An Extra Structure of Spacetime: A Space

of Points, Areas and Volumes 1

arXiv:gr-qc/0611050v1 8 Nov 2006

Matej Pavšič
Jožef Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia
email: [email protected]

Abstract

A theory in which points, lines, areas and volumes are on on the same footing is
investigated. All those geometric objects form a 16-dimensional manifold, called C-space,
which generalizes spacetime. In such higher dimensional space fundamental interactions
can be unified à la Kaluza-Klein. The ordinary, 4-dimensional, gravity and gauge fields
are incorporated in the metric and spin connection, whilst the conserved gauge charges
are related to the isometries of curved C-space. It is shown that a conserved generator of
an isometry in C-space contains a part with derivatives, which generalizes orbital angular
momentum, and a part with the generators of Clifford algebra, which generalizes spin.

1 Introduction
In current approaches to quantum gravity the starting point is often in assuming that at
short distances there exists an underlying structure, based, e.g., on strings and branes, or
spin networks and spin foams (see, e.g.,[1]). It is then expected that the smooth spacetime
manifold of classical general relativity will emerge as a sufficiently good approximation
at large distances. However, it is feasible to assume that what we have, even at large
distances, is in fact not just spacetime, but spacetime with certain additional structure.
The approach discussed in this contribution suggests that the long distance approximation
to a more fundamental structure is the space of extended events, i.e., points, lines, areas,
3-volumes, and 4-volumes [2]–[13]. All those objects can be elegantly represented [14] by
1
Talk presented at the XXIX Spanish Relativity Meeting ERE 2006, 4th-8th September 2006, Palma
de Mallorca, Spain

1
Clifford numbers xM γM ≡ xµ1 ...µr γµ1 ...µr , r = 0, 1, 2, 3, 4, and therefore the corresponding
space is called Clifford space (C-space).
It turns out that, since C-space is a higher dimensional space, it provides a consis-
tent description of quantized string theory [9]. The underlying spacetime can remain 4-
dimensional, there is no need for a 26-dimensional or a 10-dimensional spacetime. The
extra degrees of freedom required for consistency of string theory, described in terms of
variables X M (τ, σ) ≡ X µ1 ...µr (τ, σ), are due to extra dimensions of C-space, and they need
not be compactified; they are due to the volume (area) evolution, and are thus physical.
But since a generic component X µ1 ...µr denotes an oriented r-volume, associated with an
(r − 1)-brane (i.e., a p-brane for p = r − 1), we have that string itself (i.e., 1-brane) is
not enough for consistency. Higher branes are automatically present in the description
with functions X µ1 ...µr (τ, σ), although they are not described in full detail, but only up
to the knowledge of oriented r-volume. Because of the presence of two parameters τ, σ,
we keep on talking about evolving strings, not in 26 or 10-dimensional spacetime, but in
16-dimensional Clifford space. In general, the number of parameters can be arbitrary, but
less then 16, and so we have a brane in C-space, i.e., a generalized brane [4]. Since with the
points of a flat C-space one can associate Clifford numbers (polyvectors), this automatically
brings spinors (as members of left or right ideals of Clifford algebra) into our description.
A polyvector X µ1 ...µr γµ1 ...µr can be rewritten in terms of a basis spanning four independent
left ideals, and thus contains spinorial degrees of freedom. This means that by describing
our branes in terms of the C-space embedding functions X M ≡ X µ1 ...µr we have already
included spinorial degrees of freedom. We do not need to postulate them separately, as in
ordinary string and brane theories, where besides Grassmann even ( “bosonic”) variables
X µ , there occur also Grassmann odd (“fermionic”) or spinorial variables. In this formula-
tion we have a possible clue to the resolution of a big open problem, namely, what exactly
is string theory.
If we release the constraint of flat C-space, then we encounter a fascinating possibility,
namely that curved C-space provides a realization of Kaluza-Klein theory [10]–[13], and
since all of its 16 dimensions are physically observable to us, there is no need for compacti-
fication of the “extra” dimensions. The extra components of the C-space metric tensor are
related to the gauge fields that describe the fundamental interactions. When considering
the generalized Dirac equation in curved C-space it turns out that the extra components
of the C-space spin connection also manifest themselves as gauge fields [12, 13]. The con-
served charges consist of two contributions: one due to the “orbital angular momentum”,
and another one due to the “spin”in the “internal” space. This is analogous to the situation
that we know from the ordinary Dirac theory: instead of spacetime we have C-space, and
instead of gamma “matrices” we have now the full basis of Clifford algebra. So instead of
the ordinary angular momentum and spin, we have the corresponding momenta in C-space,

2
and in particular their components in the “internal” part of C-space, i.e., the part that
goes beyond spacetime.

