coordinate transformations,
serve as measurable geometric quantities in the physi-
cal space. In this case, the integrable noninvariant coor-
2 Forexample, see the relationship between gauge transformations
and relativistic transformations of reference frames in quantum
electrodynamics [11–13].
2006
106 ZAKHAROV et al.

dinate differentials are treated as auxiliary mathemati- 2. REFERENCE FRAMES IN MECHANICS

cal quantities similar to electromagnetic potentials. 2.1. Newtonian Mechanics
A great number of papers were devoted to the choice First we’ll review the initial notions by using a sim-
of reference frame in GR (see, for example, monograph ple example of one-dimensional Newtonian mechanics,
[22] and references therein). In this review, we consider defined by the action functional
only the reference frames used for describing the evo-
lution of metric and fields in the form of the Hamilto- SL = ∫ dtL ( X ( t ), dX ( t )/ dt ),
nian dynamics [20, 23, 24]. This class of reference (1)
dX ( t ) 2 m
frames was defined by Dirac [23] as a bundle of four- L ( X ( t ), dX ( t )/dt ) = -------------- ---- .
dimensional coordinate manifold of spacetime into a dt 2
set of nonoverlapping spacelike hypersurfaces. The Here, x(t) is the variable describing a particle trajectory,
group of general coordinate transformations conserv- t is the time coordinate, and m is the mass treated as a
ing such a bundle was found by Zel’manov, who fundamental parameter of the theory. Variation of
referred to it as a group of kinemetric transformation; action (1), δSL = 0, under given boundary conditions
this group involves the reparametrization subgroup of δX(t0) = δX(t1) = 0, yields the equation of motion
coordinate time. This means that the coordinate time, 2
which is not invariant with respect to gauges, is not d X(t)
m ---------------
- = 0. (2)
observable. Therefore, one encounters the problem of dt
2

setting physical gauge-invariant variables and observ-

ables (including time, energy, and interaction poten-
tials) in GR for a specific reference frame [26–28]. A
method of solving similar problems in gauge field the-
ory was outlined by Dirac [11–13] as the Hamiltonian
reduction. The method involves the calculation of
action and other quantities on a constraint surface in
order to separate out the dynamical essence of the the-
ory from extra variables and gauge arbitrariness.
This review is devoted solely to the Hamiltonian
reduction in GR problems, which is treated as a method
of defining physical gauge-invariant variables and
observables in a specific reference frame. These
observables are to be used for constructing quantum
gravity defined as the first and second quantization of
energy constraint. We will compare the gauge-invariant
method of describing field dynamics [26–28] to some
noninvariant methods, similar to island universe [20] or
so-called global-time theory [29] assuming the coordi-
nate time is not observable.
The contents of the paper are as follows. In Section 2,
we consider the notion of reference frame in mechanics
and the Hamiltonian reduction of relativistic mechanics
with constraints onto an equivalent system without con-
straints. Section 3 is devoted to the Hamiltonian reduc-
tion in GR in the homogeneous approximation. In Sec-
The general solution
P
X ( t ) = X I + -----I ( t – t I ) (3)
m
of this equation depends on the initial data X(tI ) = XI
and dX(tI )/dtI = PI /m given at the time tI . The initial data
are to be measured by a set of physical devices (in this
example, by a ruler and a watch relative to a fixed
spacetime point) and to be associated with a reference
frame. The reference frames that move with constant
relative velocities are referred to as inertial frames. The
transformation X
X̃ = X + Xg + vg(t – tI ) turns a fixed
reference frame with its origin at the point X(tI ) = XI
into the reference frame moving with the velocity vg
and with its origin at Xg(tI ) = XI + Xg. This transforma-
tion group for reference frames in Newtonian mechan-
ics is referred to as the Galilei group. Equations (2) are
independent of initial data and, therefore, reference
frame. The independence of the equations, treated as
the laws of nature, on initial data is referred to as the
relativity principle [30–32].
In the Hamiltonian approach, action (1) takes the
form
dX ( t )
SH = ∫ dt P ( t ) ------------- – H .
dt
(4)

tion 4, we present a review of papers (Dirac, Here, P(t) is the momentum, {P, X} is the phase space,
Zel’manov, Lichnerowicz et al.) devoted to defining
2
reference frames in GR in order to formulate the Hamil- P
H ( P ) = ------- (5)
tonian reduction in terms of gauge-invariant observ- 2m
ables in a specific reference frame and to apply these
results to topical problems of modern cosmology. Ref- is the Hamiltonian function, and its value on a trajec-
tory is the energy E = H(PI ). Variation of action (4)
erence frames in GR are treated in Section 5 in the tet-
yields the first-order equations
rad formalism to construct the Hamiltonian cosmologi-
cal perturbation theory and its Lorentz transformations. dX ( t ) dP ( t )
P ( t ) = m -------------- , ------------- = 0, (6)
We here use the system of units in which c =
= 1. dt dt

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

 TETRAD FORMALISM AND REFERENCE FRAMES 107

rather than second-order equations (2). According to There are the two methods of describing relativistic
Newtonian mechanics, all observers in different refer- particles in literature. The first one, based on constraint-
ence frames use the same absolute time t. free systems, was proposed by Poincaré and Einstein in
1904–1905 [33, 34]. The second method was formu-
lated by Hilbert and Einstein in 1915 in terms of sys-
2.2. Special Relativity As a Model of General Relativity tems with constraints [1, 15].
2.2.1. Relativistic mechanics as a consequence of 2.2.2. Poincaré–Einstein dynamics of a relativistic
electrodynamics. As was shown above, the notion of the particle. Any reference frame in SR is defined by a unit
spatial coordinates Xi (i = 1, 2, 3) in Newtonian 2 2 2
mechanics as dynamical variables is clearly separated timelike vector l(µ), with l ( µ ) = l ( 0 ) – l ( i ) = 1, which will
from absolute time t, treated as an evolution parameter. be referred to as a time axis. These vectors form a com-
plete set of the Lorentz reference frames. The time in
Relativistic mechanics is based on the Faraday– each frame is defined in Minkowski space X(µ) as a sca-
Maxwell electrodynamics symmetry group found by
Lorentz and Poincaré. The time t = X(0) and spatial coor- lar product of the time axis vector and the coordinate:
dinates X(i) (i = 1, 2, 3) are treated in this group as a uni- X(0) = l(µ)X(µ). The spatial coordinates are defined on the
⊥
fied space of events or the Minkowski spacetime X(α) three-dimensional hypersurface X ( µ ) = X(µ) –
(α = 0, i) with the scalar product of any pair of vectors l ( µ )(l ( ν ) X ( ν )) perpendicular to the time axis.
(A(α)B(α) = A(0)B(0)B – A(i)B(i)) [32].
Relativistic particles are described in SR by the Without loss of generality, the time axis can be cho-
action sen in the form l(µ) = (1, 0, 0, 0) defining the observer
rest reference frame. After solving the equations, arbi-
d X ( α )⎞ 2 trary Lorentz frame can be introduced. Taking out the
S SR = – m dτ ⎛ -----------
∫ - . (7) factor dX(0) /dτ from the radicand in Eq. (7), we arrive at
⎝ dτ ⎠
the action integral in the initial formulation of SR (1905):
This action is invariant with respect to the transforma- 2
dX (0) dX (i)
tions X ( α ) = X ( α ) p + Λ ( α ) ( β ) X ( β ) of the Poincaré group, ∫
S SR:1905 = – m dτ -----------
dτ
- 1 – -----------
dX (0)
-
which is the transformation group of reference frames. (10)
Its subgroup of rotations Λ(α)(β)X(β) is referred to as the dX (i) 2
Lorentz group. Fixing the indices (0), (i) in this space
of events [X(0) |X(i)] implies the choice of a specific
∫
= –m d X ( 0 ) 1 – -----------
dX (0)
- .

Lorentz reference frame. Expressing the momentum in terms of the velocity

It should be noted that Special Relativity contains a V(i) = d X ( i ) /d X ( 0 ) entering into the variation of the
new symmetry with respect to the transformations that 2
do not change the initial data; namely, action (7) is Lagrangian L = – m 1 – V ( i ) ,
invariant with respect to the reparametrization of the
coordinate evolution parameter P ( i ) = ∂L/∂ ( V ( i ) ) = mV ( i ) / 1 – V ( i ) ,
2
(11)
τ τ̃ = τ̃ ( τ ). (8) one can obtain the Hamiltonian function
This results in originating constraints between the
2 2
variables. This transformation group is referred to as a H ( P(i) ) = P(i) V (i) – L = m + P(i) ( X (0) ) (12)
gauge group, while the quantities invariant with respect
to gauge transformations are called observables. The and rewrite action (10) in the Hamiltonian form
geometric time interval
dX (i)
τ
d X ( α )⎞ 2
S SR:1905 = ∫ dX (0)
dX (0)
- – H ( P(i) ) .
P ( i ) ----------- (13)
s(τ) = ∫ dτ̃ ⎛ -----------
- (9)
⎝ dτ̃ ⎠ The energy is defined as the Hamiltonian function
0
2 2
on the world line of a particle in the space of events X(α) on the trajectory: H(PI(i)) = m + P I ( i ) . The famous
can be taken as such an observable that is invariant with formula E = mc2 (with c = 1) is a consequence of the
respect to the time reparametrization. This interval is definition of physical observables from the correspon-
measured by a comoving observer. The time variable of dence to classical mechanics and follows from the low-
the space of events X(0) is the time measured by an energy expansion of the Hamiltonian function in pow-
external observer. ers of dynamical variables:
The goal of the theory is to solve the equations 2
describing trajectories in the space of events in terms of 2 2 P(i)
H ( P(i) ) = m + P(i) - + ….
= m + ------- (14)
gauge invariants. 2m

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

108 ZAKHAROV et al.

X0 PARTICLE The following question arises. Can a relativistic par-

ψ(s|qi) ticle theory be formulated such that all physical conse-
P0 = E quences, including relativistic effects, be described by
a variational equation?
ds We will prove that such a relativistic particle theory
ψ(X0|Xi) can be formulated by perfect analogy to the Hilbert’s
X0 = X0I
“Foundations of Physics” [15], i.e., as a geometrody-
q2 namics. According to this theory, the description of a
q1
X2 physical system is based on action functional, geomet-
X1 ric interval, symmetry of reference frames, gauge sym-
metry, equations of motion, and constraint equations
for initial data.
Fig. 1. Motion of an unstable relativistic particle in a world
line in the space of events. The motion is completely 2.2.3. Geometrodynamics of a relativistic particle.
described by the two Newtonian-like sets of observables, According to Hilbert [15], the geometrodynamics of a
dynamical and geometric. Each has its proper time and relativistic particle is based on the two principles: the
wave function Ψ. The two measured life-times of the parti- action [27, 28]
cle (the time as either a dynamical variable X 0 or a geomet-
ric interval s) are interrelated by the equations of motion fol-
lowing from the action of Hilbert-type geometrodynamics ∫
S SR:1915 ≡ dτL SR:1915
rather than by Lorentz transformations. (17)
dX (α) ⎞ 2
= – ---- dτe ( τ ) ⎛ ---------------
m
2 ∫ ⎝ e ( τ ) dτ⎠
- +1
Variation of action (13) with respect to canonical
momenta P(i) and variables X(i) yields, respectively, the for the variables X(α) = [ X ( 0 ) X ( i ) ] forming the space of
velocity in terms of the momenta, events of the moving particle and the geometric interval
P(i)
V ( i ) = ------------------------
-, (15) ds = e ( τ )dτ (18)
2 2
m + P(i) in the Riemann one-dimensional space on the world
line of the particle in this space (see Fig. 1). Here, e(τ)
and the momentum conservation law: dP ( i ) /d X ( 0 ) = 0. is the only metric component, so-called the shift of the
The solution of these equations determines the particle coordinate evolution parameter.
trajectory in the space of events: Variation of the action with respect to the function
e(τ) yields the equations of geometrodynamics
X ( i ) ( X ( 0 ) ) = X ( i ) ( X I ( 0 ) ) + V ( i ) [ X ( 0 ) – X I ( 0 ) ], (16)
2 2
[ e ( τ )dτ ] = d X ( α ) . (19)
where XI(0) is the initial time relative to the observer rest
frame. Solving these equations in e(τ), we arrive at
The transformation to any reference frame is
d X ( α )⎞ 2
described by the corresponding Lorentz transformation e ( τ ) = ± ⎛ ------------
- . (20)
and equivalent to an appropriate choice of a time axis. ⎝ dτ ⎠
Every reference frame has its proper time, energy, and
momentum. The relationship between dynamical vari- It is seen that action (17) coincides on these solu-
ables and times in different reference frames is treated tions with initial action (7) of the relativistic particle up
as the relativity principle formulated, most clearly, by to a sign. The negative sign of e(τ) in Eq. (20) implies
Einstein [34]. According to the Einstein relativity prin- the change of the mass sign in action (7) for an antipar-
ciple, the Lorentz transformations contain extra informa- ticle. Equation (19) is referred to as a constraint equa-
tion on relativistic effects as compared to solutions (16) tion.
of the dynamical equations derived by variation of For the Hamiltonian relativistic-particle theory with
action (13). constraints, the corresponding action can be derived
from (17) by introducing the canonical momenta Pα =
Therefore, the appearance of relativistic effects due
to the Lorentz kinematic transformations (i.e., transfor- ∂L SR:1915 /∂ Ẋ ( α ) :
mations of reference frames) means that the Einstein τ2
theory significantly differs from Newtonian mechanics. dX (α) e ( τ ) 2
∫ dτ
2
In the latter, all the physical effects are to be deduced S SR:1915 = - + ---------- ( P ( α ) – m ) . (21)
– P ( α ) -----------
from the equations of motion by variational method dτ 2m
τ1
with due regard to initial data. In this case, the Galilei
group in Newtonian mechanics contains nothing Here, the function e(τ) of the shift of the coordinate
beyond solutions of the equations of motion. evolution parameter τ defines geometric interval (18):

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

 TETRAD FORMALISM AND REFERENCE FRAMES 109

τ function in the spatial dynamical variables [P(i), X(i)].

ds = e ( τ )dτ
s ( τ ) = ∫ dτe ( τ ). (22) According to the principle of correspondence to New-
tonian mechanic, these variables belong to so-called
0 reduced phase space. The variable X(0) is the evolution
Action (21) and interval (18) are invariant with time relative to the observer rest frame.
respect to the reparametrization of the coordinate evo- In a given Lorentz reference frame, the time compo-
lution parameter τ: nent of solution (26),
τ τ̃ = τ̃ ( τ ). (23) P 0±
X ( 0 ) ( s ) – X ( I0 ) = -------
-s (29)
Therefore, SR could be referred to as a one-dimen- m
sional GR, with the reparametrization group of coordi-
nate evolution parameter (23) serving as a group of of geometrodynamics has no an analogy in Newtonian
gauge (general coordinate) transformations. The equa- mechanics. In this case, formula (29) is a pure kine-
tion for the auxiliary shift function δSSR /δe = 0 deter- matic relation between the two times noted above,
mines the energy constraint imposed on the particle namely, the dynamical variable X(0) and the geometric
momenta P(0), P(i), interval s:
2 2 2 m
P(0) – P(i) = m , (24) s = [ X ( 0 ) – X 0 I ] -------- . (30)
P 0±
so-called a mass surface equation.
The equations This equation will be referred to as a geometric ratio
of the two times of a relativistic particle, namely, the
d X (α) d X (α) dP ( α ) time [X(0)] as a variable and the time s as an interval.
- ≡ m ------------
P ( α ) = m ------------ -, ------------ = 0 (25)
edτ ds ds The substitution of geometric ratio (30) into spatial
part
for the variables P(α), X(α) derived by a variation of
action (21) are gauge-invariant. The solution P(i)
X ( i ) ( s ) = X ( Ii ) + -------
-s (31)
PI (α) m
X ( α ) ( s ) = X I ( α ) + ----------
-s (26)
m of solution (26) gives the relativistic equation of motion
of these equations in terms of geometric interval (22) is in the reduced phase space [P(i), X(i)],
a generalization of solution (3) of the Newtonian equa- P(i)
tions in the Minkowski space. In this case, the geomet- - [ X – X ( 0 I ) ],
X ( i ) = X ( iI ) + ------- (32)
ric time interval serves as an evolution parameter, while P 0+ ( 0 )
PI(α) and XI(α) are initial data for the four variables at the with the time [X(0)] as a variable.
point s = 0:
Thus, geometrodynamics in a specific reference
X (α) ( s = 0 ) = X I (α) . (27) frame consists of constraint-free “particle dynamics”
(32) and “geometry” (31) describing purely relativistic
These equations contain three new features, as com- effects by the equations of motion in the same reference
pared to Newtonian mechanics, namely, momentum frame [27, 28].
constraint (24), the time component in solution (26) of
the equation of motion, and the initial value XI(α) of time 2.2.5. What could Einstein not know in 1905? For-
as a variable. mula (13) for the action describing a moving particle
can be derived by substitution of solution (28) into
2.2.4. Reduction of geometrodynamics to the Ein- geometrodynamic action (21). Such a substitution also
stein theory (1905). Action (21) and interval (22) were gives the action for a particle with negative energy (28):
above referred to as geometrodynamics of a particle.
The geometrodynamics of a particle is characterized by S SR:1915 P ( 0 ) = P 0–
the two times in every reference frames, namely, the
X(0 I )
time as a geometric interval measured by an observer dX (i) (33)
∫
2 2
on the world line and the time as a dynamical variable = d X 0 – P ( i ) -----------
- – P(i) + m .
measured by an fixed observer. dX (0)
X(0)
The physical interpretation of solutions (24) and
(26) of geometrodynamics is determined by the choice The equations corresponding to this action have the
of a specific Lorentz reference frame Pµ = (P(0), P(i)), so- solutions
called observer rest frame. The solution m
X i = X ( iI ) + -------- [ X ( 0 I ) – X ( 0 ) ( s ) ]
P ( 0 )± = ±
2
P(i)
2
+ m = ±H (28) P 0–
(34)
m
of constraint equation (24) in the zero momentum com- = X ( iI ) + -------- [ X ( 0 ) ( s ) – X ( 0 I ) ].
ponent P(0) in this reference frame is the Hamiltonian P 0+

