1
Riemannian Geometry Framed as a Non-Commutative Algebra of Observables
Joseph E. Johnson1
1. Distinguished Professor Emeritus, Department of Physics & Astronomy, University of South Carolina, Columbia SC, 29208
Abstract: This paper reframes Riemannian geometry (RG) as a generalized Lie algebra allowing the
equations of both RG and then General Relativity (GR) to be expressed as commutation relations among
fundamental operators. We begin with an Abelian Lie algebra of n operators, X whose simultaneous
eigenvalues, y, define a real n-dimensional space R(n). Then with n new operators defined as independent
functions, X’(X), we define contravariant and covariant tensors in terms of their eigenvalues, y’, on a
Hilbert space representation. We then define n additional operators, D,whose exponential map is to
translate Xas defined by a noncommutative algebra of operators (observables) where the “structure
constants” are shown to be the metric functions of the X operators thus allowing for spatial curvature
resulting in a noncommutativity among the D operators. The Doperators then have a Hilbert space
position-diagonal representation as generalized differential operators which, with the metric, written as a
commutator, can express the Christoffel symbols, and the Riemann, Ricci and other tensors as
commutators in this representation. Traditional RG and GR are obtained in a position diagonal
representation of this noncommutative algebra of operators. Our motivation was suggested by the fact that
Quantum Theory (QT), Special Relativity (SR), and the Standard Model (SM) are framed and well-
established in terms of Lie algebras. But GR, while also well-established, is framed in terms of nonlinear
differential equations for the space-time metric and space-time variables. We seek to provide a more
general framework for RG to support an integration of GR, QT, and the SM by generalizing Lie algebras
as described.
Keywords: Riemannian Geometry, Lie algebra, Quantum Theory, metric, Heisenberg algebra >November 18, 2019 version
2
1. Introduction
Lie algebras and the Lie groups which they generate have played a central role in both
mathematics and theoretical physics since their introduction by Sophius Lie in 1888 [1]. Both relativistic
quantum theory (QT) and the gauge algebras of the phenomenological standard model (SM) of particles
and their interactions are framed in terms of observables which form Lie algebras and are firmly
established [2,3,4,5]. The Heisenberg Lie algebra (HA) among (generalized) momentum and position
operators, [D, X], in QT also has applications in mathematics in studies related to Fourier transforms and
harmonic analysis [6,7,8,9]. Likewise, in QT one has the Poincare symmetry Lie algebra (PA) of space-time
observables whose representations define free particles.
But the theory of gravitation as expressed in Einstein’s general theory of relativity (GR), although
also firmly established, is formulated in terms of a Riemannian geometry (RG) of a curved space-time
where the metric is determined by nonlinear differential equations from the distribution of matter and
energy [10,11]. In GR there are no operators representing observables, and thus no commutation rules to
define Lie algebras, and thus no representations of such algebras. The observables in GR are (a) the
positions of events in space-time, and (b) the metric function of position in space-time (and its
derivatives) which define the distance between events, and thereby define the curvature of space-time.
Thus QT and GR are expressed in totally different mathematical frameworks and their merger into a
single theory has been a central problem in physics for over a century. However, the space-time events in
QT are the eigenvalues of the space-time operators which are an essential part of the HA which also
contains the Minkowsky metric which defines the associated translated distance when space-time is not
curved. If the associated space were curved, one would have a metric that was a function of the position
in space-time. Such a generalized HA would no longer allow closure as a traditional Lie algebra but rather
closure in the enveloping algebra of analytic functions of the basis elements of the Lie algebra.
This led us to consider a noncommutative algebra of operators (NAO), generalizing the form of a
Lie algebra, with n operators, X, and n corresponding operators D, which by definition are to execute
infinitesimal translations in the associated representation space of the X eigenvalues, y ( = 0, 1, .n-1).