2 Extending spacetime to Clifford space

In a manifold we do not have points only. We also have higher objects, such as open
and closed lines, surfaces, etc. . In a flat manifold description of such objects can be
done straightforwardly by employing vectors of various grades: ordinary vectors, bivectors,
trivectors, etc., in general r-vectors or multivectors, which are Clifford numbers.
Basis vectors (1-vectors) are generators of Clifford algebra satisfying
1
γµ · γν ≡ (γµ γν + γν γµ ) = gµν (1)
2
Higher grade vectors are given by the wedge product of vectors:
1 1
γµ ∧ γν ≡ (γµ γν − γν γµ ) ≡ [γµ , γν ] (2)
2! 2
1
γµ ∧ γν ∧ γα ≡ [γµ , γν , γα ] (3)
3!
..
.
1
γµ1 ∧ γµ2 ∧ ... ∧ γµn ≡ [γµ , γµ , ..., γµn ] (4)
n! 1 2
An oriented line element is
dx = dxµ γµ (5)
where dxµ , µ = 0, 1, 2, 3, ..., are just the ordinary differentials of coordinates (that should
not be confused with the notation in the calculus of differential forms). The dx is a vector,
expanded in terms of basis vectors γµ , its components being dxµ = dx · γ µ . Here γ µ are
reciprocal basis vectors satisfying γ µ · γν = δ µ ν .
Oriented area element is the wedge product of two line elements dx1 and dx2 :

dx1 ∧ dx2 = dxµ1 dxν2 γµ ∧ γν (6)

An arbitrary multivector element of degree r is

dx1 ∧ dx2 ∧ ... ∧ dxr = dxµ1 1 ...dxµr r γµ1 ∧ ... ∧ γµr (7)

At this stage we do not yet specify the dimensionality n of the underlying space(time);
it can be arbitrary. The grade of a non vanishing r-vector can be at most r = n.
If we are in flat spacetime manifold, we can straightforwardly integrate the above in-
finitesimal r-vector line elements, and so we obtain finite r-vectors xµ1 ...µr γµ1 ...µr describing
oriented r-dimensional areas (r-areas) associated with closed (r−1)-dimensional surfaces [6],

3
or open r-dimensional surfaces. We have thus introduced a manifold of r-areas, with xµ1 ...µr
being their coordinates. The latter manifold is called Clifford space, or C-space.
This was just a manifold of abstract r-areas that generalizes the ordinary manifold of
abstract points. We can make contact with physics by noting that in a spacetime manifold
we have “matter” consisting of all sorts of physical objects that can be point particles
or extended objects, such as strings and branes of various dimensionalities. A spacetime
manifold can be considered as a space of all possible positions that a test particle can have.
Analogously, an extended object can have many positions in a multidimensional space
of all possible object’s configurations, the so called configuration space. If extended objects
are strings or branes, then their configuration space is infinite dimensional. In refs. [4] it
was called M-space. The latter space is thus a space of all possible configurations that
a (generalized) brane can have. M-space can have metric. Different choices of M-space
metric lead to different brane theories, with different actions in the underlying spacetime.
From the point of view of spacetime we have different brane theories, but from the point
of view of M-space there is one theory which allows for different background M-space
metrics. The latter metric can become dynamical, if we include into the description a
corresponding kinetic term [4]. Thus we arrive at a background independent theory that
generalizes general relativity to M-space.
In order to avoid infinite dimensional description of branes, one can introduce a
quenched description [15, 16] working in a finite dimensional subspace of M-space. The
finite dimensional space is the space of oriented r-areas that we associate with closed
(r − 1)-branes, or open r-branes. We have thus a many–to–one mapping from the infi-
nite dimensional objects, such as branes, to finite dimensional r-areas. In other words,
we have a many–to–one mapping from infinite dimensional M-space to finite dimensional
C-space [2]–[13].
Clifford space is thus just a particular case of configuration space, associated with an
extended object. It takes into account not only an object’s center of mass position, but
also its extension, which is sampled by coordinates xµ1 ...µr .
In general, a “point” in C-space can be described by coordinates xµ1 ...µr . Two infinites-
imally separated points, with coordinates xµ1 ...µr and xµ1 ...µr + dxµ1 ...µr , define a polyvector
n
1 X
dX = dxµ1 ...µr γµ1 ...µr ≡ dxM γM (8)
r! r=0
which is a superposition of poyvectors of different grades, including the scalars.
The line element in C-space is given by the quadratic form [2]–[13]
dS 2 ≡ dX ‡ ∗ dX = dxM dxN GM N ≡ dxM dxM (9)
Here ‘*’ denotes the scalar product of two polyvectors A and B:
A ∗ B = hABi0 (10)