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

110
The problem of negative energy was solved in quan-
tum field theory [19].
It is known that quantum relativistic mechanics is
defined as the quantization of energy constraint (24),
(P0)2 – (Pi )2 = m2, by substituting the particle momen-
tum Pα = (P0, Pi ) by its operator P̂ α = –i∂α. The quan-
tization yields the Klein–Gordon–Fock equation for
wave function,
2 2
[ ( P̂ α ) – m ]Ψ [ P α X α ] = 0,
as a quantum analog of constraint equation (24). This
equation has the normalized solution,
– 1/2 + tum theory is referred to as a quantum anomaly [36].4
Ψ [ P(α) X (α) ] = ( 2 P(0) ) [ a Ψ P( 0 )+ θ ( X ( 0 ) – X I ( 0 ) )
– (36)
+ a Ψ *P( 0 )– θ ( X I ( 0 ) – X ( 0 ) ) ],
where θ is the Heaviside step function and the terms
with the coefficients a+ and a– correspond to two classi-
cal solutions of constraint equation (24) with positive
and negative energies (28).
Quantum field theory is known to be formulated as
the quantization of the coefficients a+ and a–, i.e., as the
second quantization of relativistic particles [18, 19]. In
this case, to exclude negative energies, –|P(0) |, and
therefore to ensure the stability of quantum systems,
the coefficients a+ and a– are to be treated as production
and annihilation operators, respectively, for particles
with positive energy.3
This treatment is equivalent to the postulate of the
existence of the vacuum as a lowest-energy state in the
space of events. The postulate imposes a constraint on
the motion of a classical particle in the space of events,
namely, the particle with the energy P0+ (P0–) moves
forward (backward):
P 0+ X I (0) ≤ X (0) , P 0–
The following question arises. How does causal
quantization (36) with constraint (37) influence geo-
metric interval s (22)?
2.2.6. Quantum anomaly of geometric interval. To
answer this question we perform the Lorentz transfor-
mation from the rest reference frame to the comoving
frame: [ X ( 0 ) X ( i ) ] , where P ( i ) = 0 and P 0± = ±m.
It follows from (30) and (37) that the time X ( 0 ) in
the comoving frame is related to the geometric interval
s by the equation
s ( X ( 0 ) X I ( 0 ) ) = ( X ( 0 ) – X I ( 0 ) )θ ( X ( 0 ) – X I ( 0 ) )θ ( P ( 0 ) )
+ ( X I0 – X 0 )θ ( X I ( 0 ) – X ( 0 ) )θ ( – P ( 0 ) ).
3 Moreover, the initial data XI(0) is treated in quantum theory as a
point of production or annihilation of a particle.
ZAKHAROV et al.
This expression for the geometric interval s in quan-
tum field theory looks like the Green’s causal function
of the comoving time:
2
d s ( X (0) X I (0) )
-----------------------------------
2
= δ ( X ( 0 ) – X I ( 0 ) ). (39)
d X (0)
Therefore, the positive geometric arrow of time, s ≥ 0,
is a consequence of the postulate of existence of the
vacuum as a lowest-energy state, which leads to the
(35) existence of the absolute time origin s = 0. The positive
arrow of time implies the breaking of classical symme-
try with respect to the transformation s to –s. In contrast
to classical symmetry, breaking of symmetry in quan-
Under the assumption of the existence of the vac-
uum as a physical lowest-energy state, the second quan-
tization of an arbitrary relativistic system leads to the
absolute geometric-time origin s = 0 in this system. The
question on what was before the creation of a relativis-
tic particle, string, or universe has no physical sense for
an observer measuring time because time is created
together with the quantum relativistic universe as a con-
sequence of universe stability.
2.2.7. How does the invariant reduction differ from
the choice of gauge? We now compare the gauge-
invariant method of describing field dynamics [27, 28]
to the gauge-noninvariant method, assuming that the
coordinate time x0 becomes observed.5 In the case of
SR under consideration, this assumption implies the
use of the synchronous gauge e(x0) = 1 in action (21):
τ2
dX (α) e ( τ ) 2
∫ dτ
S SR = – P ( α ) -----------
2
- + ---------- ( P ( α ) – m ) . (40)
dτ 2m
τ1
This yields a constraint-free theory:
S SR [e = 1]
X I (0) ≥ X (0) . (37) (41)
τ2
dX (i) dX (0) 1
∫
2 2 2
= dτ P ( i ) ----------- – P ( 0 ) ------------ – ------- ( –P ( 0 ) + P ( i ) + m ) .
dτ dτ 2m
τ1
From the viewpoint of quantization, Eq. (41)
describes an unstable system because it contains the
variable X(0) making a negative contribution to the
2 2
energy E = (– P ( 0 ) + P ( i ) + m2)/2m, which is defined in
the interval (–∞ < E < ∞). The particle action on the
three-dimensional hypersurface defined by the condi-
4 The anomaly associated with Dirac fields also follows from the
vacuum existence postulate. This fact was first pointed out by Jor-
dan [35] and then rediscovered by a lot of authors [36]. The vac-
uum existence postulate is verified by a number of experimental
(38)
effects, in particular, anomalous decays of pseudoscalar bound
states (neutral pion and parapositronium) into two photons.
5 It is the assumption of the coordinate time x0 as an observable in
SR that is used in the island universe [20] and in so-called global-
time theory [29].
PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006
 TETRAD FORMALISM AND REFERENCE FRAMES 111

tion P(0) = 0 (the similar constraint in GR is referred to Both action (43) and interval (44) are invariant with
as a minimal surface [20]) coincides with the Newto- respect to the general coordinate transformations
nian action µ µ µ 0 1 2 3
τ2
x x̃ = x̃ ( x , x , x , x ). (45)
2
P(i) They serve as a generalization of the action consid-
S SR [ e = 1, P ( 0 ) = 0 ] = S Newton = ∫ dτ P ( i ) Ẋ ( i ) – -------- (42)
2m ered above and as an interval for a relativistic particle,
τ1 invariant with respect to the reparametrization group of
up to a constant factor, as well as with the Einstein coordinate time. In the case of the homogeneous
action (13) in the nonrelativistic limit, where the time approximation,
X(0) in the rest frame coincides with the time interval s. 2 0 2 0 0 2 i 2
2 2
ds = a ( x ) [ N 0 ( x ) ( d x ) – ( d x ) ], (46)
Theory (41) under the constraint = + m2 P(0) P(i)
the GR action reduces to the Hamiltonian cosmology
reduces to SR because the gauge symmetry is restored. action [39–41]:
2.2.8. Relativity as a principle of gauge symmetry.
Poincaré [33] and Einstein [34] found that, in contrast S cosmic–1915
to classical mechanics, just the two observers are 2 (47)
Pϕ
∫ d x – P ϕ ∂ 0 ϕ + N 0 ⎛ --------
- – ρ 0 ( ϕ )V 0⎞ .
0
needed to completely describe the motion of a relativ- =
istic particle: the first is at rest, and the second moves ⎝ 4V 0 ⎠
with the particle. For example, every Einstein observer Here, V0 is the volume and the matter energy density is
measures its proper lifetime of an unstable particle. approximated by the homogeneous density ρ0(ϕ)
Therefore, time is a relative quantity. dependent only on the scale factor
Einstein described this time relativity as a pure kine- 0 0
matic effect by using the Lorentz transformations from ϕ ( x ) = ϕ 0 a ( x ). (48)
a fixed reference frame to a moving one.
In the homogeneous approximation, the GR equa-
As was shown above there exists a geometrody- tions of motion, as well as the conformal time interval
namic generalization of the Poincaré–Einstein dynam-
0
ics to the gauge theory with constraint (21) that allows x
us to describe this two-time relativity as a consequence
∫ dx N ( x ),
0 0
dη = N 0 ( x )d x , η =
0 0
0 (49)
of the equations of motion rather than the Lorentz kine-
matic transformations. This geometrodynamic descrip- 0

tion defines the new two-time relativity as a ratio of the are invariant with respect to reparametrization of the
dynamical particle evolution parameter X0 to geometric coordinate parameter, x0
0 0 0
x = x ( x ) , which serve
interval s (22).
as gauge transformations. As shown above, the rep-
We now illustrate this inference with a mini-uni- arametrization group of the coordinate parameter
verse. In this case, purely relativistic effects can’t be means that one of the variables (here, the only variable
described kinematically by transformations of Lorentz- is ϕ) is identified with time as a variable, while the
type variables. momentum Pϕ is taken as the corresponding Hamilto-
nian function, with its value on the equations of motion
3. RELATIVISTIC QUANTUM MINI-UNIVERSE becoming the energy of events.
3.1. Homogeneous Approximation
of General Relativity 3.2. Hilbert Variational Principle
As shown in the preceding section, relativistic Variation of action (47) with respect to the lapse
effects in variational equations can be dynamically function N0, δScosmic /δN0 = 0, yields the energy con-
described within the framework of SR formulated by straint equation
analogy to the Hilbert’s variational description of GR 2 2
[15]. According to Hilbert, GR geometrodynamics is P ϕ = E ( ϕ ), (50)
based on the two basic notions: the action where
2
ϕ0 E ( ϕ ) = 2V 0 ρ 0 ( ϕ )
∫d x
– g – ------ R ( g ) +  matter , (51)
4
S GR =
6
(43) is treated by a terrestrial observer as the universe energy.
2 3 2 Equation (50) gives two values, positive and nega-
ϕ 0 = ------M Planck
8π tive, of universe energy:
in the variables of a field space of events and the geo- ±
P ϕ = ± E ( ϕ ) = ± 2V 0 ρ 0 ( ϕ ). (52)
metric interval of the Riemannian coordinate manifold
µ ν Variation of action (47) with respect to momentum
ds = g µν d x d x . (44) Pϕ, δScosmic /δPϕ = 0, yields a geometric relation

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

112 ZAKHAROV et al.

between the two times in the form of the Friedmann dif- For each of these states, Eq. (57) can be solved in
ferential equation: terms of the conformal time η, under the initial condi-
tions a(η0) = 1 and a'(η0) = H0:
± dϕ
P ϕ = 2V 0 ------ ≡ 2V 0 ϕ' = ± 2V 0 ρ 0 ( ϕ ). (53)
dη a rigid ( η ) = 1 – 2H 0 r , a rad ( η ) = 1 – H 0 r,
The integral solution 1 2
1 (59)
ϕ ϕ a M ( η ) = 1 – --- H 0 r , a Λ ( η ) = ------------------,
dϕ̃ dϕ̃ 2 1 + H0r
± ∫
η ( ϕ I ϕ ) = 2V 0 --------------
P ϕ ( ϕ̃ )
∫
- = ± ------------------
ρ 0 ( ϕ̃ )
(54)
where r = η0 – η. The conformal time dη is defined in
ϕ ϕ
I I
observational cosmology as the instant of time at which
of this equation is referred to as the Hubble law [37, 38]. the photon is radiated by an atom at a cosmic object
moving with the velocity c = 1 in a geodetic line on the
world cone:
3.3. Reduction of Action
2 2 2 2
The reduction of Hilbert action (47) implies the cal- ( ds ) = a ( η ) [ ( dη ) – ( dr ) ] = 0, (60)
culation of its values on solutions (52) of constraint
equation (50): 2 2 2
where r = x 1 + x 2 + x 3 is the radius. This allows us to
±
S cosmic–1915 ± = S cosmic–1905 find the relation of the distance
Pϕ = Pϕ
(55) η0
= − ∫
+ 2V 0 dϕ ρ 0 ( ϕ ). r(η) = ∫ dη̃ ≡ η 0 – η, (61)
Reduced action (55) leads to equations not containing η
a geometric interval. To solve the initial geometrody-
covered by a photon to its conformal time η0 – η. Here,
namics completely, we should supplement the reduced
η0 is the present-day conformal time of the photon mea-
theory with relation (54) between the geometric interval
sured by a terrestrial observer with a(η0) = 1, and η is
and time as a variable. As will be shown below, relation
the time instant at which the photon is radiated by an
(54) describes classical cosmology, i.e., the Hubble law,
atom at the distance r from the Earth. Therefore, η is
and is an auxiliary relation to the Wheeler–De Witt quan-
the difference between the present-day conformal time
tum cosmology, provided that this cosmology is defined
η0 and the time of a photon flight to the Earth, coincid-
as a quantization of constraint equation (50). The quanti-
ing with the distance. Equation (61) gives
zation is defined by the substitution of variables by oper-
ators, Pϕ P̂ ϕ = – id/dϕ , acting on the Wheeler–De η = η 0 – r. (62)
Witt wave function Ψ [39]:
In observational cosmology, density (58) can be
2 2 expressed in terms of the present-day critical density
P̂ ϕ Ψ – E ( ϕ )Ψ = 0. (56)
2 2
We will show below that both classical [37] and ρcr = ϕ'2 ≡ ϕ2(ϕ'/ϕ)2 = ϕ 0 H 0 ,
quantum [39–41] cosmologies follow from the Hilbert
geometrodynamics. This allows us to combine them to ρ 0 ( a ) = ρ cr Ω ( a ),
settle their troubles, namely, the quantization of classi- –2 4
(63)
Ω ( a ) = Ω rigid a + Ω rad + Ω M a + Ω Λ a
cal cosmology and the description of the Hubble law in
quantum cosmology. and the relative ones Ωrigid, Ωrad, ΩM, and ΩΛ, satisfying
the condition Ωrigid + Ωrad + ΩM + ΩΛ = 1 [37].
3.4. Friedmann Equation and Classical Cosmology Classical cosmology describes the redshift of the
We now consider universe evolution equation (53), radiation spectrum E(η) at a cosmic object relative to
2 2
the spectrum E(η0) at the Earth. Redshift is defined as
ϕ 0 a' = ρ 0 ( a ) (57) the scale factor,
in a Euclidean space, where a = ϕ/ϕ0. The universe is E(η) –1
filled with homogeneous matter with the density ρ0 -------------- = a ( η ) = ( 1 + z ) (64)
E ( η0 )
conformally dependent on the scale a(η):
–2 4 versus the coordinate distance (given by conformal
ρ 0 ( a ) = ρ rigid a + ρ rad + ρ M a + ρ Λ a . (58) time (62)) to the object.
Here, ρrigid is the contribution of the equation of rigid Taking these relations into account and substituting
state, ρrigid = prigid. The densities ρrad, ρM, and ρΛ a = 1/(1 + z) and η = η0 – r, we can write out scale evo-
describe the contributions of radiation, baryon matter, lution equation (57) at the light ray geodetic line dr/dη =
and Λ-term, respectively. –1 in the form