The X are to form an Abelian algebra whose eigenvalues represent a “space-time” manifold of four or a
larger number of dimensions as the associated X eigenvalues are simultaneously measureable. But we
will allow the space to be curved so the corresponding D operators will not in general commute as the
translations in such a curved space can interfere with each other. We found that this approach generalized
the HA “structure constants” to be the Riemann metric thus allowing the metric to be a function of the
position operators, X, in the algebra [12]. This generalizes the concept of a Lie algebra to allow for
“structure constants” that are functions of the operators in the algebra and thus are no longer constants
except approximately in small neighborhoods. This paper formally reframes RG [13] in terms of a NAO or
generalized HA. We show that the fundamental concepts in RG such as the coordinate transformations,
contravariant and covariant tensors, Christoffel symbols, Riemann and Ricci tensors, and the Riemann
covariant derivative can all be expressed in terms of commutation relations among the fundamental
operators. This framework is reminiscent of contractions of Lie algebras where the structure constants as
functions are modified to vary smoothly among different algebras based upon certain external parameters [14, 15, 16, 17, 18, 19] but not in the algebra itself. In a similar way, our algebra allows the structure constants to
be dependent upon the X operators in the algebra so that RG is retrieved as a position-diagonal
representation of the algebra as one moves over the Riemann manifold of X eigenvalues.
3
2. Riemannian Geometry Framed as a Noncommutative Algebraic Geometry of Observables
Consider a set of n independent linear self adjoint operators, X, which form an Abelian Lie
algebra of order n, where
[XXand where = 0, 1, 2, … (n-1). (2.1)
Consider a Hilbert space of square integrable complex functions |> as a representation space for this
algebra where a scalar product is used to normalize the vectors to unity: . The simultaneous
eigenvectors of the Abelian Lie algebra can be written as the outer product of the Xeigenvectors with the
Dirac notation
|y0> |y1>|y2> …|yn-1> = | y0, y1, y2, …yn-1 > =|y> (2.2)
where the eigenvalues y label the associated eigenvectors |y> of the X operators and where we use the
notation
X|y> = y|y> where the y are real numbers. (2.3)
These independent real variables y can be thought of as the coordinates of an n-dimensional space Rn
since each set of values defines a point in Rn. Let the eigenvectors be normalized to be orthonormal with
the scalar product
<ya|yb> = (y0a-y0
b) (y1a-y1
b)… (yn-1a-yn-1
b). (2.4)
Let the decomposition of unity
1 = dy|y><y| (2.5)
project the entire space onto the basis vectors |y> where <y|, using Dirac notation, is the dual vector to
|y>. A general vector in the representation (Hilbert) space of this Lie algebra can then be written as
|> = dy|y><y|> = dyy) |y>, (2.6)
where the functiony) gives the “components” of the abstract vector |> on the basis vectors |y>. Thus
<|> = 1 = dy <|y><y|> = dy y y
Now consider another set of n linear operators, X’, which are independent analytic functions, X’X),
of the Xoperators also forming an Abelian Lie algebra on the same representation space for this algebra
where it follows that
[X’X’
Let the X’ have eigenvectors |y’> and eigenvalues y’ given by
X’|y’> = y’|y’> where y’ are real numbers (2.9)
The same orthonormality and decomposition of unity also obtain for the |y’> vectors which are also a
complete basis for the space. Then we can let the X’X act to the left on the dual vector y’| and also act
to the right on the vector |y> as
<y’| X’X|y> = <y’| X’X|y> to give (2.10)
y’y’|y> = y’y)y’|y>. (2.11)
Thus the eigenvalues y’ = y’y)give the transformation from the y coordinates to the y’ coordinates if the
Jacobian does not vanish i.e. | y’ y| ≠ 0 which we require to be the case. Thus the operators X’X)
define a coordinate transformation in Rn between the eigenvalues (coordinates) y and the eigenvalues y’
(transformed coordinates) define Rn’. Then the set of n real variables y and the alternative variables y’
both can be interpreted as specifying the coordinates of points in this n-dimensional real space Rn with
coordinate transformations given by the functions
y’y’ y
4
It now follows that
dy’ y’ y) dy
and any set of n functions V(y) that transform as the coordinates,
V’(y’) = y’ y) V(y) is to be called a contravariant vector. (2.14)
We use the summation convention for repeated identical indices. The derivatives y transform as
y' = y y’) y (2.15)
and any such vector V(y) which transforms in this manner as
V’(y’) = y y’) V(y) is defined as a covariant vector. (2.16)
Upper indices are defined as contravariant indices while lower indices are covariant indices. Functions
with multiple upper and lower indices that transform as the contravariant and covariant indices just shown
are defined as tensors of the rank of the associated indices.