4
where h i0 denotes the scalar part of the Clifford product AB. Symbol ‘‡’ denotes reversion,
i.e., the operation that reverses the order of vectors, e.g.,

(γ1 γ2 γ3 )‡ = γ3 γ2 γ1 (11)

The metric is given by the scalar product of two basis Clifford numbers:
‡
GM N = γ M ∗ γN (12)

Reversion in the above definition is necessary for consistency reasons [9].

If the signature of n-dimensional spacetime we start from is pseudoeuclidean of the form
+ − − − ... , then the signature of C-space is (p, p), with p = 2n /2. This has important
consequences for string theory in particular, and for field theory in general [9, 17].
We can now envisage that physical objects are described in terms of xM =
(ω, xµ , xµν , ...). The first straightforward possibility is to introduce a single parameter
τ and consider a mapping
τ → xM = X M (τ ) (13)
where X M (τ ) are 16 embedding functions that describe a worldline in C-space. From the
point of view of C-space, X M (τ ) describe a wordlline of a “point particle”: at every value
of τ we have a point in C-space. But from the perspective of the underlying 4-dimensional
spacetime, X M (τ ) describe an extended object, sampled by the center of mass coordinates
X µ (τ ) and the coordinates X µ1 µ2 (τ ), X µ1 µ2 µ3 , X µ1 µ2 µ3 µ4 (τ ). They are a generalization of
the center of mass coordinates in the sense that they provide information about the object
2-vector, 3-vector, and 4-vector extension and orientation2 .
The dynamics of such an object is determined by the action
Z Z
1 1
I[X] = M †
dτ (Ẋ ∗ Ẋ) = M2 dτ (Ẋ M ẊM ) 2 (14)

The dynamical variables are given by the polyvector

X = X M γM = Ω1 + X µ γµ + X µ1 µ2 γµ1 µ2 + ... + X µ1 ...µn γµ1 ...µn (15)

whilst
Ẋ = Ẋ M γM = Ω̇1 + Ẋ µ γµ + Ẋ µ1 µ2 γµ1 µ2 + ... + Ẋ µ1 ...µn γµ1 ...µn (16)
is the velocity polyvector, where Ẋ M ≡ dX M /dτ .
In the action (14) we have a straightforward generalization of the relativistic point
particle in M4 : Z
1
I[X µ ] = m dτ (Ẋ µ Ẋµ ) 2 , µ = 0, 1, 2, 3 (17)
If a particle is extended, then (17) provides only a very incomplete description. A more
complete description is given by the action (14), in which the C-space embedding functions
X M (τ ) sample the objects extension.
2
A systematic and detailed treatment is in refs. [6].