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

 TETRAD FORMALISM AND REFERENCE FRAMES 113

1 dz 2 2 –1 –4
------ ----- = ( 1 + z ) ρ cr [ Ω rigid(1 + z) + Ω rad + Ω M(1 + z) + Ω Λ(1 + z) ],
H 0 dr

where H0 = ρ cr /ϕ 0 . The solution where

1+z 1 2 2 2
dx H U = --- [ P Ψ + E ( ϕ )Ψ ] (70)
H0r(z) = ∫ ---------------------------------------------------------------------------- (65)
6
Ω rigid x + Ω rad x + Ω M x + Ω Λ
4 3
2
1 is the Hamiltonian. The definition of the energy E(ϕ) of
(coinciding with solution (54)) of this equation deter- the universe makes it possible to write out the Hamilto-
mines the coordinate distance as a function of redshift nian HU as the product of E(ϕ) and the universe number,
z and gives formulas (59) for every state. Formula (65), + –
being a basis of observational cosmology (for example, NU = A A , (71)
see [37]), is used to find the equation of the state of the
1 + – – + 1
universe matter from astrophysical red-shift data under H U = E ( ϕ ) --- [ A A + A A ] = E ( ϕ ) N U – --- , (72)
the assumption of flat space. We here consider the effect 2 2
of the quantization on Hubble law (54) in observational where the holomorphic variables
cosmology defined by (65), namely, the consequences
of the Hamiltonian action reduction and cosmological 1 (+) (–)
time interval. Ψ = -------------------- { A + A },
2E ( ϕ )
(73)
E ( ϕ ) (+) (–)
3.5. Quantum Cosmology as a Quantization PΨ = i ------------ { A – A }
of Constraints 2
3.5.1. Primary quantization. The first quantization are introduced [42]. To exclude negative energies, A–
should be postulated as the annihilation operator for the
i [ P ϕ, ϕ ] = 1 (66) universe with positive energy; this implies the existence
of the cosmological scale ϕ means that energy con- of the vacuum as a lowest-energy state:
straint equation (50) turns into Wheeler–De Witt equa- (–)
A |0〉 A = 0. (74)
tion (56) for the universe moving in the space of events ϕ:
2 2 However, the universe number NU = A+A– is not con-
∂ ϕ Ψ + E ( ϕ )Ψ = 0. (67) served, because the energy E(ϕ) depends on ϕ. It is just
This equation can be derived by variation of the cor- the ϕ-dependence of E(ϕ) that leads to an extra term in
responding Klein–Gordon-like field theory, the action written in terms of the holomorphic variables
in functional space:
1
∫ dϕ --2- [ ( ∂ Ψ ) ∫ PΨ ∂ϕ Ψ
2 2 2
SU = ϕ – E ( ϕ )Ψ ] ≡ dϕL U , (68)
i + – + – i + + (75)
which will be referred to as the field theory of uni- = --- ( A q ∂ ϕ A – A ∂ ϕ A ) – --- ( A A – AA )∆ ( ϕ ) ,
verses. 2 2
The negative energy in solutions (52) means that the where
energy of the relativistic system under consideration
has no minimum; therefore, an arbitrary small interac- ∂ϕ E ( ϕ )
∆ ( ϕ ) = -----------------
-. (76)
tion causes this system to be unstable. In the quantum 2E ( ϕ )
field theory resulting from the second quantization of
the Wheeler–De Witt equations for Ψ, a system can be The last term in Eq. (75) describes the cosmological
made stable if the vacuum is postulated as a lowest- creation of the universe from the vacuum if ∂ϕE(ϕ) ≠ 0.
energy state. As shown in the preceding section con- 3.5.3. Bogoliubov transformation and universe cre-
cerning the second quantization of a particle, such a ation. To define the vacuum and a set of conserved
vacuum originates if the creation of the universe with numbers, so-called integrals of motion, we can use the
negative energy is treated as the annihilation of the uni- Bogoliubov transformations [43] of the variables (A+, A–)
verse with positive energy. (similar to the case of cosmological particle creation [42]),
3.5.2. Secondary quantization. By introducing the + +
A = αB + β*B ,
– –
A = α*B + β A
– +

canonical momenta PΨ = ∂L U /∂(∂ ϕ Ψ) , we obtain an 2 2

(77)
action in the Hamiltonian form: ( α – β = 1 ),
so the corresponding equations, expressed in terms of
SU = ∫ dϕ { P Ψ ∂ϕ Ψ – H U }, (69) the universes (A+, A–),

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

114 ZAKHAROV et al.

+
( i∂ ϕ + E ) A = i A ∆ ( ϕ ),
– dN
⎧ ----------U = ∆ ( ϕ ) 4N U ( N U + 1 ) – R U2
(78) ⎪ dϕ
–
( i∂ ϕ – E ) A = i A ∆ ( ϕ ),
+
⎨ (88)
⎪ d---------
RU 2
take the diagonal form in terms of the quasiuniverses ⎩ dϕ- = – 2E ( ϕ ) 4N U ( N U + 1 ) – R U
(B+, B–),
under the initial conditions NU(ϕ = ϕI) = RU(ϕ = ϕI ) = 0.
+ –
( i∂ ϕ + E B )B = 0, ( i∂ ϕ – E B )B = 0. (79) In the model of the equation of rigid state with the
energy E(ϕ) = Q/ϕ, functional (86) can be found explic-
This means that the Bogoliubov transformation itly because Eqs. (88) can be solved exactly. In this
coefficients satisfy the Bogoliubov equations case, the distribution function takes the form
( i∂ ϕ + E )α = iβ∆ ( ϕ ), 1 2 2 1 ϕ
(80) N U = ------------------ sin Q – --- ln ----- ≠ 0, (89)
( i∂ ϕ – E )β* = iα*∆ ( ϕ ). 4Q – 1
2 4 ϕI
Introducing the variables r and θ,
where ϕ = ϕ I 1 + 2H I η , where ϕI and HI = ϕ 'I /ϕ I =
iθ ( ϕ ) iθ ( ϕ )
α = e cosh r ( ϕ ), β* = e sinh r ( ϕ ), (81) 2
Q/(2V 0 ϕ I ) are initial data.
(referred to as a shift parameter and a rotation parame- 3.5.4. Quantum anomaly of conformal time. The
ter, respectively), we arrive at the equations vacuum existence postulate restricts the motion of the
( i∂ ϕ θ – E ( ϕ ) ) sinh 2r = – ∆ ( ϕ ) cosh 2r sin 2θ, universe in the field space of events and means that the
(82) universe moves forward (ϕ > ϕI) or backward (ϕ < ϕI)
∂ ϕ r = ∆ ( ϕ ) cos 2θ. if the energy of events is positive (Pϕ ≥ 0) or negative
(Pϕ ≤ 0), respectively, where ϕI are initial data. In quan-
The quasiuniverse energy entering into Eqs. (79) is tum theory, the quantity ϕI is considered as a creation
given by the expression point of the universe with positive energy Pϕ ≥ 0 or as
E ( ϕ ) – ∂ϕ θ an annihilation point of the anti-universe with negative
E B ( ϕ ) = --------------------------
-. (83) energy Pϕ ≤ 0. We can assume that the singular point
cosh 2r ϕ = 0 belongs to the anti-universe: Pϕ < 0. A universe
By virtue of Eqs. (79), the quasiuniverse number B = with positive energy of events has no the cosmological
(B+B–) is conserved: singularity ϕ = 0.
According to the vacuum existence postulate, solu-
d B d ( B+ B– ) tion (54) takes the form
----------- ≡ -------------------- = 0. (84)
dϕ dϕ ϕ0
dϕ
Therefore, we arrive at the definition of the vacuum
as a state of quasiuniverses in the form
∫
η ( ϕ I, ϕ 0 ) = θ ( P ϕ ) ------------------ θ ( ϕ 0 – ϕ I )
ρ0 ( ϕ )
ϕI
– ϕI
(90)
B |0〉 U = 0. (85) dϕ
The number of universes created from the Bogoli- ∫
+ θ ( – P ϕ ) ------------------ θ ( ϕ I – ϕ 0 ) ≥ 0,
ρ0 ( ϕ )
ϕ0
ubov vacuum can be found by averaging operator (71)
of the universe number (wheelerons) over the Bogoli- where θ is the Heaviside step function. Therefore, this
ubov vacuum. Therefore, this number is proportional to postulate leads to conformal time (90) being positive
the Bogoliubov coefficient squared defined in Eq. (77): both for the universe (Pϕ > 0, ϕ > ϕI) and for the anti-
+ – 2 universe (Pϕ < 0, ϕI < ϕ); i.e., it results in the conformal
NU(ϕ) = U〈 A A 〉U ≡ β . (86) arrow of time (90) η > 0.
This quantity can be referred to as a functional of
universe distribution NU(ϕ), whereas the quantity 3.6 Levi-Civita Transformations and Hubble Law
R U ( ϕ ) = i ( α*β* – αβ ) ≡ U 〈 P Ψ Ψ〉 U The conformal time η and corresponding momen-
(87) tum Π can be introduced, respectively, as a new field
= – sinh 2r sin 2θ variable and a proper energy of geometric space of events
can be referred to as a rotation functional, because the if the Levi-Civita canonical transformation [27, 44],
variables r and θ are shift and rotation parameters, (Pϕ |ϕ) (Π|η), is used to turn energy constraint (50)
respectively. into the new canonical momentum Π. Let us consider
In terms of the functional of universe distribution this transformation for the universe filled with photons
NU(ϕ) and the rotation functional RU(ϕ), the Bogoli- with ρ0(ϕ) = const. In this case, the transformation
ubov equations take the form takes the form Pϕ = ± 2 ΠV 0 and ϕ = ± ( 1/2 ) Π/V 0 η ,

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

 TETRAD FORMALISM AND REFERENCE FRAMES
∫
0
with action (47) reducing to Sc = d x [–Π∂ 0 η + N 0 (Π –
ρ 0 V 0 )] .
Solving the constraint equation Π – V0ρ0 = 0 means
that the nonzero energy Π = V0ρ0 corresponds to the
conformal time.
In this case, the reduced action takes the form S =
η0
–V0 ∫0
dηρ 0 = – V 0 ρ 0 η 0 . In quantum theory, with the
geometric energy Π substituted by the operator Π̂ =
id/dη , the geometric evolution of wave function is
determined from the quantum version of the constraint
equation ( Π̂ – V0ρ0)ψgeometric(η) = 0, which has the
solution
– iV 0 ρη
ψ geometric ( η ) = e θ ( η ). (91)
The Hubble evolution ϕ = ϕ(η) could be considered
as a relativistic effect associated with two complemen-
tary methods of describing the relativistic universe,
namely, by using either field or geometric wave func-
tion, exp(–iPϕϕ) or (91).
3.7. Quantization Results
Thus, the unified geometrodynamic formulation of
both the theories (SR and GR), which is based on the
Hilbert variational principle [15], makes it possible to
quantize the cosmological models similarly to the first
and second quantization of a relativistic particle. The
latter is a basis of the modern quantum field theory [18],
which is verified by a great number of high-energy
experiments. A similar approach to the quantization
within the framework of GR was first formulated by
Wheeler and De Witt [39]. They assumed that cosmo-
logical time, treated as a variable, be identical with the
cosmological scale factor. Moreover, they introduced
into GR the notion of a field space of events, in which
the relativistic universe moves, by analogy to the notion
of Minkowski space. However, the Wheeler–De Witt
formulation [39] does not contain time as a geometric
interval and, therefore, its scale-factor dependence
(interpreted in the Friedmann cosmology as the Hubble
law). Thus, as noted above, classical cosmology fails to
quantize [37], while quantum cosmology fails to
describe the Hubble law [39–41].
In this section, we use the invariant reduction of the
Wheeler–De Witt cosmology [44], considered as a rel-
ativistic-universe geometrodynamics, to restore the
relation of observational cosmology (i.e., the Hubble
law) to the first and second quantization of the universe
and calculate the distribution of created universes. This
reduction allows us to solve a series of problems,
namely, Hubble evolution, universe creation from the
vacuum, arrow of time, initial data, and elimination of
the cosmological singularity, under the assumption that
the Hamiltonian is diagonal and the universe is stable.
PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1
115
In this case, we attain the physical description of nature
that looks like quantum field theory in particle physics.
In what follows, we consider a similar invariant reduc-
tion in GR in order to define physical observables,
quantize gravity, and formulate a low-energy perturba-
tion theory.
4. REFERENCE FRAMES IN GENERAL
RELATIVITY IN METRIC FORMALISM
4.1. Dynamics and Geometry of General Relativity
We now present the general gauge-invariant formu-
lation of GR in metric formalism based on the action
ϕ
2
∫
S [ ϕ 0 F ] = d x – g – -----0- R(g) +  matter(ϕ 0 g, f ) (92)
4
6
depending on the fields F = (g, f ) and the geometric
interval
2 µ ν
ds = g µν d x d x . (93)
Here, the Newton constant
2 3 2
ϕ 0 = ------M Planck
8π
defines a mass scale.
Action (92) and interval (93) are invariant with
respect to the general coordinate transformations
µ µ µ 0 1 2 3
x x̃ = x̃ ( x , x , x , x ). (94)
If these transformations are treated as gauges and
the corresponding four equations for g0α are considered
as constraint equations similar to the Gauss equations
in electrodynamics, then the coordinates x µ can not be
taken as measurable quantities. This was illustrated
above with the similar formulation of SR. To define
truly measurable coordinates one should solve con-
straint equations defined in a specific reference frame.
Reference frames, in relativistic theory, are associ-
ated with corresponding sets of devices measuring ini-
tial data [22]. As noted above, the choice of a reference
frame means variable, coordinates, and equations of
motion should be subdivided into space and time com-
ponents. In this case, the number of variables entering
into the action is preserved. In the Hamiltonian
approach, time components of metric tensor serve, as a
rule, as Lagrange multipliers. Variation of the action
with respect to these multipliers yields the above-men-
tioned constraint equations for the initial data.
4.2. Dirac Hamiltonian Approach
Until now, most Hamiltonian approaches to GR
were based on the Dirac–ADM formulation [23, 24].
According to this formulation, the choice of a reference
frame is a classification of coordinates and variables
such that interval (93) takes the form
2006
116 ZAKHAROV et al.

2 µ ν 1/3 ij
ds = g µν d x d x The transversality conditions, ∂j ( γ γ )
0, and the
(95) Dirac gauge [23]
0 2 i i 0 j j 0
= [ γ N d d x ] – γ ij ( d x + N d x ) ( d x + N d x ). 2
ϕ
Here, γij is the metric of the three-dimensional hyper- Tr P γ = – -----0- Tr D
space embedded into the four-dimensional manifold, N j 3
(100)
is the three-dimensional origin shift vector depending γ0
2
k
on time, and Nd is the Dirac lapse function [23] (see a - [ ∂ 0 γ – ∂ k ( γ N ) ]
0,
≡ – ----------------
more detailed consideration in monograph [22]). In 3N d γ
terms of these variables, action (92) takes the form defining the minimal spatial hypersurface, both can
∫ d x[K(ϕ
S [ ϕ0 F ] =
4
g ) – P ( ϕ0 g ) serve as examples.
0
(96) The Hamiltonian form of action follows from
+ S ( ϕ 0 g ) + – g  matter ( ϕ 0 g, f ) ]. Eqs. (97) and (98) [23]:
Sh [ ϕ0 F ]
Here,
0
x2
⎧ 0⎧ ⎫ (101)
∫ ∫ ∑P +  – N d  t ( ϕ 0 F ) ⎬.
3
F ∂0 F
2
⎪ K ( ϕ g ) = N ----- ϕ0 2 2 = dx ⎨ d x
⎪ 0 d - [ Tr D – ( Tr D ) ] ⎩ ⎭
6 0 F
⎪
x1

⎪ 6 1
∑ ∑
ij
ij 2
⎪ = P γ D ij – N d ------2 Tr P γ – --2- ( Tr P γ )
2
Here, F
P F ∂ 0 F = P γ ∂ 0 γ ij + f
P f ∂0 f ;
⎪ ϕ0
⎨ (97)  = N T k ( ϕ0 F ) + C0 D + Ci ∂ j ( γ
k 0 1/3 ij
γ ) (102)
⎪ ϕ0
2
(3)
⎪ P ( ϕ 0 g ) = N d ------ γ R is the constraint sum with the Lagrange multipliers N k,
⎪ 6
C0, and Ci;
⎪ 2 2
⎪ ϕ0 ϕ0
t ( ϕ0 F ) = γ T 0 ( ϕ0 F )
k i 0
⎪ S( ϕ 0 g ) = -----
- [ ∂ D – ∂ ( N D ) ] – -----
- ∂ ( γ ∂ ( γ N d) ) (103)
3 0 k
3 i
⎩
is the Dirac Hamiltonian; and
are kinetic, potential, and surface terms, respectively;
⎧ 0 6 2 1 2
1 ⎪ T 0 ( ϕ 0 F ) = ------------2 Tr P γ – --2- ( Tr P γ )
D ij = --------- ( ∂ 0 γ ij – ∇ i N j – ∇ j N i ) ⎪ γ ϕ0
2N d ⎪
(98) ⎪ ϕ 02 ( 3 )
1 0
⎪ + ------ R + T 0matter
l l
= --------- ( ∂ˆ 0 γ ij – γ il ∂ j N – γ jl ∂ i N )
2N d ⎨ 6 (104)
⎪ 0 j 0
are the covariant rates of change of the metric (i.e., sec- ⎪ T k ( ϕ 0 F ) = 2∇ j P γk + T k matter
ond quadratic forms); and ⎪
⎪ ϕ 02 j 0
ϕ
2 ⎪ = ------ [ ∇ j D k – ∂ k Tr D ] + T k matter
k k k
P γi = -----0- [ D i – δ i SpD ] ⎩ 3
6
(99) are the energy–momentum tensor components. The
2
⎛Pk ϕ
≡ Tr P γ = – -----0- SpD⎞
identification of these tensors with observables means
⎝ γk 3 ⎠ that the observable evolution parameter is identified
with noninvariant coordinate time x0. In this case, as
are the corresponding momenta of the spatial metric. was shown in [20], the perturbation theory is based on
These momenta are defined with the use of the Leg- the Euclidean metric Nd = 1, γij = δij and can be
endre transformation zD2/2 = PD – z–1P2/2 in the K-term described by the effective action
in (97). The quantity R(3) is the three-dimensional cur-
vature. S [ ϕ0 F ] (3)
= S Bjern
g Nd
1
Because of GR symmetry (92) with respect to trans- 0
x2
formations (94), treated as gauges, the theory has extra, 0⎧ ⎫ (105)
∫ ∫ ∑ P F ∂ 0 F +  – γ T 0 ( ϕ 0 F ) ⎬.
3 0
purely gauge, degrees of freedom. These degrees of = dx ⎨ d x
freedom can be eliminated by using gauge transforma- 0
⎩ F ⎭
x1
tions. The condition that decreases the number of fields
and metric components with the help of a symmetry This action, first written by Bjorn [29], results from
group of interval and action is referred to as a gauge. GR action in so-called synchronous reference frame