One would like to have transformations that translate one in the Rn space of the operators X (and
thus their eigenvalues y). We define a new additional set of n operators, D, that by definition translates a
point an infinitesimal distance, ds, in the Rn respectively in each corresponding direction y by using the
transformation generated by the D elements of the algebra via the exponential map with transformations:
G(ds, ) =exp(ds D/b) where b is a constant (2.17)
In this transformation is to be a unit vector in the y space, b is an unspecified constant, and ds is
defined to be the distance moved in the direction as defined below. Then
X’G XG
By taking ds to be infinitesimal, then one gets
X’Xsds)exp( ds D/b) Xsexp(- ds D
/b)
= (1 + ds D/b) Xs- ds D/b)
= X s ds D, X/bhigher order in ds., (2.19)
Thus the commutator D, Xdefines the way in which the transformations commute (interact)
with each other in executing the translations in keeping with the theory of Lie algebras and Lie groups
although in general the D & X do not close as a standard Lie algebra. If the space is flat then there is no
dependence of the commutator upon location, and thus there is no interference among the D. Then D
Xcan be normalized to ±(since D
is defined to translate X ) thus
D, XI± b ±
(2.20)
where ± is the diagonal n x n matrix with ±1 on the diagonal with off-diagonal terms zero. This is the
customary Heisenberg Lie algebra with structure constants ± and with D, D 0 for ≠. The
operator I is to commute with all elements and by definition has a single eigenvalue b, and is needed to
close the basis of the Lie algebra which is of dimension 2n+1. Thus confirming that the distance is ds:
dX s ds b/b± ds + higher order terms in ds. (2.21)
We now wish to allow for curvature in the space Rn of the X eigenvalues. Thus the [D, X]
commutator is now allowed to be dependent upon the operators X and can vary from point to point in the
space. We define the functions g(X) as generalized structure functions (no longer constants):
D, X bg(X) (with the requirement that |g|≠ 0 ) (2.22)
wherehas the single eigenvalue b with the commutators
[D , ] = 0 = [X , ], and [X , X] = 0 . (2.23)
These generalized structure functions can also be written as
5
g(X) = D, X/b
where g(X) are assumed to be analytic with g(X) defined by
g(X) g(X) = in the representation space. (2.25)
Then using (2.21) one gets
Xsds) - Xs) = dXds g(X) = ds . (2.26)
Then
g(X) dXdXds2 g(X)
= ds2 since is a unit vector on this metric, or (2.27)
ds2 g(X) dXdX proving that g(X) is the metric for the space. (2.28)
One notes that equation [2.22] would be identical to the Heisenberg algebra if one were to set
b = iħ to get D, X iħ g(X) (2.29)
since D is the translation operator on X, and then invoke the operator interpretation of quantum theory.
But for now we retain the purely mathematical framework where b is simply an undefined constant.
One notes that g(X) can have an antisymmetric component as well as a symmetric. But only
the symmetric portion of g(X) contributes to the distance metric for the space since it is contracted
with the symmetric form dXdX. The antisymmetric component of g(X) can however support a torsion
(twisting) for the transformation although not contributing to the distance function ds. This results in a
2n+1 dimensional noncommutative algebraic geometry with D, X, and I as the algebras basis elements
(observables).
One notes that the commutator [D , D ] has not yet been defined. D can be represented on the
basis vectors of the Hilbert representation space where X is diagonal as
<y|[D , X ] = <y| b g(X) as (2. 30)
<y| Db g(y) /yA(y) )<y| = b A(y) )where ∂ = g(y) (∂/∂y) (2.31)
as the representation of Don the space of eigenvectors <y| and where A(y) is an arbitrary collection of
vector functions of y. Note that this arbitrary vector function A(y) can include other terms such as
b g(y) y/y . So one could write
D’ = DA(X) (2.32)
in the commutators with X as this would not alter the commutation rules of D with X (but which does alter
the commutators [D,D]. This is the most general representation of the commutation rules with the
operators available using the scalar, vector and second rank tensor representations. Both the vector
function A(y) and the scalar function (y) could consist of multiple higher order tensor components
including g(X), arbitrary scalar functions, arbitrary contravariant vector function A(X) and derivatives of
such objects because any contravariant vector function of the X will commute with the X in the defining
commutator of D and X. The A(y) can thus support a Yang Mills gauge transformation group, acting both
on the representation space and on the A(y) vector functions. In that case the A(y) will have the
commutation rules of that algebra with additional indices supporting Yang Mills gauge transformations.
If that gauge algebra were to be extended to include g(X) then the commutators are more complex.
Since [D, X] = I g(X), this is a generalization of the normal definition of a Lie algebra since
g(X) is now a function of the position operators X which, in the position representation |y>, become the
eigenvalues which determine the position in the n dimensional space. Consequently, this “generalized Lie
algebra” has “structure functions” (not constants), g(y), which vary from point to point in the space.