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

 TETRAD FORMALISM AND REFERENCE FRAMES
γ N d
1. (106)
In fact, this condition concerning the synchronous
reference frame should be considered, by Dirac [26], as
a gauge that can only be used in equations of motion as a
weak equality. Substituting gauge (106) into action (101),
we arrive at a new theory (105), which is not equivalent
to initial one (101) and could coincide with it only in a
nonrelativistic limit. This was illustrated with the GR
absolute time in Section 2.2.7. The conventional pertur-
bation theory based on the Euclidean metric Nd = 1,
γij = δij , and theory (105) contains no energy con-
straints. Therefore, they can not solve the problems
mentioned above concerning the initial data, energy
localization, arrow of time, and elimination of cosmo-
logical singularity. It is well known [20] that the Dirac
GR formula considered above encountered just these
problems. Moreover, minimal surface condition (100)
contradicts to the experimental observations on the dis-
tance-dependence of redshift (i.e., the Hubble law) pro-
vided that this dependence is related to the dilation pro-
portional to the quadratic form: SpD ~ H0 ≠ 0, where
H0 is the Hubble constant.
These problems can be solved by using the experi-
ence of solving the problems mentioned above of rela-
tivistic mechanics and cosmology, as shown in [27, 28,
45–47]. This experience is based on the following sub-
jects: the determination of gauge symmetry of GR ref-
erence frames, where the Hamiltonian formula (101) is
defined; the definition of invariant evolution parame-
ters; and the invariant reduction of the theory in a spe-
cific reference frame.
4.3. Gauge Symmetry of Reference Frames
in General Relativity
In the reference frame considered above, the gauge
group of the Hamiltonian approach is the diffeomor-
phism group of metric parameterization (95) [22, 25]:
x
0 0 0
x̃ = x̃ ( x ),
0 zero Fourier harmonic), and log γ = log γ – 〈 log γ 〉 is
0
xi x̃ i = x̃ i( x , x 1, x 2, x 3), (107)
0 0 i
i ∂x̃ d x ∂x̃ ∂x
k k
dx k
Ñ = N --------0 , Ñ = N -------- --------0 – -------- --------0 . (108)
d x̃ ∂x i d x̃ ∂x i ∂x̃
This transformation group conserves the set of
hypersurfaces x0 = const and is referred to as a kine-
matic subgroup [22, 42] of the group of general coordi-
nate transformations (94). The group of kinemetric
transformations contains the reparametrization of coor-
dinate time (107).
4.4. Class of Functions of Dynamical Variables
There are the two physical consequences of GR
invariance with respect to the reparametrization of
coordinate time (107). Firstly, such an invariance
PHYSICS OF PARTICLES AND NUCLEI Vol. 37
117
means that no physical instruments exist which could
measure the coordinate time x0. If this invariance is
treated as a gauge principle, then (as illustrated above
with SR) the Dirac Hamiltonian considered as a sum of
constraints, as well as coordinate evolution parameter x0,
is not a gauge invariant. As shown in [27, 28, 45–47], the
invariance with respect to the reparametrization of
coordinate time in GR means that reference frame (95)
should be predetermined by choosing two invariant
observable times, namely, one as a variable and the
other as an interval. The SR example considered in the
preceding section shows that the homogeneous degree
of freedom makes it possible to introduce in GR a
gauge-invariant dynamical evolution parameter, i.e.,
time as a variable.
Secondly, the function class on which the parame-
ters of transformations of physical variables are defined
should coincide with the function class on which the
physical variables are defined. This means that physical
variables in reference frame (95) include the class of
homogeneous functions (107) depending only on coor-
dinate time f(x0). In the perturbation theory, these
homogeneous degrees of freedom can be treated as zero
Fourier harmonics, which are picked out by averaging
∫
3
over spatial volume V0 = d x .
In particular, the logarithm of spatial metric deter-
0 i
minant log γ ( x , x ) can be written out as the sum
0 i 0 0 i
log γ ( x , x ) ≡ 〈 log γ 〉 ( x ) + log γ ( x , x ). (109)
Here,
1
∫
0 3 0 i
〈 log γ 〉 ( x ) = ------ d x log γ ( x , x ) (110)
V0
V0
∫
3
is the averaging over spatial volume V0 = d x (i.e., a
the deviation satisfying the strong condition
∫ d x log γ ≡ ∫ d x [ log γ – 〈 log γ 〉 ] ≡ 0,
3 3
(111)
with accordance with (110).
The similar zero Fourier harmonic also enters into
the spur of canonical momenta of metric components:
TrP γ = 〈 TrP γ 〉 + TrP γ , (112)
where
2 2
ϕ ⎛ TrP = – ϕ
-----0- Sp D⎞ . (113)
k k k
P γi = -----0- [ D i – δ i TrD ]
6 ⎝ γ
3 ⎠
It is known that, as a rule, the infrared spectrum 〈Pγ〉
in gauge theories is nonperturbative, while deviations
from it P γ are to be calculated by perturbation theory.
No. 1 2006
118 ZAKHAROV et al.

4.5. Universe Evolution As a Zero Mode Solution
of the Constraint Equation
The separation of actual dynamical variables from
nondynamical ones is a crucial stage of the extraction
of the physical information adequate to symmetry prin-
ciples, as well as the variational equations including
both equations of motion and constraint equations. This
goal can be realized by two methods, namely, the con-
ventional method of imposing gauge conditions [23]
(type of Dirac’s minimal surface condition (100)) and
the method of explicitly solving the variational equa-
tions with constraints. The latter was developed by
Dirac, who considered the Gauss equations of quantum
electrodynamics as an example [11].
The explicit solution of the variational constraint
equations, similar to the Gauss conditions in electrody-
0
namics or the conditions T k ( ϕ 0 F ) = 0 in GR, makes
it possible to allow for the dynamics in the function
class that is defined by the gauge transformation group
and by the condition of finite energy density. For exam-
ple, the Gauss condition ∂1E(x0, x1) = 0 in the
Schwinger two-dimensional electrodynamics was
shown in [48] to have the explicit, nontrivial, homoge-
neous solution E(x0, x1) = E0(x0), so-called zero mode,
which determines the topology and mass spectrum of
the theory.
A similar, nontrivial, homogeneous solution of the
Gauss equation
δS GR 0 i
-----------k ≡ T k = 2∇ i P γk = 0
δN
also exists in GR, where SGR is metric action (92) in GR.
As is seen, the zero Fourier component of metric
momentum (112)
i 1 i
[ P γk ] particular = --- δ k 〈 TrP γ 〉 ≠ 0
3
arises as a particular solution of the Gauss equation
0 2 i 2
T k = --- ∇ i [ δ k 〈 TrP γ 〉 ] = --- ∂ k 〈 TrP γ 〉 = 0.
3 3
If all the variables, except for the scalar factor a(x0),
0
are ignored, then the tensor T k ( ϕ F ) is identically
equal to zero. Therefore, the scale factor a(x0) also can
be referred to as a zero mode solution of constraint
equation (117).
As was shown in Section 4, it is the factor a(x0) that
is unambiguously related to cosmological evolution
and to the time as a dynamical variable [45]. To avoid
double counting of variables in the conventional cos-
mological perturbation theory [49] the strong condition
∫
3
d xD ≡ 0 must be imposed, where D is determined by
Eq. (98) conserving the number of GR variables [47].
In what follows we show that the allowance for the
0
zero mode of the Gauss condition T k ( ϕ F ) = 0 makes
it possible to combine the three GR branches: nonivari-
ant island-universe theory [20] (in which the cosmolog-
ical scale factor is identically equal to unity and the
symmetry with respect to the reparametrization of the
coordinate time is violated by its global synchroniza-
tion), Hamiltonian cosmology [41] (in which the cos-
mological scale factor is treated as a dynamical vari-
able), and cosmological perturbation theory (in which
the variable similar to the cosmological scale factor is
allowed for two times [49–51]).
4.6. Separation of Zero Mode and Reduction of Action
Using transformation (118) we rewrite action (92) in
(114) the form
2
0 ( ∂0 ϕ )
N0∫
S [ ϕ 0 F ] = S [ ϕ F ] – d x ---------------
-. (119)
Here, Sh[ ϕ F ] coincides with Dirac action (101), the
fields F are substituted by the fields F with running
(115) masses ϕ, the quantity
0 0
ϕ ( x ) = ϕ0 a ( x ) (120)
is the running mass scale, and the equation
(116)

∫ d xN
0 –1 –1 3 –1 0 i –1
The existence of such a homogeneous solution of N0( x ) = V0 d (x , x ) ≡ 〈 Nd 〉 (121)
the Gauss equation in GR follows from the invariance
V0
of the equation
–1
0 0
T k ( ϕ0 F ) = T k ( ϕ F ) = 0 (117) is the averaging of the inverse lapse function N d over

∫
3
spatial volume V0 = d x . Therefore, the lapse function
with respect to scale transformations of the metric and
all the fields, with a scale defined in the class of homo- N0 determines the geometric time ζ
geneous functions 0
x
(n)
∫ dx N ( x ),
(n) 0 0 0
dζ = N 0 ( x )d x , ζ( x ) =
n 0 2 0 0 0
F = F a ( x ), g µν = g µν a ( x ). (118) 0 (122)
Here, (n) is the conformal weight and ϕ(x0) = ϕ0a(x0) is
the running mass scale. in reference frame (152).

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

 TETRAD FORMALISM AND REFERENCE FRAMES 119

The Hamiltonian action (up to a total derivative) SR and cosmology examples. The quantum anomaly
takes the form implies an absolute origin of geometric time.
2
Pϕ
∫
0
S h [ ϕ 0 F ] = S h [ ϕ F ] – d x P ϕ ∂ 0 ϕ – N 0 --------
- , (123) 4.7. Problem of Initial Data in General Relativity
4V 0
where the canonical momentum The initial data ϕI and HI = ϕ I' /ϕ I can be treated as
a creation (an annihilation) point of the universe with a
P ϕ = 2V 0 ϕ' ≡ 2V 0 dϕ/dζ positive (negative) energy Pϕ ≥ 0 (Pϕ ≤ 0).
defines the energy in the space [ ϕ F ] of events. The initial value problem in GR is a problem of find-
ing the dependence of metric and fields on a coordinate
δS [ ϕ F ] time from the equations invariant with respect to the
The energy constraint --------------------- = 0 yields the equa- reparametrization of this time. To solve this problem
δN d
we pick out the invariant dynamical evolution parame-
tion ter ϕ as a scale factor and show that the action contains
2
N 0 Pϕ
2 the invariant homogeneous geometric interval dζ =
2
- =  t ( ϕ F ),
------------ (124) 0
N 0 d x corresponding to this parameter. The extraction
Nd
2 2
of the scale factor a = ϕ/ϕ0 ( ϕ 0 = M Planck 3/8π ) by the
where t( ϕ F ) = γ
0
T 0 (ϕ
F) coincides in form with n (n)
conformal transformation F(n) = a F , where n is the
expression (103) after the corresponding change of
conformal weight, allows us to describe the universe
variables ( ϕ 0 F ) ( ϕ F ). This equation has the motion covering a hypersurface in the field space
solution [ ϕ F ] of events (Fig. 2).
〈 〉 These equations allow us to unambiguously deter-
N d = N 0 ---------------t - , (125) mine the dependence of the fields F on the dynamical
t
evolution parameter ϕ if the initial data F I (x i ) = F (ϕI , x i )
P ϕ ( ± ) = ± 2V 0 ϕ' = ± 2V 0 〈  t〉 . (126) and PFI (x i ) = PFI (ϕI , x i ) are given. Here, ϕI and PϕI are
treated as the coordinates of a point of universe creation
Substituting this solution into action (123), we arrive (or annihilation) with a positive energy E = Pϕ ≥ 0 in the
at the relativistic-universe action
phase space.
S± [ ϕ I ϕ0 ] energy constraint In what follows we assume that both GR and SR can
ϕ0 be interpreted only within the framework of quantum
⎧ 3 ⎫ (127) theory with the vacuum as a lowest-energy state. The
= ∫ ⎩
∫
dϕ ⎨ d x ∑P F ∂ϕ F +  ± 2 t ( ϕ F ) ⎬
⎭
universe stability condition in the form of this postulate
ϕI F leads to the quantum anomaly and an absolute origin of
the global time, ζ = 0. Cosmology involves the problem
moving in the field space of events [ ϕ F ], where  = of defining initial data for the universe creation, ϕ(ζ =
/∂0ϕ, ϕI is the point of universe creation or annihila- 0) = ϕI and Pϕ(ζ = 0) = PϕI . Time, matter, and tempera-
ture (as a characteristic of matter motion) are also cre-
tion. Equations (127) unambiguously determine all the
ated together with the universe. For this problem to be
fields F (ϕ, x i ) as functions of the dynamical evolution solved theoretical quantities should be identified with
parameter ϕ and initial data of these fields at the time observable ones in accordance with the correspondence
instant ϕ(ζ = 0) = ϕI of universe creation. The geomet- principle for classical theory in flat spacetime.
ric time can be found from Eq. (126) [27]:
ϕ0
4.8. Low-Energy Dynamics
dϕ
〈  t〉
∫
ζ ( ϕ 0 ϕ I ) = θ ( ϕ 0 – ϕ I ) ---------------- and Correspondence Principle
ϕ I Thus, there is a direct analogy [39] between GR and
ϕI
(128) the relativistic particle dynamics considered in Section 2.
dϕ This analogy can be extended to the definition of
∫
+ θ ( ϕ I – ϕ 0 ) ---------------- ≥ 0.
〈  t〉
observables (similar to time, particle number, and one-
ϕ 0 particle energy) by using the correspondence of GR to
We here allow for the quantum anomaly of geomet- the theory of classical fields F in a flat spacetime. In
ric interval, which follows from the stability condition this case, field energy is much less than a huge energy
for the quantum system. This was illustrated above with of homogeneous cosmological density ρ0(ϕ).

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

120 ZAKHAROV et al.

ϕ UNIVERSE
ψ(s|G)
P0 = Lϕ = E
ϕ = ϕ0
ds

ψ(ϕ|F) = A+ψ + A–ψ*

ϕ = ϕ1

Fig. 2. Universe moving on its hypersurface in the world field space of events. Every observer in the universe has the two sets of
+
measurable quantities, namely, the field set (mass ϕ, particle number densities aq a q with quantum numbers q) and the geometric
+
set (time interval η and initial data for Bogoliubov quasiparticle number density bq b q ) [54].