6
From now on we assume the general case where g = g (y) is to be understood in the position
representation.
In the position representation one now has
<y| D|> =(b g(y) (∂/∂y) + A (y) ) (y) = ( b ∂ + A (y) ) (y) where (2.33)
(y) = <y|> and (2.34)
∂ = g(y) (∂/∂y) (2.35)
and A(y) is an yet undetermined vector function of X. In the position representation, one can write
g (∂ /∂y) (y) <y| = ∂ (y) <y| = <y| ( 1/b) [D, (X)]) (2.36)
for any function (X) allowing one to convert differential operators into commutators with D. It follows
that [D, [D, X]] ≠ 0 so that this Heisenberg type algebra is no longer nilpotent. But instead one gets
<y| [D, [D, X]] = b2 g (∂g/∂y) <y| since (2.37)
[A, g] = 0 (2.38)
as they both are only functions of X.
We have not specified the commutators [D, D ] (often written as F ) yet as they are no longer
zero but which in the position representation with b=1 give
<y| [D, D ] = [(g(y) (∂/∂y) + A (y)), (g(y) (∂/∂y)+ A (y))] <y| or (2.39)
= -(( g(y)(∂g(y)/∂y)– (g(y) (∂g(y)/∂y)) (∂/∂y)
-(( g(y)(∂A(y)/∂y)– (g(y) (∂A(y)/∂y)) + [A, A ] ) <y|. (2.40)
If the vector fields are generalized to Yang-Mills gauge fields such as in the standard model in physics
one gets
[D, D ] = -(∂g- ∂g ) ∂ (∂AC ∂AC )CABC AA, AB (2.41)
where [AA, AB ] = CABC AA ABwhere the A, B, & C indices are those of the SM. ()
3. Reframing of Riemannian Geometry in the Non-Commutative Algebra
The Christoffel symbols are given by
= (½) (∂, g + ∂, g- ∂, g ) (3.1)
and can be written in the position diagonal representation, in terms of the commutators of D with the
metric as
= (½) (1/b) ( [D, g] + [D, g] - [ D, g ] ). (3.2)
Then using
g (X) = (1/b) [D, X] one obtains (3.3)
= (½) (b-2) ( [D, [D, X]] + [D, [D, X]] - [ D, [D, X] ] ). (3.4)
The Riemann tensor then becomes
R = (1/b) ( [D, ] - [D, ] ) + ( -
) (3.5)
where is to be inserted for the Christoffel symbols using [3.4] giving only commutators. One then
defines the Ricci tensor using [3.5] for the Riemann tensor as
R = g R = (1/b) [D, X] R and also defines (3.6)
R = g R = (1/b) [D, X] R . (3.7)
It is well known that the ordinary derivative of a scalar function, V = /y , in Riemann
geometry will transform under arbitrary coordinate transformations as a covariant vector. But such a
derivative of a vector function of the coordinates will not transform as a tensor. The covariant derivative
with respect to y of a contravariant vector is
7
/y
and the covariant derivative of a covariant vector is given by
/y
where both and transform as tensors with respect to the metric g.
One recalls for Riemannian geometry that there is a Christoffel symbol on the right hand side for
each index of the tensor being differentiated. In this algebraic framework one can write the covariant
differentiation of a contravariant vector as:
i [D ](½) ( [D, [D, X]] + [D, [D, X]] - [D, [D, X] ] ) (3.10)
assuming that A is at most a function of the X operators. Thus we are able to write both the regular
derivative (first term) and complete it with the index contraction with the Christoffel symbol (second
term). It is important to distinguish this covariant differentiation from the regular differentiation that
occurs as a representation of the operator D in the position representation. It follows that we can write the
covariant derivative of any tensor in the same way but with a contraction of the Christoffel symbol with
each of the tensor indices as is well known in Riemannian geometry. When the SM is incorporated then
the [D, A] commutator must not only include the Riemann covariant correction given by (3.10) but must
also include the covariant correction due to the [B, C from the SM.
4. General Relativity Framed as a Noncommutative Algebraic Geometry of Operators:
Up to now, the framework is for an abstract space. To connect this with GR and the foundations
of physics, we first connect this framework to special relativity with the assignment of the four-position
operators X and their eigenvalues as X0 = ct, and Xi = (x, y, z) with i = 1, 2, & 3 where c is the speed of
light with SI units of meters and seconds and D as the effective four momentum with eigenvalues (E/c,
Pi). In a theory of additional dimensions, subsequent values of Xi would refer to hidden dimensions.