0
The local energy density T t0 (ϕ F) entering into considered above in Section 2. The second term is the
Eq. (103) can be written out as a sum of the homoge- conventional field action written in terms of the confor-
neous cosmological density ρ0(ϕ) and the local density mal time,
of particle-like perturbations in field theory: η0

∫ dη ∫ d x ∑ P
(±) 0
+  – γ  t0 , (133)
3
0 S field = F ∂η F
F ) = ρ0 ( ϕ ) + 
0
T t0 ( ϕ t0 ( ϕ F ). (129)
ηI F

0 with the running masses m(η) = a(η)m0 describing cos-

We then expand the square root T t0 in reduced
mological particle creation from the vacuum [55]. In
action (127) in powers of the local density, the limiting case, in which the evolution of the universe
and particle creation are ignored (ϕ
ϕ0), action (133)
0
dϕ2 ρ 0 ( ϕ ) +  t0 coincides with GR action (105) in the synchronous-
time gauge:
0
 t0 (130) 0
S Bjern [ x synchro = η, T t0 synchro =  t0 ]
0 0
= dϕ 2 ρ 0 ( ϕ ) + ------------------ + …,
ρ0 ( ϕ ) (±)
(134)
= S field [ ϕ ( η ) = ϕ 0 ].
use the definition dη = dϕ/ ρ 0(ϕ) of the conformal As is seen from the comparison of invariant ac-
time, which coincides with the geometric time ζ in this tion (133) with action (105) in the synchronous-time
approximation, and present reduced action (127) in the gauge, the conformal time η serves as world observable
form6 time, just so the geometric interval in Newtonian
mechanics (42) serves as absolute observable time.
(±) (±) (±) Thus, according to the correspondence principle, the
S [ ϕ I ϕ0 ] energy constraint
= S cosmic + S field + …. (131) measured cosmological time is to be identified with the
conformal time dη rather than with the Friedmann time
Here, the first term is cosmological action (47) dt = a(η)dη.
ϕ0

ϕ0 ] = −
(±)
∫
S cosmic [ ϕ I + 2V 0 dϕ ρ 0 ( ϕ ) (132) 4.9. Relative Units of Measurement
ϕI
The identification of measured quantities with con-
formal ones corresponds to the choice of relative length
units. In this case, a homogeneous dilation of all the
6 In quantum field theory, the interaction ρ0 + ( 2 ) +  I =
spatial intervals, L = aLc implies the dilation of the
ρ 0 +  ( 2 ) +  I ρ 0 +  ( 2 ) is initially separated. This yields length units, l = alc, so that a measured interval, defined
a form factor, which makes it possible to partially eliminate ultra- as the ratio Lr = aLc /alc = Lc /lc of the two expanding
violet divergences [27]. quantities, does not vary.

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

 TETRAD FORMALISM AND REFERENCE FRAMES
This yields a cosmological model of the evolution of
the universe, in which all measurable quantities are
identified with conformal ones. Namely, the conformal
time dη, coordinate distance r, running masses m =
m0 /(1 + z), and conformal temperature Tc(z) = T(z)/(1 + z)
are identified with Friedmann time dt = a(η)dη, Fried-
mann distance R = r/(1 + z), constant masses m0, and
conventional temperature T(z), respectively. In this
case, the redshift of spectral lines of atoms at cosmic
objects,
E emission m atom ( η 0 – r ) ϕ ( η 0 – r )
- ≡ ----------------------
- = ------------------------------
----------------
E0 m atom ( η 0 ) ϕ0
1 On the other hand, cosmic evolution allows us to find
= a ( η 0 – r ) = -----------,
1+z
is explained by the running masses. In the conformal
cosmological model, the conformal observable dis-
tance r loses the factor a, in comparison with the Fried-
mann distance R = ar. In this case, the recent distance–
redshift data from Ia supernovae [52] was shown in [38]
to be compatible with the equation of rigid state
2
Q
p = ρ ( ϕ ) = ---------------
2 2
4V 0 ϕ
and with the dependence a(η) = 1 + 2H 0 ( η – η 0 ) of
the scale factor on the conformal (i.e., observable) time.
It is the dependence that is used to explain the chemical
evolution of the universe [53].
The evolution of the universe in relative units was
shown to differ from that in absolute units of Standard
Cosmology. In fact, the temperature evolution of the
expanding universe in relative units looks like the mass
evolution of a cold universe with a constant tempera-
ture of cosmic background radiation [38, 54, 55].
There is one more argument in favor of relative
units, namely, the large deficit of luminous masses,
M/ML ≥ 10 (where ML is the mass of luminous matter),
in all superclusters with a mass of M ≥ 1015M
and a
size of R  5 Mpc [56]. In this case, the Newtonian
velocity is less than the cosmic one by an order of mag-
nitude. In terms of relative units, this deficit can be
caused [56] by galaxy deceleration in the course of the
cosmological evolution of galaxy masses (135).
The evolution of the universe as a mechanism of
cosmic-object deceleration in a central gravitational
field appears to be proposed by Einstein and Strauss in
1945 [57]. This idea was then developed in terms of
conformal variables and coordinates in [56], where the
cosmological evolution was shown to cause an energy
decrease and the formation of galaxy and its clusters
because of the capture of cosmic objects by central
gravitational field. In fact, the definition of total particle
2 universe we consider the universe evolution model
p αm
energy, E(η) = --------------------- – -------- (where p is the momen-
2m 0 a ( η ) r kinetic energy of a slightly self-interacting scalar field
tum and α = GM
m = rgm/2 is the Newton constant),
PHYSICS OF PARTICLES AND NUCLEI
121
follows from the definition of the conformal one-parti-
cle energy with the running mass m(η) = a(η)m0 and
from Newtonian potential. Therefore, energy is not
conserved in contrast to the energy of a particle with
constant mass in Newtonian mechanics. The scale fac-
tor a(η) increases as the energy varies from positive to
negative values. The energy goes to zero at a time η = ηI ,
which is the time of galaxy capture described in [56].
The galaxy capture could be a crucial argument to obvi-
ate one of fundamental troubles of the modern theory of
galaxy formation: how does a system of unbound parti-
cles with positive energy turn into a system of bound
particles with negative energy?
(135)
the range of applicability of the Newtonian mechanics
when describing galaxies. The range is determined by
the critical radius Rcr
1020 cm (M/M
)1/3, at which the
Newtonian orbital velocity in absolute units coincides
with the Hubble velocity of galactic dilation relative to
the center of the supercluster containing the galaxy.
According to [56], the Newtonian dependence of the
orbital velocity on the distance to the supercluster center,
(136) v Newtonian orbital ( R circle ) =
rg
---------------,
2R circle
for circular trajectories with Ṙ circle = Ṙ˙circle = 0 (Rcircle =
a(t)r, dt = a(η)dη) should be substituted by the depen-
dence
rg 2
v Cosmic orbital ( R circle ) = --------------- + γ ( R circle H ) , (137)
2R circle
provided that cosmic evolution is taken into account.
Here, γ = 2 – ⎛ --- Ωmatter + 3Ω Λ ⎞ . In the cosmology
3
⎝2 ⎠
model with cosmological quantities identified with
conformal ones, we have
γ ( Ω rigid = 1 ) = 2.
In this case, the deficit of visible matter decreases in
contrast to the Standard Cosmology, in which the quan-
tity γ is negative,
γ ( Ω Λ = 0.7, Ω matter = 0.3 )
–1/2,
and much more dark matter is required. Moreover, as is
seen from Eq. (137), in the conventional model with
absolute units of measurement the orbital velocity
could be imaginary.
4.10. Universe Creation
Thus, to define the initial data of the creation of the
(with relative units of measurement), in which the
can serve as a primordial vacuum dark matter in an
Vol. 37 No. 1 2006
122 ZAKHAROV et al.

NZ (x)
0.06
0.04
0.02
4
0 3
5
10 2 x
τ 15
1
20
25 0

Fig. 3. Longitudinal NZ(x) components of the boson distribution function versus the dimensionless time τ = 2ηHI and dimensionless
momentum x = q/MI under the initial condition MI = HI (γv = 1) (according to [54, 55, 58]).

empty universe.7 This yields the evolution law in the Vacuum postulate (85) can exclude negative ener-
form of the equation of rigid state (136). gies but it can’t forbid the Hubble parameter to be neg-
ative.
The equation of state (136) is the most singular one. In the model with the event energy taking the form
Therefore, it is natural to assume that the universe is E(ϕ) = Q/ϕ, the universe distribution function NU given
governed by this state at instant of creation. There were by Eq. (86) can be found since Eqs. (88) can be solved
no particles, time, or temperature before the instant of exactly. In this case,
the creation of the universe. We remind that tempera-
ture of a system is a characteristic of collisions and 1 2 2 1 ϕ
scattering of its particles. If particles are absent, the N U = ------------------ sin Q – --- ln ----- ≠ 0. (139)
4Q – 1
2 4 ϕI
temperature is also absent.
We will describe the quantum universe creation in In what follows, we will mainly consider the case of
GR in the low-energy approximation considered in expanding universe.
Section 4.8, when the action is divided into the parts
describing universes by Eq. (47) (or (132)) and parti- 4.11. Matter Creation from the Vacuum
cles by Eq. (133).
As was shown in [54, 55, 58] within the framework
In this approximation, the second quantization with of the conformal cosmological model considered
the postulate of the vacuum as a lowest-energy state is above, there is an intense resonance vector–boson pro-
a general method of describing the cosmological cre- duction from the vacuum in the early universe when its
ation of both universe (47) and particles (133) [42]. –1 2 –1
horizon H I = a I H 0 coincides with the Compton
In the case of the equation of rigid state (136), the –1 –1 –1
energy takes the form E(ϕ) = Q/ϕ, where the parameter wavelength M I = a I M W of W- and Z-bosons. This
2
Q coincides with the integral of motion Q = 2V 0 ϕ I H I = effect is due to these bosons being the only Standard
Model particles having the zero–mass production sin-
2
2V 0 ϕ 0 H 0 , which is referred to as a scale charge. gularity. The initial data for the boson pair production,
2
If the scale charge of the vacuum is equal to zero, the a I = [H0 /MW]2/3 = 10–29, follow from the uncertainty
two universes with opposite scale charges ±|Q| are cre- relation ∆E∆η = 1 for the boson pair production with
ated simultaneously. In terms of the Standard Cosmol- the energy 2MI during the universe lifetime (2HI )–1.
ogy, they correspond to expanding (+) and contracting (–) 29
universes: This corresponds to the redshift zI
10 . According
to calculations, the boson density in units of the initial
2 2 Q Hubble parameter reaches equilibrium very rapidly, rel-
ϕ ±Q = ϕ I ± --------- η. (138) ativistic bosons make a dominant contribution (see Fig. 3)
2V 0
[54, 55], and the boson temperature is equal to
7 Inpresent-day literature, the vacuum dark matter is referred to as 1/3 1/3
2 2
a quintessence [38]. T boson ∼ ( M I H I ) = ( MW 0 H0) ∼ 3K, (140)

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

 TETRAD FORMALISM AND REFERENCE FRAMES 123

which is an integral of motion of the supernova evolu- functions defined by the gauge group of the Hamilto-
tion [52]. The cosmic background radiation tempera- nian approach and by the finiteness condition for
ture succeeds to the temperature of decaying bosons. energy density.
The boson lifetime [55] 0
(ii) The zero mode of the Gauss condition T k = 0 is
2 2/3
τ W = 2H I η W
⎛ -----⎞
16, identified with the cosmological scale factor. This
⎝ α g⎠ makes it possible to join the two theories: the nonivari-
(141)
ant theory of the island universe [20] (where the cosmo-
2/3
τ Z ∼ 2 τ W ∼ 25 logical scale factor is identically equal to unity, and the
symmetry with respect to the reparametrization of the
determines the present-day baryon density coordinate time is violated by global synchronization)
2 and Hamiltonian cosmology [41] (where the cosmolog-
Ω b ∼ α g = α QED / sin θ W ∼ 0.03. (142) ical scale factor is considered as a dynamical variable).
The ratio of the created matter density ρv (ηI ) to the (iii) The identification of the cosmological scale fac-
primordial scalar–field (quintessence) density ρcr(ηI ) = tor and its canonical momentum with the invariant evo-
2 2 lution parameter and the universe energy, respectively,
H I ϕ I is extremely small: together with the postulate of the quantum theoretical
2 2 vacuum as the lowest-energy state, allow us to define an
ρv ( η I ) M I MW – 34
---------------- - ∼ 10 .
∼ ------2- = -------- (143) absolute origin of the geometric interval s. The vacuum
ρ cr ( η I ) ϕ ϕ0
2
postulate eliminates the cosmological singularity of the
I
universe if its reparametrization-invariant energy of
Transverse bosons produce a baryon asymmetry of events is positive [28].
the universe because they cause the polarization of the
(iv) The problem of universe homogeneity is
Dirac left-fermion vacuum according to the Standard
resolved by defining the cosmological equations as
Model selection rules [59]: the difference of baryon and
exact GR equations averaging over spatial volume in a
lepton numbers is conserved, but their sum is not. The
specific reference frame [54, 55].
superweak interaction [60], observed experimentally
and responsible for the CP-violation with the constant (v) The reparametrization-invariant evolution
XCP ~ 10–9, makes the baryon asymmetry of the uni- parameter ϕ given in a specific reference frame and the
verse freeze up at the density absolute origin ζ = 0 of the geometric interval deter-
mine the following initial data for the GR field vari-
–9 – 34
ρ b ( η = η L )
10 × 10 ρ cr ( η = η L ). (144) ables F(ϕ, x): FI = F(ϕI , x), where ϕI = ϕ(ζ) ζ = 0 .
The consequent evolution in the cold universe (vi) The theoretical description in terms of the Stan-
repeats the well-known hot-universe scenario [53] dard Model [54, 55], when applied to the creation of
because such evolution is determined by the confor- matter in the universe, its energy balance, and a series
mally invariant mass-temperature ratio m/T. The baryon of cosmological observations and astrophysical data,
asymmetry increases with mass, while the primordial testifies to the existence of an initial value of the scale
quintessence density varies inversely with the mass factor ϕI, which corresponds to the redshift z2 ~ 1029.
squared. Because of this, the present-day baryon den- However, the metric formalism does not define iner-
sity, estimated from Eqs. (141), (143), and (144), tial reference frames interrelated by Lorentz transfor-
ϕ 3
– 43 43 η 3/2
– 43
mations. Such transformations can be defined only in
Ω b ( η 0 ) = -----0- × 10 ∼ 10 ------I × 10 the tetrad formalism of GR.
ϕL ηL
(145)
α QED
∼ ---------------
2
- ∼ 0.03 5. REFERENCE FRAMES
sin θ W IN TETRAD FORMALISM
is in close agreement with experimental data. 5.1. Action, Linear Forms, and Symmetry
Thus, there is a reparametrization–invariant descrip- Tetrad formalism is required in GR to separate gen-
tion of the Dirac Hamiltonian formulation of GR and eral covariant transformations from transformations of
cosmology within the framework of metric formalism. reference frames.
We consider Hilbert–Einstein action (92) in the
4.12. Results space defined by the interval
2 αβ µ ν
The reparametrization-invariant description of ref- ds = η ω ( α ) ω ( β ) ≡ g µν d x d x , (146)
erence frames in GR yields the following results.
ab
(i) A zero–mode solution of the GR variational η = diag ( 1 – 1 – 1 – 1 ), (147)
0
equation T k (ϕ 0 F) = 0 is proved to exist in the class of Here,

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

124 ZAKHAROV et al.

µ 6
ω ( α ) = e ( α )µ d x (148) N = Ndψ . (155)
are the Cartan linear forms [20, 21], and the coefficients The symmetric parameterization w(a) = e ( a )i d x of
i
e(α)µ of expansion over differentials of a coordinate
space are referred to as tetrads. These forms allow us to Cartan forms can be used as a nonlinear realization of
include fermions and other fields f of the Standard equi-affine symmetry [61].
Model, The Dirac parameterization defines a set of hyper-
surfaces x0 = const with a unit vector να = (1/N, –N k/N)
S [ ϕ 0 F = e, f ] = S GR [ ϕ 0 e ] + S SM [ ϕ 0 F ] (149) perpendicular to the hypersurfaces. The second form
(extrinsic curvature) is defined by the equation
and to separate gauge transformations from transforma-
tions of reference frames [61]. 1 l l
π ( a )i = ------ [ ( ∂ 0 – N ∂ l )e ( a )i – e ( a )l ∂ i N ]
The Cartan forms ω(α) are invariant with respect to Nd (156)
the general coordinate transformations 2
= ψ [ v ( a )i + 2e ( a )i v ψ ],
µ µ µ 0 1 2 3
x x̃ = x̃ ( x , x , x , x ), (150)
where the velocities
which are treated as gauges. The forms transform as a v ( a )i
vector representation of Lorentz group treated as a ref-
erence frame transformation group. An example of 1 l 1 l l (157)
such transformation is = ------ ( ∂ 0 – N ∂ l )e ( a )i + --- e ( a )i ∂ l N – e ( a )l ∂ i N
Nd 3
⎧ ω(0) – V ω(1) and
⎪ ω(0) = ----------------------------
⎪
2
1–V 1 l 1 l
⎪ v ψ = ------ ( ∂ 0 – N ∂ l ) ln ψ – --- ∂ l N (158)
⎪ ω(1) – V ω(0) Nd 6
⎨ ω(1) = ---------------------------- (151)
indicate that this hypersurface is embedded into a four-
⎪
2
1–V
⎪ dimensional spacetime.
⎪ ω(2) = ω(2) After performing transformation (154), one can
⎪ write out the intrinsic three-dimensional scalar curva-
⎩ ω(3) = ω(3) .
ture (3)R(e) in the form
It is known that transformations of reference frames (3) 1 (3) 8
do not change the number of variables, while gauge R ( e ) = ------4 R ( e ) + ------5 ∆ψ, (159)
transformations result in the constraints decreasing this ψ ψ
number [23]. where (3)R(e) is the curvature expressed in terms of the
i j
tetrads e(a)i: ∆ψ = ∂ i(e ( a ) e ( a ) ∂ j ψ) .
5.2. Reference Frame
The choice of a Lorentz reference frame in GR 5.3. Invariant Evolution Parameter
implies the fixation of Lorentz indices (α) in Cartan
forms (148) and the subdivision of their components Thus, GR and cosmology provide a series of argu-
into timelike ω(0) and spacelike ω(α) components. ments in favor of the identification of the GR invariant
evolution parameter with the cosmological scale factor.
The Hamiltonian dynamics is formulated in a spe- This factor can be introduced into theory (149) by the
cific Lorentz reference frame in terms of the Cartan conformal transformation of all the fields with the con-
forms [22–24]: n (n)
0
formal weight (n), F(n) = a F , including the Cartan
ω ( 0 ) = Nd x , (152) forms
2
ω ( α ) = aω ( α ) , ψ = aψ .
i i 0 2
ω ( a ) = e ( a )i ( d x + N d x ). (153) (160)
Here, the triads e(a)i form the spatial metric In general, such a transformation is used as a defini-
tion of the cosmological perturbation theory [50, 51]. In
ij j i
γ ij = e ( a )i e ( a ) j , γ = e(a) e(a) . the gauge-invariant Hamiltonian approach to GR con-
sidered in the preceding section, the scale factor should
According to Dirac [23], the spatial metric determi- be treated as a dynamical variable used as an invariant
nant can be extracted by taking out the factor ψ2 from evolution parameter. To define a complete set of Hamil-
the triads tonian canonical momenta the number of variables of
2 the theory should be restored by eliminating the zero
e ( a )i = ψ e ( a )i , det e = 1, (154) Fourier harmonic from the spatial metric determinant