In general relativity the Einstein equations for the metric:
R - ½ g R + g = (8 π G/c4) T can now be written (4.1)
R + (b [D, X]) ( ½ R - ) = (8 π G/c4) T
where R and R are now given in terms of commutators as shown above in (3.6) and (3.7) while Tis the
energy-momentum tensor. Thus all terms on the LHS consist only of commutators and (4.2) is an exact
reproduction of the Einstein equations in GR expressed totally as NAO commutators thus framing GR as
a non-commutative algebra of the X, D, I, and g(X) observables and their eigenvalues. The expression of
Einstein’s equations in the form of commutators of operators does not lead to new solutions, but one can
utilize the solutions that are already known. In a strong gravitational field near a star, such as a non-
rotating white dwarf, one can treat the metric as constant using the Schwarzschild solution over a region
that is small relative to the size of the star. The radial direction can be taken as the y1 direction as the
distance to the center of the star, with
g00 = (1 – rs/y1 ) and g11 = -1/(1 – rs/y1) (4.3)
where rs = 2GM/c2 with g22 = g33 = -1 (4.4)
and where G is the gravitational constant, M is the mass of the star, c is the speed of light, and y1 is the
distance to the center of the star giving g(X) as the Schwarzschild solution. Equation 4.2 is still exactly
the classical equation for the metric but recast in NAO commutation relations.
5. On Integrating this NAO framework of GR into a Quantum Theory Framework:
At this point we have reframed RG and GR both as a NAO thus providing a classical framework
that is similar to the Lie algebra framework of the HA and PA of relativistic quantum theory as well as the
LA structure of the SM. At this point, both RG and GR are fundamentally the same exact theory as
8
before. But one can see suggestive parallels between the NAO as a generalization of the HA where the
four momentum D is not only the translation generator for space time, X, but where that commutator not
only fixes the constant b as the value iћ, but also naturally supports arbitrary vector fields as a
component of D and thus support the Yang Mills gauge basis for the standard model with D interpreted
physically as the effective four momentum. From these parallels, we invoke quantization with Din the
reframed GR (1) be taken as the four momentum with (b) the general formalism as in standard quantum
field theory with (c) the metric and vector fields Aas those of the SM, and (d) quantized with b = iћ.
This is a very difficult program which we have just begun to investigate. However, there are
certain results that follow immediately that could test the general concept of this approach: The position
and momentum operators are now to have the interpretational structure given by quantum theory with free
particles as representations of the position extended Poincare algebra. The essentially new feature is that
by virtue of the presence of a particle in a gravitational field such as near a star, the commutation rules
with the rationalized Planks constant, ћ, are effectively modified by the metric in the radial (X1) and time
(X0 ) directions with the specific prediction that in a small region, with the Schwarzschild metric, one gets
the altered uncertainty principles:
Xr Pr ≥ (ћ/2)(1/(1-rs/r)) and (5.1)
t Er ≥ (ћ/2) (1-rs/r) (5.2)
where rs = 2GM/c2 and where r = the distance to the center of the spherical mass. This is because the
generalized algebra effectively alters the value of Planks constant in both the X1 and X0 directions as well
as the wave nature of particles in to the altered local Fourier transform. This would in turn alter the
creation rate of virtual pairs in the vacuum in a strong gravitational field. What is maintained is a more
general form of the Heisenberg uncertainty principle obtained by multiplying (5.1) and (5.2) together to
obtain
t Er Xr Pr ≥ (ћ/2)2 (5.3)
while the other two uncertainty relations remain unchanged. Because the metric would be quantized along
with the vector fields in D, it follows that distance and angle in space-time would now be observationally
“granular” or “quantized” while leaving the underlying space itself as continuous.
One would now use traditional quantum field theory with the standard model intact, with all free
fields quantized as creation and annihilation operators as representations of the Poincare algebra. Then
the gravitational field g now included in D, would be quantized as the spin 2, (b0 = 0, b1 = 3 symmetric
tensor Poincare representation) massless field determined by equation (4.2) with other spin and helicity
components gauged away. Then the RHS would be expressed in terms of the symmetrized energy-
momentum density operator, T where D not only contains the vector fields of the standard model but
now also contains the gravitational tensor field g in parallel with the mediating forces of the vector
fields. The Twould be formed in the traditional way containing all fields for the energy-momentum
tensor in the Lagrangian as the source of the gravitational field as well as source terms for vector fields.
The generalization of the Fourier transform follows from <y|D|k> = <y|D|k> where the D acts
first to the left on the bra vector and then to the right on the ket vector, which is to be an eigenstate of D
with eigenvalue k giving the differential equation
(iħ g(y) (∂/∂y) + A(y) )<y|k> = (k + A(y)) <y|k>. (5.4)
When there is no vector field present and when g is constant (no y dependence & using a Minkowski
metric), then this can be solved (with normalization for a four dimensional space-time) with:
9
<y|k> = (2 )-2 exp (g yk ) as in standard quantum theory. (5.5)
But in the general case with g(y) as a function of y this is no longer a solution and in the general case
one cannot solve this equation except formally. In fact, since the D do not commute among themselves,
one does not generally have a complete set of simultaneous Deigenvectors. However, one can consider
very small regions of space where the metric is effectively a constant as giving by a modified Fourier
transform. Then the general solution would be approximately the smoothing of these local traditional
solutions into a global solution maintaining functional and derivative continuity.
6. Lie Algebra Limit with no Gravitation as the Extended Poincare Algebra.
As quantum theory is founded upon the relationship between momentum and position operations
as defined in the HA, with [X, P] = iħ, and [E, t] = iħ, then a full Lie algebra of space-time observables
must also include a four-position operator X in order to formally include the foundations of quantum
theory as well as, M, the generators of the Lorentz group. This led us previously to extend the Poincare
algebra [20,21] (P & M) by adding a four-vector position operator, X ( … = 0, 1, 2, 3) whose components
are to be considered as fundamental observables using a manifestly covariant form of the HA. We briefly
review that algebra whose representations give free particles (fields). As the X generate translations in
momentum, they do not generate symmetry transformations or represent conserved quantities but do
provide the critical observables of space-time. We choose the Minkowski metric g = (+1, -1, -1, -1) and
write the HA in the covariant form as [P, X] = iħ I g where I is an operator that commutes with all
elements and has the unique eigenvalue “1” with P = E/c, P= Px ..., X = ct, X.= x … . Here “I” is needed to
make the fifteen (15) fundamental observables in this extended Poincare algebra (X, P, M, I) close into a
Lie algebra with the structure constants as follows:
[I, P] = [I, X ] = [I, M ] = 0 (6.1)
thus I commutes with all operators and has “1” as the only eigenvalue.
[P, X] = iħ g where the I eigenvalue is iħ (6.2)
which is the covariant Heisenberg Lie algebra – the foundation of quantum theory,
[P, P] = 0 (6.3)
which insures noninterference of energy momentum measurements in all four dimensions.
[X, X] = 0 (6.4)
which insures noninterference of time and position measurements in all four dimensions.
[M, P] = iħ (gPgP
which guarantees that P transforms as a vector under Mand thus the Lorentz group M
[M, X] = iħ (gX gX
which guarantees that Xalso transforms as a vector under Mand thus the Lorentz group.
M, M] = iħ (gM + gM - gM - gM). (6.6)
which guarantees that Mtransforms as a tensor under the Lorentz group generated by M.
The representations of the Lorentz algebra are well-known [8, 9 ] and are straight forward but the
extension to include the four-momentum with the Poincare algebra make determination of the Poincare
representations are somewhat complicated. But with our extension of the Poincare algebra to include a
four-position operator, X, the representations are clearer. Because X is now in the algebra, one can now
define the orbital angular momentum four-tensor, operator L as:
LX P – X P
From this it follows that
10
[L, P] = iħ (gPgP
[L, X] = iħ (gX gX (6.9)
L, L] = iħ (g L + g L - g L - g L).
One can then define an intrinsic spin four-tensor as:
S = M - L (6.11)
with the result that
[S, P] = 0 ; [S, X] = 0 ; S, L] = 0; and (6.12)
S, S] = iħ (g S + g S - g S - g S) (6.13)
Now one can separate this EP algebra into the product of two Lie algebras, the nine parameter
HA (consisting of X, P, I) and the six parameter homogeneous Lorentz algebra (consisting of the S & I).
Thus one can write all representations as products of the representations of the two algebras. For the HA
one can choose the position representation:
X y > = y y > or the momentum representation (6.14)
P k > = k k > (6.15)
or equivalently diagonalize the mass and the sign of the energy and three momenta as
P P= m2 c2; ( Pkiwith eigenstates written as |m, ( Pk > (6.16)
The transformations between the position diagonal and momentum diagonal basis vectors are given by the
Fourier transform as is well known.