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

 TETRAD FORMALISM AND REFERENCE FRAMES 125

with the help of a constraint condition. To find this con- Here, the first term is due to the kinetic part K[ ϕ e ];
dition, we perform transformation (160) of action (149) the second term is derived from the quasisurface part
and obtain the equation
S[ ϕ e ], which is not a total derivative because of time-
2
0 ( ∂0 ϕ ) dependence of the scale factor; and the third term is the
S [ ϕ 0 F ] = S [ ϕ F ] – V 0 d x ---------------
N0
-. ∫ (161) quadratic action for the scale factor.
The scale factor is a dynamical variable, therefore its
∫
3
Here, V0 = d x is the volume of the Dirac coordinate canonical momentum can be obtained by variation of
Lagrangian (170) with respect to the time derivative ∂0ϕ:
space. The averaged lapse function N0(x0),
∂L SD
P ϕ ≡ ----------------
-
∂ ( ∂0 ϕ )
∫ d xN
0 –1 –1 3 –1 0 i –1
N0( x ) = V0 d (x , x ) ≡ 〈 Nd 〉 (162) (171)
∫
3
V0 = – d x [ 4ϕv ψ + 2v ϕ ] ≡ – [ 4ϕV ψ + 2V ϕ ].
determines the geometric time ζ,
In this case, the average of the canonical momentum
dζ = N 0 ( x )d x ,
0 0
(163) of the spatial determinant takes the form
∂  SD
and S[ ϕ F ] is the sum of the GR action
∫
3
P ψ ≡ – d x -------------------------
-
∂ ( ∂ 0 log ψ ) (172)
S GR [ ϕ e ]
∫
3
(164) = – d x p ψ = – 2ϕ [ 4ϕV ψ + 2V ϕ ],
∫
0 3
= dx d x ( K [ ϕ e ] – P [ ϕ e ] – S [ ϕ e ] )
∫ ∫
3 3
and SM action (149), with all masses, including the where Vϕ = d xv ϕ , Vψ = d xv ψ , and ϕ' = dϕ/dζ. It
Planck mass, scaled by the factor ϕ = ϕ0a. Action (164) is easy to prove that the velocities Vϕ and Vψ can not be
is the sum of the kinetic K, potential P, and quasisur- expressed in terms of canonical momenta Pϕ and Pψ
face S terms: because, Eqs. (171) and (172) form a degenerate sys-
2 tem having no solutions.
2 2 v ( ab )
K [ ϕ e ] = N d ϕ – 4v ψ + ----------
- , (165) This means that action (170) is singular, because of
6 the double counting of the spatial determinant.
To avoid double counting, the field variable log ψ in
2 12
N d ϕ ψ (3)
P [ ϕ e ] = --------------------
- R ( e ), (166) Eq. (169) should be defined in the class of functions
6
under strong constraint conditions
2 l
S [ ϕ e ] = 2ϕ [ ∂ 0 v ψ – ∂ l ( N v ψ ) ]
(167)
∫ d x log ψ ≡ 0, ∫ d xv
3 3
ϕ
2
2 j 6 ψ ≡ 0. (173)
– -----∂ j [ ψ ∂ ( ψ N d ) ].
3 V0 V0

Here, we used definitions (157), (158), v(ab) = These conditions mean that the scale factor and its
1 i i velocities must be orthogonal to the logarithm of spatial
--- (e ( a )i v ( b ) + e ( b )i v ( a ) ) , and determinant log ψ .
2
Under these conditions, action (161) takes the form
–1 l 1 l
v ψ = N d ( ∂ 0 – N ∂ l ) log ψ – --- ∂ l N , (168)
6 ∂0 ϕ
0⎧
∫ dx ⎨⎩ V ϕ ( x )∂ ⎛⎝ --------
-⎞
0
S [ ϕ0 F ] =
which is the velocity of the logarithm of spatial deter- 0
N ⎠ 0
0
minant (174)
2
log ψ = log a ( x ) + log ψ .
0 2 ⎫
∫
--- + d x ( K [ ϕ e ] – P [ ϕ e ] +  SM ( ϕ F ) ) ⎬,
(169) 3

After removing scale factor (160), the part of ⎭

action (163) that contains the spatial determinant takes
the form where SM is the SM Lagrangian density. The quanti-
ties N0, K[ ϕ e ], and P[ ϕ e ] are defined by Eqs. (121),
∫
3 2 2 2
L SD = – d xN d [ 4ϕ ( v ψ ) + 4ϕv ϕ v ψ + ( v ϕ ) ] (165), and (166), respectively.
(170)
This separation allows us to formulate the Hamilto-
∫ d x  SD .
3
= nian approach to the theory under consideration.

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

126 ZAKHAROV et al.

5.4. Hamiltonian and Constraints 1 i i

p ( ab ) = --- [ p ( a ) e ( b )i + p ( b ) e ( a )i ]. (182)
The equations derived from action (164) have no 2
uniquely defined solutions for metric components, Variation of action (177) with respect to the
which depend on initial data and gauge. For gauge Lagrange multipliers N i, C0, and C(a) yields the first-
ambiguity to be excluded the coordinates should be class constraints
fixed by imposing gauge conditions allowed by the
equations of motion. 0
T kt = 0 (183)
Following Dirac [23], we fix the physical coordi-
nates by imposing the transversality condition and the second-class constraints
k
i
∂i e(a) = 0 (175) p ψ = 0, ∂ k e ( a ) = 0, (184)
as the minimal-surface and transversality conditions,
and the minimal-surface condition [23] respectively.
v ψ = 0, (176)
5.5. Equations of Exact Theory
where v ψ is given by Eq. (168). The time lapse function is determined by the equa-
All the equations considered above can be derived tion
by variation of the Hamiltonian action
δS [ ϕ 0 F ]
2 N d ----------------------- = 0, (185)
Pϕ δN d
∫ dx ∫
0
S = – P ϕ ∂ 0 ϕ + N 0 ---------
4V 0 which serves as an energy constraint and takes the form
(177) 2
⎛ ⎞ ϕ'
------ =  t .
∫ ∑ P F ∂ 0 F +  – N d  t ( ϕ F )⎟ , (186)
3
+ d x⎜ 
⎝ F
⎠
Here,  = N d /N 0 is the reparametrization-invariant
i
where PF is the set of the field momenta pψ, p ( b ) , and part of function of the running time (121) satisfying
pf , defined as the condition 〈–1〉 = 1, and the quantity t is given
by Eq. (180).
∂K [ ϕ e ] 2
p ψ = ----------------------- = – 8ϕ v ψ , The averaging of Eq. (185) over the volume V0
∂ ( ∂ 0 ln ψ ) yields the Friedmann-type equation of the exact theory,
(178)
i ∂K [ ϕ e ] δS [ ϕ 0 F ]
p(a) = ----------------------. 2
N d ----------------------- = 0 ⇒ ϕ' = ρ t ,
∂ ( ∂ 0 e ( a )i ) (187)
δN d
The quantity where
 = N
i 0 k
Ti + C0 pψ + C(a) ∂k e(a) (179) ρ t ≡ 〈  t〉 (188)
is the sum of constraints (175) and (176), where Nd, N i, is the total generator of evolution relative to time ζ
C0, and C(a) are the Lagrange multipliers. In Eq. (177), (163) for all the dynamical variables, with the excep-
tion of the scale factor. Thus, in the Hamiltonian
6p
( ab ) ( ab ) p 16 approach with the extracted scale factor, energy con-
 t ( ϕ F ) = ------------------------
2
- – -----2- p ψ straint equation (185) is subdivided into local (189) and
ϕ
2
ϕ global (187) parts.
(180)
2
ϕ ψ (3)
7
Substituting (187) into (185), we arrive at the equa-
+ ------------ [ R ( e )ψ + 8∆ψ ] + ψ T 0 ( SM ) ,
12 0
6 tion
〈  〉
0 1
T kt = – p ψ ∂ k ψ + --- ∂ k ( p ψ ψ ) ----------------t - =  t , (189)
6 
(181)
i which unambiguously determines the local part of the
p(b) ∂k ei(b)
0
+ + T k ( SM ) , lapse function
0 0
where T 0 ( SM ) and T k ( SM ) are the energy–momentum 〈 〉
 = ---------------t - . (190)
tensor components of the matter fields F and t

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

 TETRAD FORMALISM AND REFERENCE FRAMES 127

Substituting this solution into global constraint Pϕ

2
2
equation (187), we reduce it to the form -2 = 〈  t〉
-------- (197)
2 4V 0
ϕ' = 〈  t〉 ≡ ρ t .
2
(191)
has the two solutions
Thus, after isolating the evolution parameter as a
dynamical variable, the energy constraint equation is P ϕ ( ± ) = ± 2V 0 〈  t〉 = 2V 0 ϕ' (198)
concerned with only the zero Fourier harmonic. This
completely corresponds to the diffeomorphism group corresponding to positive and negative energies of
of the Hamiltonian approach [22–24]. For this reason, events.
the quantum gravity, if defined as quantization of a glo- Substitution of solutions (198) into action (177)
bal constraint, is significantly simplified and, in yields reduced action (127), considered above in the
essence, reduced to the quantum cosmology considered metric approach,
in Section 3. On the other hand, there is only the local S(±) [ ϕ I ϕ0 ]
energy constraint t = 0 in the conventional Hamilto- energy constraint
nian approach, in which the scale factor is not isolated ϕ0
⎧ 3 ⎫ (199)
[20]. This contradicts the diffeomorphism group of the
Hamiltonian approach and complicates considerable
= ∫ ∫
dϕ ⎨ d x
⎩
∑ PF ∂ϕ F +  −
+ 2  t ( ϕ ) ⎬,
⎭
the quantization problem considered as quantization of ϕI F

a constraint. which is similar to reduced action (33) in GR. Here,

The second equation of the Friedmann cosmology is
derived by variation of action (174) with respect to the  = /∂0ϕ and the scale factor ϕ serves as a dynamical
scale factor ϕ: evolution parameter in the space of events [ ϕ F ]. It is
easy to verify that the variation of this action with
δS [ ϕ 0 F ]
ϕ ----------------------- = 0 ⇒ 2ϕϕ'' = ρ t – 3 p t , (192) respect to log ψ yields Eq. (196), with  determined
δϕ by Eq. (190). Action (199) describes the field evolution
where in terms of the red-shift parameter related to the scale
factor ϕ by the equation ϕ = ϕ0 /(1 + z).
3 pt
(193)
≡ 〈 3K [ ϕ e ] – P [ ϕ e ] + 2S [ ϕ e ] +  ψ T k ( SM )〉
12 k 5.7. Geometric Sector
Solutions of the equations derived from (177) in
is the precise pressure of all the fields including the SM terms of geometric time (122) define geometric interval
fields. The relationship 2 2
given by Eqs. (9), (152), and (153), ds2 = ω ( 0 ) – ω ( a ) ,
2ϕ' + 2ϕϕ' ≡ ( ϕ )'' = 3 ( ρ t – p t ) ≡ 〈 Â t 〉
2 2
(194) with the Cartan forms
follows from Eqs. (187) and (192), where the expres- ω ( 0 ) ( ζ, x ) = a ( ζ )ψ  dζ,
6
(200)
sion
ω ( a ) ( ζ, x ) = a ( ζ )ψ ( e ( a )i d x +  ( a ) dζ ).
2 i
 t  ≡ 4P [ ϕ e ] – 2S [ ϕ e ] (201)
(195) In this case, the Dirac minimal-surface and transver-
+ ψ ( 3T 0 ( SM ) – T k ( SM ) ) 
12 0 k
sality conditions given by Eq. (184) are to be imposed
on the metric components [23]:
determines the equation for log ψ :
∂ ζ [ ψ ] = ∂ ( b ) [ ψ  ( b ) ],
i
∂i e(a)
0
6 6
1 δS [ ϕ 0 F ] (202)
--- ----------------------- = 0 ⇒ Â t  = 〈 Â t 〉 . (196) ( det e = 1 ).
2 δ log ψ
Equations (189) and (196) in the infinite-volume We remind that invariant geometric time ζ (122) is
limit V0 ∞ reduce to the zero–density and zero– determined by the Friedmann-type equations of exact
theory (187):
pressure equations, t = 0 and  t  = 0, because the
dϕ
averages 〈〉 and 〈 Â t 〉 tend to zero in this limit. ------ ≡ ϕ' = ± 〈  t〉 . (203)
dζ
The solution
5.6. Resolution of Constraint Equation ϕ0
and Reduction of Action
dϕ
The Hamiltonian form of global constraint equa- ζ ( ϕ0 ϕ I ) = ∫ ---------------
〈 〉
-
t
(204)
tion (191), ϕI

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

128 ZAKHAROV et al.

of Eq. (203) can be treated as a relativistic relation 〈 〉

between the evolution parameter ϕ and geometric δS [ ϕ ]
N d --------------- = 0 ⇒  = ---------------t - , (214)
time (122). Equations (192) and (203) describe the δN d t
Friedmann-like universe evolution (under no assump-
tion of homogeneity) as a pure relativistic effect in the δS [ ϕ ]
ψ --------------- = 0 ⇒ Â t  = 〈 Â t 〉 , (215)
GR field space of events. This effect is similar to the 2δψ
Lorentz time dilation in SR. In this case, all fields
(n) n (n) where
F = a(ζ) F are to be considered in the space
 t  ≡ 4P [ ϕ e ] – 2S [ ϕ e ]
ω ( 0 ) ( ζ, x ) = ψ  dζ,
6
(205)
(216)
4y⎛
+ψ T k ( SM )⎞ .
0 k
ω ( a ) ( ζ, x ) = ψ ( e ( a )i d x +  ( a ) dζ )
2 i 3T –
(206) ⎝ 0 ( SM ) ⎠
with mass multiplied by a scale factor a(ζ): m = a(ζ)m.
(n) 2 n (n) 5.8. Hamiltonian Cosmological Perturbation Theory
The variables F = (ψ ) F defined in the space
with a basis We now consider the expansion of Cartan forms (200)
and (201),
ω ( 0 ) ( ζ, x ) = ψ  dζ,
4
(207) ω ( 0 ) = a ( η ) ( 1 + Φ )dη, (217)
( TT )
ω ( a ) ( ζ, x ) = ( e ( a )i d x +  ( a ) dζ ),
i i i
(208) ω ( a ) = a ( η ) ( 1 – Ψ ) ( d x ( a ) + h ( a )i d x
(218)
(T )
with mass multiplied by the scale a(ζ) ψ , i.e., m =
2
+ ∂ ( a ) νdη + N ( a ) dη )
a(ζ) ψ m , have been also considered [63–65]. In what
2
( TT )
by the perturbation theory, where h ( a )a = 0. The trans-
follows, we will refer to the fields F as Lichnerowicz ( TT )
versality condition ∂ i h ( a )i = 0 and minimal-surface
variables. In general, it is convenient to introduce
energy density (180), condition pψ = 0 (202) are imposed. The equation
6p p 16 3
( ab ) ( ab )
 t = ------------------------
- – -----2- p ψ
2
∆ν = --- Φ h' (219)
2
ϕ ϕ 4
(209) gives the longitudinal component ∂(a)ν of origin shift
2
ϕ ψ (3)
7
+ ------------ [ R ( e )ψ + 8∆ψ ] + ψ T 0 ( SM )
4y 0
vector (222).
6 To avoid double counting in the cosmological per-
and pressure density (193), turbation theory, we define cosmological perturbations
of the metric components,
3 pt
(210)  = ( 1 + Φ + 3Ψ ), ψ 2 = ( 1 – Ψ ), (220)
≡ 〈 3K [ ϕ e ] – P [ ϕ e ] + 2S [ ϕ e ] +  ψ T k ( SM )〉 ,
4y k
( TT )
e ( a )i = δ ( a )i + h ( a )i , (221)
with an arbitrary weight y, referred to as a Lichner-
owicz index. Here, we use the Hamiltonian form of N (a) = ∂(a) ν + N (a) ,
T T
∂ ( a ) N ( a ) = 0, (222)
notations (165)–(167):
6 p ( ab ) p ( ab ) 16 2 in the class of functions with the nonzero Fourier har-
K [ ϕ e ] =  ------------------------
- – -----2- p ψ , (211) monics
ϕ
2
ϕ
∫ d xΦ ( x )e
3 ikx
Φ̃ ( k ) = , (223)
2
ϕ ψ (3)
7
P [ ϕ e ] =  ------------ [ R ( e )ψ + 8∆ψ ], (212) satisfying the strong conditions
6