All representations of the homogeneous Lorentz group have been found by Bergmann and by
Gelfand, Neimark, and Shapiro [2,9] to be given by the two Casimir operators b0 and b1 defined as:
b02 + b1
2 – 1 = ½ ggS S (6.17)
where b0 = 0, ½, 1, 3/2, …(|b1|-1) and where b1 is a complex number defined by
b0b1 = - ¼ S S (6.18)
with the rotation Casimir operator as S2 which has the spectrum s(s+1) with the total spin
s = b0 , b0+1, …, (|b1| - 1) (6.19)
and the z component of spin:
= -s, -s+1, ….s-1, s (6.20)
Thus the homogeneous Lorentz algebra representation can be written as | b0 , b1 , s, > which joined with
the Heisenberg algebra gives the full representation space as either
|kb0 , b1 , s, > = a+ kb0, b1, s,|0> for the momentum representation or (6.21)
| y b0 , b1 , s, > = a+ y b0,b1,s,0> for the position representation. (6.22)
To obtain the effective extended Poincare algebra, one must generalize the equations above by
replacing Pwith Dand the appropriately altered commutators which contain both the gravitational
metric and the vector bosons A as in equation (2.30) and likewise change (5.7) to
LX D–X D
7. Conclusions and Discussion:
1. The primary objective and result of this work is to reframe RG as a noncommutative algebraic
geometry following the customary expression of differential equations in terms of commutators of
operators in Lie algebras thus casting differential equations in a more powerful and formal
mathematical context. A RG can now be seen in the form of commutation rules with the differential
equations for the Christoffel connections and the Ricci, Riemann, and other tensor expressions
11
expressed as commutators of operators generalizing the framework of a Lie algebra. The potential
applicability to geometry is then of interest mathematically.
2. The second result is the subsequent reframing of Einstein’s equations and GR as a noncommutative
algebra of operators corresponding to observables thus providing a framework for GR that is more
compatible with the foundation of QT and the SM as a Lie algebras of observables expressed as
commutator products with the associated interpretation of interference of measurements. This
reframing does not in itself change GR but reexpressed it as commutators as described.
3. Our next objective will be to replace the effective momentum of QT and the SM with this more
general form for D described above that includes the metric as a function of operators that is to be
quantized as a massless spin 2 graviton as well as the vector particles that mediate the electroweak
and strong forces as described in the SM. Our first result in that direction is that the generalized HA,
[D, X] = iћ g(X), in the framework of QT implies that the interference of the simultaneous
measurement of momentum and position is to be generalized from the constant ћ to the operator ћ
g(X). This in turn suggests that the metric of space-time curvature represents the degree of
momentum-position interference of measurement and conversely. It also has the potential observable
and testable result that the uncertainty relations, between position and momentum (in the direction of
the gravitational field) and between energy and time, are altered. This could be possibly tested for a
Schwarzschild or Kerr metric near a gravitating body both with the altered Fourier transform, atomic
energy states, and virtual pair production.
4. The fourth result is the suggested replacement of the momentum operator in the Poincare,
Heisenberg, and Lorentz algebras with the D operator resulting in a generalization of the Lie algebra
framework as a NAO with the inclusion of the metric operator. The resulting commutation rules for
rotation and Lorentz transformations in a Riemannian geometry are very complex expressions but are
well defined and can be investigated in small regions of space-time where the metric is known and is
approximately constant. The Casimir operators are g(X) DD = m2c2, ½ ggL L = b0
2 + b12 – 1,
and - ¼ L L = b0b1 and as well as with the L operator for orbital angular momentum replaced
by S and also by M. We are currently investigating these equations.
5. The fifth result is that this framework exactly reduces to Einstein’s equations in an astronomical
environment with no quantum effects and where the energy-momentum tensor is taken with its
classical value. Likewise, if there is no gravitational force then the metric is Minkowski and one has
the standard QT and SM systems. But when all forces and scales are present, then this formulation is
formidable but can yet can be viewed in approximation when the metric is known and for small
domains where it can be taken as constant.
6. It is also of interest to note that the functions g(y) can contain an antisymmetric component related
to torsion although this component does not contribute to the metric for distance [22, 23,24], but does
allow more freedom in the connection as explored by Einstein and Cartan. And most important,
this framework can be extended to higher dimensions as in string theory as there is no restriction of
the space-time to four dimensions.