∫ d xΨ ( η, x ) ≡ 0,
3
S [ ϕ e ] = –4 [ ∂ζ pψ – ∂l ( N pψ ) ]
l 〈 Ψ〉 = i
(224)
(213)
∫ d x∂  /∂Ψ̇ ( η, x i ) ≡ 0.
3
ϕ
2
〈 P Ψ〉 =
– -----∂ j [ ψ ∂ ( ψ  ) ].
2 j 6
3
The energy–momentum tensor components can be
In this case, the scalar sector , ψ , and ∂ζ[ ψ ] =
6 also expanded:
0 k
∂ ( b ) [ ψ  ( b ) ] is defined by the equations
0 k
T 0sm = ρ s + δT 0 , T ksm = 3 p s + δT k ,
6
(225)

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

 TETRAD FORMALISM AND REFERENCE FRAMES 129

where ρs and ps are the density and pressure, respec- 2

µ 2 + ( α ψ – 2 )h
2 2
µ 2 + ( α φ – β )h
2
tively. Under the minimal-surface condition pψ = 0, the γΦ = -------------------------------------
2 2
-, γΨ = ------------------------------------
2 2
-, (235)
first-order terms of expressions (211)–(213) are µ1 + µ2 µ1 + µ2
2 2
(1) (1) 2ϕ (1) ϕ
K = 0, P = – --------- ∆Φ, S = -----∆Ψ. (226) µ2 + αψ h
2 2
βh
2
3 3 ζ Φ = -----------------------
2 2
-, ζ Ψ = -----------------
2
-2 . (236)
( TT ) ( TT ) µ1 + µ2 µ1 + µ2
Under the conditions ∂ i T ik = 0, T ii = 0, and
0(T )
∂k T k = 0, the equations for the vector and tensor The Hamiltonian version of the cosmological per-
components take the form turbation theory, with the cosmological factor isolated
from action, does not require its convergence to be
2
0(T ) ϕ T proved because the perturbations are given in a differ-
Tk = – ------ ∆N k , ent class of functions (with nonzero Fourier harmonics)
12
than the cosmological dynamics described by exact
2 TT (227) equations (188) and (192).
TT ϕ
2
( TT ) ( ϕ h ik ' )'
T ik = ------ – ∆h ik + ---------------------
- .
12 ϕ
2 Let us compare the Hamiltonian cosmological per-
turbation theory described by Eqs. (228) and (229) with
They coincide with the equations of the cosmologi- the standard one [49, 50], in which the zero Fourier har-
cal perturbation theory [49, 50]. monic of the logarithm of the spatial metric determi-
In this case, the scalar sector,  = 1 + Φ and ψ = 1 – nant is taken into account twice: as a cosmological
Ψ/2, defined by Eqs. (214), (215) in view of Eqs. (226) scale factor and as a spatial potential of the metric.
takes the form Equations (228) and (229) basically differ from the
0
t0
2
= [ ∆ – α ψ h ]Ψ – 2h Φ,
2
(228) equations

– 3  (  Φ + Ψ' ) + ∆Ψ = 4πGδT 0 ,
0 k 2 2 0
[ t 0 + t k ] = – βh Ψ + ( ∆ – α φ h )Φ, (229)
2
0 0
where t 0 = δT̃ 0 3/ ( 2ϕ ) and
2 3 [ ( 2  ' +  )Φ +  Φ' + Ψ'' + 2  Ψ' ] (237)
i
2 + ∆ ( Φ – Ψ ) = 4πGδT i
αφ = 5 + γ , α ψ = 6 – 2y,
2 p 2 3ρ (230) of the standard cosmological perturbation theory
β = 15 – 2y, γ = -----s , h = ---------2s .
ρs 2ϕ (where 4πG = 3/(2ϕ2),  = a'/a, and ∆ = ∂ i , see
2

The solutions of Eqs. (228) and (229) in terms of Eq. (4.15) in [51]) in the absence of time derivatives of
Fourier components can be written out in the form the metric components. These derivatives are responsi-
ble for the primordial temperature perturbation spec-
1 – γΨ γΨ trum of cosmic microwave background (CMB) in the
Ψ̃ = – ----------------
2
-2 + ---------------
2
-2 t 00 inflation model [51, 62]. The time derivatives of Φ and
k + µ1 k – µ2 Ψ are eliminated by imposing strong conditions (224)
(231)
ζΨ ζΨ to avoid double counting. In this case, scalar metric
+ ----------------
2
-2 – ---------------
2
-2 t kk , components serving as dynamical variables turn into
k + µ1 k – µ2 conventional potentials similar to the Coulomb electro-
dynamic potential. Thus, the explanation of the primor-
1 – γΦ γΦ
Φ̃ = – ----------------
2
-2 + ---------------
2
-2 t 00 dial temperature perturbation spectrum of CMB in the
k + µ1 k – µ2 inflation model [51, 62] is based on double counting,
(232) which makes it impossible to construct the Hamiltonian
1 – ζΦ ζΦ formalism and quantum theory.
+ ----------------
2
-2 + ---------------
2
-2 t kk ,
k + µ1 k – µ2 On the other hand, it is the Hamiltonian quantum
where theory that allows us, as is shown in Section 4, to treat
CMB and its primordial temperature perturbation spec-
2 2
µ 1 = h [ ( 2β – α ψ α φ ) + ( α ψ + α φ ) /4
2 trum as a result of decays of primordial W- and Z-bosons
(233) created from the Bogoliubov stable vacuum at the time
+ ( α ψ + α φ )/2 ], when the boson mass was equal to the Hubble constant.
In this case, the equations for longitudinal components
2 2 2
µ 2 = h [ ( 2β – α ψ α φ ) + ( α ψ + α φ ) /4 of W- and Z-bosons [55] are close to Eqs. (237) of the
(234) inflation model, which describe the primordial temper-
– ( α ψ + α φ )/2 ], ature perturbation spectrum.

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

130 ZAKHAROV et al.

5.9. Central Gravitational Fields In the case of (239), we obtain the solutions that
In the Newtonian case, when ps, ρs  ϕ2k2, solutions reduce in the coordinate representation to the Yukawa
of Eqs. (228) and (229) coincide with those of the con- potentials with Jeans-like oscillations:
ventional perturbation theory with time-depending r – µ 1 ( z )r
masses: Ψ = – ----g [ ( 1 – γ Ψ )e – γ Ψ cos µ 2 ( z )r ], (246)
r
4πG 0 4πG 0 k
Ψ̃ = – ----------
2
- T̃ 0 , Φ̃ = – ----------2
- [ T̃ 0 + T̃ k ], r – µ 1 ( z )r
k k Φ = – ----g [ ( 1 – γ Φ )e – γ Φ cos µ 2 ( z )r ]. (247)
(238) r
3
G = ------------.
2 In the case of a point mass
8πϕ

∑M
Let us consider the central gravitational field of a 0 3 1
T0 = J δ ( x – y J ) – ------ , (248)
unit mass with V0
J
0 3 –1
T 0 = M [ δ ( x ) – V 0 ], T kk = 0. (239) solutions (231) and (232) take the form
Here, by definition, ϕ = ϕ0a, M = M0a, and ϕ0 =
Ψ( x )
3/8πG . Solutions (238) can be written in the conven-
r gJ
∑ -----
– µ 1 ( z )r J
tional form -[(1 – γ
= – Ψ )e – γ Ψ cos µ 2 ( z )r J ],
r J
3 T 00 r
∫
3 J
Ψ( x) = Φ( x) = ------------2 d x -------------
- = ----g , (240) (249)
y –x r Φ( x )
3
4πϕ T 00 = Mδ ( x )

r gJ
∑ -----
2 2 2 – µ 1 ( z )r J
where r = x 1 + x 2 + x 3 and = – -[(1 – γ Φ )e – γ Φ cos µ 2 ( z )r J ],
r J
J
3M
r g = ------------2 = 2GM. (241) ( y = 1, 3 )
4πϕ where rgJ = 2GMJ, rJ = x – y J , and µ 1, 2 (z) is given
As follows from Eq. (219), the shift vector N i is by Eq. (245). The minimal-surface condition deter-
mines the shift of the source coordinate in the course of
3r ' x i
N = ⎛ -------g⎞ ---- .
i evolution:
(242)
⎝ 4⎠r i
3r 'gI ( x – y J )
∑
i
Substituting Eqs. (240) and (250) into the conformal N = --------
- --------------------
4 x – yJ (250)
integral, we arrive at the equation J
– µ 1 ( z )r J
× [ ( 1 – γ Ψ )e – γ Ψ cos µ 2 ( z )r J ].
i 2
r r
ds c = ⎛ 1 – ----g⎞ dη – ⎛ 1 + ----g⎞ ⎛ d x i + --- ---- r 'g dη⎞ , (243)
2 2 3x
⎝ r⎠ ⎝ r⎠⎝ 2r ⎠
The conformal integral squared
2 2 2
where ds c = ds /a (η) . 2 2
ds c = ( 1 + Φ )dη – ( 1 – Ψ ) ( d x + N dη )
i i 2
(251)
yields the equation for photon momentum,
5.10. Cosmological Generalization
of Newtonian Potential µν i i 2 2
pµ pν g
( p 0 + N p ) (1 – Φ) – p j ( 1 + Ψ ) = 0, (252)
In the case of the equation of the rigid state γ2 = 1,
we have which gives
α φ = 6, α ψ = 6 – 2y, β = 15 – 2y, i i
p 0
–N p + ( 1 + Φ + Ψ ) p , p = pi .
2
(253)
2 2 2 4
(244)
h = 3ρ s / ( 2ϕ ) = ( 3/2 )H 0 ( 1 + z) Finally, the relative value of spatial fluctuations of
instead of Eqs. (230); hence, Eqs. (233) and (234) take photon energy, expressed in terms of the metric compo-
the form nents (the potentials Φ , Ψ and the shift vector N i ), is
2
µ 1, 2 p0 – p
-----------------
i i
= – ( N n – Φ – Ψ ),
i p
n = ------i . (254)
(245) p p
2 4 2
= ( 3/2 )H 0 ( 1 + z ) [ 26 + ( y – 2 ) ± ( 6 – y ) ].
Thus, the spatial anisotropy of photon energy fluctu-
Here, y = 3, 1 is the Lichnerowicz index, and γΨ and γΦ ations in the photon flow is due to minimal-surface con-
are given by Eqs. (235). dition (250)

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

 TETRAD FORMALISM AND REFERENCE FRAMES 131

5.11. Cosmological Generalization gives the radial component of the shift vector satisfying
of Schwarzschild Solution the minimal-surface condition
In order to isolate the Laplace operator ∆ = ∂(a)∂(a) p ψ
0. (268)
from Eqs. (189) and (196), we substitute
Thus, the Schwarzschild solution is an approxima-
ψ = ψ 
7
(255) tion similar to the Coulomb potential in electrodynam-
ics, when both source motion and field dynamics are
into them: disregarded.
〈  〉
 t  = 〈  t 〉 ,  t = ----------------t -,
 5.12. Transformation of Reference Frames
where The Lorentz invariance of GR in a specific reference
frame can be treated similar to the Lorentz invariance of
 t  ≡ 4P – S + ψ  ψ ( 3T 0 – T k ),
5 0 k electrodynamic equations in SR [12]. Therefore, to per-
(256)
form the transformation to a new reference frame one
 t ≡ P + K + ψ  ψ T 0 .
5 0
(257) should return to the initial variables and interval
2 2 2
As a result, the expressions for P and S take the form ds = ω ( 0 ) – ω ( a ) (269)

4ϕ
2
ϕ
2 and choose new variables and linear forms ω ( 0 ) and
P = ---------  ψ ∆ψ + -----  ψ R ( e ), (258)
3 6 ω ( a ) related to the initial forms by the Lorentz transfor-
2
mations
ϕ
S = ----- (  ψ ∆ψ – ψ∆  ψ ). (259) 1
3 ω ( 0 ) ( ζ, x 1, x 2, x 3 ) = ------------------- ω ( 0 ) ( ζ, x 1, x 2, x 3 )
2
If P and S are equal to zero, then 1–v
v
∆  ψ = 0, ∆ψ = 0, R ( e ) = 0. (260) + ------------------- ω ( 1 ) ( ζ, x 1, x 2, x 3 ),
2
These equations have the solutions 1–v

rg ( ζ ) 1
ω ( 1 ) ( ζ, x 1, x 2, x 3 ) = ------------------- ω ( 1 ) ( ζ, x 1, x 2, x 3 )
ψ = 1 + -----------
-, (261) 2
4r 1–v
r (ζ) v
+ ------------------- ω ( 0 ) ( ζ, x 1, x 2, x 3 ).
 ψ = 1 – -----------
g
-, (262) 2
4r 1–v
where rg(ζ) = M3/4πϕ2 is the gravitational radius. Then, the Hamiltonian description of the Dirac vari-
It is easy to verify that the conventional Schwarzs- ables and coordinates is to be formulated; in this case,
µ
child metric in the vacuum, T ν = 0, can be treated as a the dependence of the new forms ω on the spacetime
solution of the Einstein equations under the assumption coordinates ( ζ , x ) should coincide with the depen-
of mass independent of the geometric time: rg(ζ) ~ dence of ω on (ζ, x).
rg(ζ0) = const. The generalization of the Schwarzschild
metric to the conformally flat metric can be written in 6. CONCLUSIONS
terms of Cartan forms (152) and (153):
The main goal of theoretical physics is to establish
 several physical principles to explain all observable
ω ( 0 ) = -------ψ- dζ, (263)
ψ effects just as a few Euclidean axioms and logical laws
make it possible to prove many geometry theorems.
ω ( r ) = ψ ⎛ dr + ----------
- dζ⎞ ,
2 V' In modern physics, symmetry properties serve as
(264)
⎝ r ψ ⎠
2 6
such fundamental principles. The following statement
of Weyl [66] is worth reminding: {The conclusion hav-
2 2
ω ( θ ) = ψ r dθ, (265) ing become the key principle of modern mathematics is
2 2
ω ( ϕ ) = ψ r sin θdϕ. (266)
as follows: every time you deal with an object hav- ∑
ing a structure, you should try to find its automorphism
Here, group; i.e., the group of the transformations unaffect-
r
ing structure relationships of the object.} From this
viewpoint, transformations of reference frames form an
∫ drr ψ ( r, ζ )
6
V ( r, ζ ) =
2
(267) automorphism group in mechanics, while the equations

PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1 2006

132 ZAKHAROV et al.
of motion derived by variation of action are invariant
structure relationships.
General Relativity first introduced new geometric
principles into the theory of dynamical systems. The
principles were explicitly formulated by Hilbert in
“Foundations of Physics” [15], where the action func-
tional, the symmetry of reference frames, and the equa-
tions of motion were complemented by the definition of
geometric interval, the gauge symmetry, and the con-
straint equations. This stimulated the establishment of
the gauge symmetry of Maxwell equations [4] and,
finally, the development of modern unified gauge theory.
As shown in this review, the same geometric formu-
lation of SR makes it possible to describe relativistic
effects by dynamical variational equations rather than
Lorentz kinemetric transformations. Moreover, the
notions of “Foundations of Physics” [15] help us under-
stand the history of SR. It should be noted that the mod-
ern quantum field theory [18, 19] is nothing but the first
and second quantization of the energy constraint equa-
tion referred to as a mass surface of relativistic particle.
In the spirit of the unified geometric formulation of
both SR and GR with the use of the Hilbert variational
principle, the following question arises. Is it possible to
adopt modern quantum-field methods (based on relativ-
istic dynamics in a specific reference frame) to the Hil-
bert formulation of GR? If this was the case, then GR
would be quantized by analogy to the first and second
quantization in SR. Almost all notions of such a quan-
tization of GR turned out to be formulated by a number
of authors. They used relativistic and gauge symmetries
to define physical gauge-invariant variables and observ-
ables of GR in a specific reference frame, as well as to
construct a GR action. These notions are listed below.
(i) The difference between transformations of refer-
ence frames and gauge transformations in GR was first
pointed out by Fock in 1929 [21]. He introduced a
gauge-invariant orthogonal basis depending on the ref-
erence frame.
(ii) A specific reference frame corresponding to the
quantization of GR was defined by Dirac in 1958 [23].
(iii) A group of gauge transformations leaving the
geometric interval invariant in a kinematic reference
frame was considered by Zel’manov in 1976 [25]. The
group contains a subgroup of transformations of a coor-
dinate evolution parameter and requires that the gauge-
invariant time be introduced as a variable of field space
of events.
(iv) The time as a variable was identified with a cos-
mological scale factor by Wheeler and De Witt in 1967
[39]. They introduced into GR the notion of a field
space of events. The relativistic universe moves in this
space on a world hypersurface characterized by a geo-
metric interval measured by a comoving observer, with
this space similar to the Minkowski space in SR.
(v) The transformation that isolates the cosmologi-
cal scale factor (treated in GR as an independent vari-
PHYSICS OF PARTICLES AND NUCLEI
able [27]) from all field variables was found by
Lichnerowicz in 1944 [63].
(vi) To define gauge-invariant physical measurable
quantities (time, one-particle energy, particle number,
number of universes, etc.) the Hamiltonian reduction
[27, 28] and the correspondence principle (formulated
by Einstein [34] in SR to define particle energy) were
employed. Expanding the reduced action in physical
fields as suggested by Einstein, we obtain the Wheeler–
De Witt cosmology [39] as the first-order approxima-
tion. The second-order approximation yields the con-
ventional field theory in the conformally flat spacetime.
In this case, gauge-invariant observables coincide with
conformal ones or relative coordinates and time, intro-
duced by Friedmann. In terms of these variables, the
universe is at rest and all mass increases with the cos-
mological factor [37, 38].
(vii) The quantum gravity is defined by Wheeler and
De Witt as the first and second quantization of energy
constraint in a specific reference frame [39]. This com-
pletely accommodates to the modern quantum field the-
ory [18, 19] verified by the huge amount of high-energy
physics data.
We demonstrated that the quantum gravity consid-
ered above makes it possible to resolve a series of prob-
lems of modern cosmology: Hubble evolution, universe
and matter creation from the vacuum, evolution and
horizon of the universe, arrow of time, initial data, and
cosmological singularity. We listed a number of theo-
retical and experimental arguments in favor of this
quantum gravity, which allows us to answer the topical
questions concerning origin of dark energy and dark
matter, the universe and matter creation, and the matter
evolution in the universe.
In this case, the primordial temperature perturbation
spectrum of the CMB temperature in the inflation
model was shown to be explained by the double count-
ing of the scale factor. This makes it impossible to con-
sistently deduce the Hamiltonian formalism and quan-
tum theory. On the other hand, it is the Hamiltonian
quantum theory considered in Section 4 that allows us
to describe CMB and its primordial temperature pertur-
bation spectrum as a product of decays of primordial W-
and Z-bosons. The bosons were created from the
Bogoliubov stable vacuum when their mass coincided
with the Hubble parameter. In this case, the equations
for longitudinal components of W- and Z-bosons [55]
are similar to Eqs. (237) of the inflation model, which
describe the primordial temperature perturbation
spectrum.
The version described above of the GR quantization
by extracting the scale factor reveals hidden scale sym-
metry of GR and the Standard Model. Namely, physical
laws, when treated as equations of motion, are indepen-
dent of measurement units, as well as values of initial
data in a specific reference frame. The Planck mass, as
a fundamental parameter, disappears from equations of
motion and emerges as the present-day value of a
Vol. 37 No. 1 2006
 TETRAD FORMALISM AND REFERENCE FRAMES
dynamical variable evolved from a possible initial data.
For this reason, GR is mathematically equivalent to the
theory of conformal scalar field [67]. The scale symme-
try indicates the possible geometrization of Higgs par-
ticle (which need not be introduced together with Higgs
potential). Moreover, such a geometrization could pre-
dict some physical effects for experimental verification
with modern high-energy particle accelerators [68].
ACKNOWLEDGMENTS
We are grateful to B.M. Barbashov, D. Blaschke,
S.I. Vinitskii, E.A. Ivanov, E.A. Kuraev, A.N. Lipatov,
V.V. Nesterenko, S.A. Smolyanskii, and A.N. Sisakyan
for fruitful discussions. The work of A.F.Z. was sup-
ported in part by the National Natural Science Founda-
tion of China (project no. 10233050) and the National
Key Basic Research Foundation of China (project
no. TG 2000078404).
REFERENCES
1. A. Einstein, Sitzungsber. K. Preuss. Akad. Wiss. B 44
(2), 778 (1915); 48 (2), 844 (1915); Collection of Scien-
tific Works, Ed. by I. E. Tamm, Ya. A. Smorodinskii, and
B. G. Kuznetsov (Nauka, Moscow, 1965), Vol. 1, Article 34,
p. 423; Article 37, p. 448 [in Russian].
2. A. Einstein, Vossissche Zeitung 26, 33 (1914); Collec-
tion of Scientific Works, Ed. by I. E. Tamm, Ya. A. Smoro-
dinskii, and B. G. Kuznetsov (Nauka, Moscow, 1965),
Vol. 1, Article 31, p. 397; pp. 282, 284, 295, 425, 456,
560 [in Russian].
3. H. Weyl, Sitzungsber. Berl. Akad., 465 (1918).
4. H. Weyl, Z. Phys. 56, 330 (1929).
5. V. Fock, Z. Phys. 39, 226 (1926).
6. V. A. Fock, Theory of Space, Time, and Gravitation
(Gostekhizdat, Moscow, 1955; Pergamon, Oxford, 1964).
7. C. N. Yang and R. L. Mills, Phys. Rev. 96, 191 (1954).
8. I. Yang, Usp. Fiz. Nauk 132 (1), 169 (1980).
9. R. Utiyama, Phys. Rev. 101, 1597 (1956).
10. T. W. B. Kibble, J. Math. Phys. 2, 212 (1961).
11. P. Dirac, Proc. R. Soc. London, Ser. A 114, 243 (1927);
Can. J. Phys. 33, 650 (1955).
12. I. V. Polubarinov, Phys. Part. Nucl. 34, 377 (2003).
13. V. N. Pervushin, Phys. Part. Nucl. 34, 348 (2003); hep-
th/0109218; L. D. Lantsman and V. N. Pervushin, Yad.
Fiz. 66, 1418 (2003) [Phys. At. Nucl. 66, 1384 (2003)].
14. V. V. Nesterenko, “On Interpretation of Noetherian Iden-
tities,” Preprint No. R2-86-284, OIYaI (Joint Institute for
Nuclear Research, Dubna, 1986).
15. D. Hilbert, Nachr. Ges. Wiss. Goettingen, Math.-Phys.,
No. 3, 395–407 (1915); Variational Principles of Mechan-
ics, Ed. by L. S. Polak (Fiz.-Mat. Literatura, Moscow,
1959), p. 589 [in Russian].
16. D. Hilbert, Math. Ann. 92, 1 (1924).
17. E. Noether, Nachr. Ges. Wiss. Goettingen, Math.-Phys.,
No. 2, 235 (1918); Variational Principles of Mechanics,
Ed. by L. S. Polak (Fiz.-Mat. Literatura, Moscow, 1959),
p. 611 [in Russian].
PHYSICS OF PARTICLES AND NUCLEI Vol. 37 No. 1
133
18. N. N. Bogoliubov, A. A. Logunov, and I. T. Todorov, Intro-
duction to Axiomatic Quantum Field Theory (Nauka, Mos-
cow, 1969; Benjamin, New York, 1975).
19. N. N. Bogoliubov and D. V. Shirkov, Introduction to the
Theory of Quantized Fields (Nauka, Moscow, 1973;
Wiley, New York, 1980).
20. L. D. Faddeev and V. N. Popov, Usp. Fiz. Nauk 111, 427
(1973) [Sov. Phys. Usp. 16, 777 (1973)].
21. V. A. Fock, Z. Phys. 57, 261 (1929).
22. Yu. S. Vladimirov, Frames in Gravitation Theory (Ener-
goizdat, Moscow, 1982) [in Russian].
23. P. A. M. Dirac, Proc. R. Soc. London A 246, 333 (1958);
Phys. Rev. 114, 924 (1959).
24. R. Arnowitt, S. Deser, and C. W. Misner, Phys. Rev. 116,
1322 (1959); 117, 1595 (1960); 122, 997 (1961).
25. A. L. Zel’manov, Dokl. Akad. Nauk SSSR 227, 78
(1976) [Sov. Phys. Dokl. 227, 78 (1976)].
26. P. A. M. Dirac, Lectures on Quantum Mechanics
(Yeshiva University, New York, 1964; Mir, Moscow,
1968).
27. M. Pawlowski and V. N. Pervushin, Int. J. Mod. Phys. A
16, 1715 (2001); hep-th/0006116.
28. B. M. Barbashov, V. N. Pervushin, and D. V. Proskurin,
Teor. Mat. Fiz. 132, 181 (2002).
29. D. E. Burlankov, Usp. Fiz. Nauk 174 (8), 899 (2004)
[Phys. Usp. 47 (8), 833 (2004)].
30. H. Poincaré, Bull. Sci. Math., Ser. 2 (Paris) 28, 302
(1904).
31. A. Einstein, Collection of Scientific Works, Ed. by
I. E. Tamm, Ya. A. Smorodinskii, and B. K. Kuznetsov
(Nauka, Mocsow, 1965), Vol. 1, Article 15, p. 175 [in
Russian].
32. A. A. Logunov, Lectures in Relativity and Gravitation
Theory (Nauka, Moscow, 1987) [in Russian].
33. H. Poincaré, C. R. Acad. Sci. (Paris) 140, 1504 (1905);
Rendiconti del Cirocolo Matematico di Palermo 21, 129
(1906); H. Lorentz, H. Poincaré, and H. Minkowski,
Principle of Relativity (ONTI, Moscow, 1935), pp. 51–
129 [in Russian].
34. A. Einstein, Anal. Phys. 17, 891 (1905).
35. P. Jordan, Z. Phys. 93, 464 (1935).
36. V. N. Pervushin, Fiz. Elem. Chastits At Yadra 15, 1073
(1984) [Sov. J. Part. Nucl. 15, 481 (1984)]; Riv. Nuovo
Cimento 8 (10), 1 (1985); N. Ilieva and V. N. Pervushin,
Fiz. Elem. Chastits At. Yadra 22, 573 (1991) [Sov. J. Part.
Nucl. 22, 275 (1991)].
37. J. V. Narlikar, Introduction to Cosmology (Jones and
Bartlett, Boston, 1983).
38. D. Behnke, D. B. Blaschke, V. N. Pervushin,
and D. V. Proskurin, Phys. Lett. B 530, 20 (2002);
gr-qc/0102039]; in Proceeding of the XVIIIth IAP Collo-
quium “On the Nature of Dark Energy,” Paris, France,
2002; Report MPG-VT-UR 240/03; astro-ph/0302001.
39. J. A. Wheeler, Lectures in Mathematics and Physics
(Benjamin, New York, 1968); B. C. DeWitt, Phys. Rev.
160, 1113 (1967).
40. C. Misner, Phys. Rev. 186, 1319 (1969).
41. M. P. Ryan, Jr. and L. C. Shapley, Homogeneous Relativ-
istic Cosmologies (Princeton Univ. Press, Princeton,
2006
134 ZAKHAROV et al.
1975); M. P. Ryan, Hamiltonian Cosmology (Springer
Verlag, Berlin, 1972), Lect. Notes Phys., No. 13.
42. V. N. Pervushin and V. I. Smirichinski, J. Phys. A: Math.
Gen. 32, 6191 (1999).
43. N. N. Bogoliubov, J. Phys. 11, 23 (1947).
44. T. Levi-Civita, Prace Mat.-Fiz. 17, 1 (1906); S. Shan-
mugadhasan, J. Math. Phys. 14, 677 (1973); S. A. Gogilidze,
A. M. Khvedelidze, and V. N. Pervushin, J. Math. Phys.
37, 1760 (1996); Phys. Rev. D 53, 2160 (1996); Phys.
Part. Nucl. 30, 66 (1999).
45. L. N. Gyngazov, M. Pawlowski, V. N. Pervushin, and
V. I. Smirichinski, Gen. Relativ. Gravit. 37, 128 (1998).
46. M. Pavlovski, V. V. Papoyan, V. N. Pervushin, and
V. I. Smirichinski, Phys. Lett. B 444, 293 (1998).
47. B. M. Barbashov, V. N. Pervushin, V. A. Zinchuk, and
A. G. Zorin, “Hamiltonian Cosmological Dynamics of
General Relativity,” presented at The International Work-
shop Frontiers of Particle Astrophysics, Kiev, Ukraine,
2004; V. N. Pervushin, V. A. Zinchuk, and A. G. Zorin,
“Conformal Relativity: Theory and Observations,” pre-
sented at The International Conference HADRON STRUC-
TURE, Smolenice, Slovakia, 2004; gr-gc/0411105.
48. S. Gogilidze, N. Ilieva, and V. Pervushin, Int. J. Mod.
Phys. A 14, 3531 (1999).
49. E. M. Lifshits, Usp. Fiz. Nauk 80, 411 (1963) [Sov.
Phys. Usp. 6, 255 (1963)]; Adv. Phys. 12, 208 (1963).
50. J. M. Bardeen, Phys. Rev. D 22, 1882 (1980);
H. Kodama and M. Sasaki, Prog. Theor. Phys. Suppl. 78,
1 (1984).
51. V. F. Mukhanov, H. A. Feldman, and R. H. Branden-
berger, Phys. Rep. 215, 206 (1992).
52. A. G. Riess et al., Astron. J. 116, 1009 (1998); S. Per-
lmutter et al., Astrophys. J. 517, 565 (1999); A. G. Riess
et al., Astrophys. J. 560, 49 (2001); astro-ph/0104455.
53. S. Weinberg, The First Three Minutes: A Modern View of
the Origin of the Universe (Basic Books, New York, 1977).
PHYSICS OF PARTICLES AND NUCLEI
54. V. N. Pervushin, D. V. Proskurin, and A. A. Gusev,
Gravit. Cosmol. 8, 181 (2002).
55. D. B. Blaschke, S. I. Vinitsky, A. A. Gusev, et al., Phys.
At. Nucl. 67, 1050 (2004).
56. A. A. Gusev, P. Flin, V. N. Pervushin, et al., Astrophys.
47, 242 (2004).
57. A. Einstein and E. Strauss, Rev. Mod. Phys. 17, 120
(1945).
58. A. Gusev et al., in Problems of Gauge Theories, Ed. by
B. M. Barbashov and V. V. Nesterenko, JINR D2-2004-
66, (Joint Institute for Nuclear Research, Dubna, 2004),
pp. 127–130.
59. A. D. Sakharov, Pis’ma Zh. Eksp. Teor. Fiz. 5, 24 (1967)
[Sov. J. Part. Nucl. 5, 17 (1967)]; V. A. Kuzmin, V. A. Ruba-
kov, and M. E. Shaposhnikov, Phys. Lett. B 155, 36
(1961); V. A. Rubakov and M. E. Shaposhnikov, Usp.
Fiz. Nauk 166, 493 (1996) [Phys. Usp. 39, 461 (1996).
60. L. B. Okun’, Leptons and Quarks (Nauka, Moscow,
1981; North-Holland, Amsterdam, 1984).
61. A. B. Borisov and V. I. Ogievetsky, Teor. Mat. Fiz. 21,
329 (1974).
62. A. D. Linde, Physics of Elementary Particles and Infla-
tion Cosmology (Nauka, Moscow, 1990) [in Russian].
63. A. Lichnerowicz, J. Math. Pures Appl. 23, 37 (1944).
64. Y. W. York, Jr., Phys. Rev. Lett. 26, 1656 (1971).
65. J. W. York, Jr., J. Math. Phys. 14, 456 (1973).
66. H. Weyl, Symmetry (Princeton Univ. Press, Princeton,
1952; Mir, Moscow, 1970).
67. R. Penrose, Relativity, Groups, and Topology, (Gordon
and Breach, London, 1964); N. Chernikov and E. Tagi-
rov, Ann. Inst. Henry Poincaré 9, 109 (1968).
68. M. Pawlowski and R. Raczka, Found. Phys. 24, 1305
(1994); M. Pawlowski and R. Raczka, hep-ph/9503269,
hep-ph/9503270.
Vol. 37 No. 1 2006