Acknowledgements: I am deeply appreciative of discussions and suggestions from Dr. James Knight.
12
References:
1. S. Lie, "Theorie der Transformations-gruppen I, II, III," Unter Mitwirkung von F. Engel. Leipzig
1888, 1890, 1893.
2. Weinberg, Steven (1995). “The Quantum Theory of Fields“1, 2, 3. Cambridge: Cambridge University
press. ISBN 978-0-521-55001-7
3. Schweber S. S. (1961, 1989) “An Introduction to Relativistic Quantum Field Theory”. Row, Peterson,
and Company. ISBN 978-0-486-44228-0
4. J.F. Donoghue, E. Golowich, B.R. Holstein (1994). “Dynamics of the Standard Model.” Cambridge
University Press. ISBN 978-0-521-47652-2.
5. Mathew D. Schwartz (2014) “Quantum Field Theory and the Standard Model,” Cambridge, ISBN 9-
781107-034730
6. www.math.columbia.edu/~woit/notes20.pdf
7. Binz, Ernst; Pods, Sonja (2008). “Geometry of Heisenberg Groups”. American Mathematical
Society. ISBN 978-0-8218-4495-3.
8. Hall, Brian C. (2015). “Lie Groups, Lie Algebras, and Representations: An Elementary
Introduction”. Graduate Texts in Mathematics. 222 (second ed.). Springer. ISBN 978-3319134666.
9. Howe, Roger (1980). "On the role of the Heisenberg group in harmonic analysis". Bull. Amer. Math.
Soc. 3 (2): 821–843. MR 578375. doi:10.1090/s0273-0979-1980-14825-9.
10. Weinberg, Steven (1972), Gravitation and Cosmology, John Wiley, ISBN 0-471-92567-5
11. Misner, Charles W.; Thorne, Kip. S.; Wheeler, John A. (1973), Gravitation, W. H. Freeman, ISBN 0-
7167-0344-0
12. Johnson, Joseph E. (2016), An Integration of General Relativity with Quantum Theory and the
Standard Model. http://arxiv.org/abs/1606.00701
13. Eisenhart, Luther Pfahler Eisenhart, Riemannian Geometry, Princeton University Press, 1997, ISBN
9-80691023533.
14. Y. S. Kim; Group Contractions: Inonu, Wigner, and Einstein arXiv:hep-th/9610183
15. Dooley, A. H.; Rice, J. W. (1985). "On contractions of semisimple Lie groups" (PDF). Transactions
of the American Mathematical Society. 289 (1): 185–202. ISSN 0002-
9947. MR 779059. doi:10.2307/1999695.
16. Gilmore, Robert (2006). “Lie Groups, Lie Algebras, and Some of Their Applications.” Dover Books
on Mathematics. Dover Publications. ISBN 0486445291. MR 1275599.
17. Inönü, E.; Wigner, E. P. (1953). "On the Contraction of Groups and Their Representations". Proc.
Natl. Acad. Sci. 39 (6): 510–24. Bibcode:1953PNAS...39..510I. PMC 1063815
. PMID 16589298. doi:10.1073/pnas.39.6.510.
18. Saletan, E. J. (1961). "Contraction of Lie Groups". Journal of Mathematical Physics. 2 (1):
1. Bibcode:1961JMP.....2....1S. doi:10.1063/1.1724208. ).
19. Segal, I. E. (1951). "A class of operator algebras which are determined by groups". Duke
Mathematical Journal. 18: 221. doi:10.1215/S0012-7094-51-01817-0.
20. Johnson, Joseph E. (1969) . “Position Operators and Proper Time in Relativistic Quantum
Mechanics”. Phys. Rev Vol 181, 1755-1764
21. Johnson, Joseph E. (1971), “Proper-Time Quantum Mechanics II”, Phys. Rev. D, Vol 3, No. 8, 1735-
1747.
22. Ferreira, Ana Cristina Castro (2010), “Riemannian geometry with skew torsion”, PhD Thesis
23. Élie Cartan. "Sur une généralisation de la notion de courbure de Riemann et les espaces à torsion." C.
R. Acad. Sci. (Paris) 174, 593–595 (1922).
24. Élie Cartan. "Sur les variétés à connexion affine et la théorie de la relativité généralisée." Part I: Ann.
Éc. Norm. 40, 325–412 (1923) and ibid. 41, 1–25 (1924); Part II: ibid. 42, 17–88 (1925).6