5
3 Strings and Clifford space
Usual strings are described by the mapping (τ, σ) → xµ = X µ (τ, σ), where the embedding
functions X µ (τ, σ) describe a 2-dimensional worldsheet swept by a string. The action is
given by the requirement that the area of the worldsheet be “minimal” (extremal). Such
action is invariant under reparametrizations of (τ, σ). There are several equivalent forms
of the action including the “σ-model action” which, in the conformal gauge, can be written
as
κ
Z
I[X µ ] = dτ dσ (Ẋ µ Ẋµ − X ′µ Xµ′ ) (18)
2
where Ẋ µ ≡ dX µ /dτ and X ′µ ≡ dX µ /dσ. Here κ is the string tension, usually written as
κ = 1/(2πα′).
If we generalize the action (18) to C-space, we obtain
κ
Z
I[X] = dτ dσ (Ẋ M Ẋ N − X ′M XN′ )GM N (19)
2
Taking 4-dimensional spacetime, there are D = 24 = 16 dimensions of C-space. Its sig-
nature (+ + +... − − − ...) has 8 plus and 8 minus signs. This particular form of metric
suggest to define vacuum according to Jackiw et al. [18, 19] (see also [9, 17]). Then one
finds that such generalized string theory is consistent. There are no negative norm states,
and the Virasoro algebra has no central charges [10, 9]). My proposal is that, instead of
adding extra dimensions to spacetime, we can start from 4-dimensional spacetime M4 with
signature (+ − −−) and consider the Clifford space CM4 (C-space) whose dimensionality
is 16 and signature (8+, 8−). The necessary extra dimensions for consistency of string
theory are in C-space. This also automatically brings spinors into the game. It is an old
observation [20, 21] that spinors are the elements of left or right ideals of Clifford algebras.
In other words, spinors are particular sort of polyvectors. Therefore, the string coordinate
polyvectors contain spinors. This is an alternative way of introducing spinors into the
string theory.

4 Curved Clifford Space

As we can pass in the ordinary theory of relativity from flat to curved spacetime manifold,
so we can pass from flat to curved Clifford space C. This is a manifold such that at every
point X ∈ C its tangent space TX C is Clifford algebra, a vector space, whose elements are
polyvectors [13]. Among them one can chose those independent polyvectors which form a
basis or frame. For variable X we have a frame field.
It is convenient to distinguish two kinds of frame field:
(i) Coordinate frame field
γM ≡ γµ1 ...µr (20)

6
 (ii) Orthonormal frame field

γA ≡ γa1 ...ar = γa1 ∧ γa2 ∧ ... ∧ γar (21)

While the basis elements γA are defined at every point point of C-space as the wedge
product of vectors, this is not the case for γM . The relation between γM and γA is given
by
γM = eM A γA (22)
where eM A is the C-space vielbein. Eqs. (2)–(4) may hold at one point X, but not at
different points of curved C space.
At every point X of C a polyvector, an element of the tangent space TX C, can be
expanded, e.g., im terms of the basis (20) or (21). Since X may run over the manifold C,
we thus have a polyvector field.
At this point we encounter an important concept, namely, the derivative of a polyvector
field.
If acting on a scalar field, the derivative is just the ordinary partial derivative
∂φ
∂M φ = (23)
∂xM
If acting on a frame field, it defines connection, e.g., the connection for the coordinate frame
field
∂M γN = ΓJM N γJ (24)
or the connection for the orthonormal frame field

∂M γA = −ΩA B M γB (25)

If acting on a generic polyvector field, we have

∂M (AN γN ) = ∂M AN γN + AN ∂M γN = (∂M AN + ΓN K N
M K A )γN ≡ DM A γN (26)

The components DM AN are covariant derivative of the tensor analysis.

Contrary to the usual practice, we use in eqs. (23)–(26) the same symbol ∂M . In ref. [13]
we argue in detail why usage of different symbols for derivatives of different geometric
objects is unnecessary.
Curvature of C-space is defined, as usually, by the commutator of derivatives acting on
basis polyvecors:
[∂M , ∂N ]γJ = RM N J K γK (27)
or
[∂M , ∂N ]γA = RM N A B γB (28)

7
Introducing the reciprocal basis polyvectors γ M , γ A satisfying

(γ M )‡ ∗ γN = δ M N , (γ A )‡ ∗ γB = δ A B (29)

we find from (27) or (28) the explicit expressions for the components of curvature in the
corresponding basis:

RM N J K = ∂M ΓK K L K L K
N J − ∂N ΓM J + ΓN J ΓM L − ΓM J ΓN L (30)

or
RM N A B = −(∂M ΩA B N − ∂N ΩA B M + ΩA C N ΩC B M − ΩA C M ΩC B N ) (31)
A consequence of non vanishing curvature is that after the parallel transport of a
polyvector along a closed path we obtain a polyvector with different orientation. In partic-
ular this means that, if initially we have, e.g., a vector at a given point of C, then after a
round trip parallel transport to the same point we can obtain, e.g., a bivector, or in general
any superposition of vectors, bivectors, 3-vectors, etc. .
Instead of 4-dimensional spacetime we have thus 16-dimensional C-space. Since the
latter space is higher dimensional, it can provide a realization of Kaluza-Klein theory.
Good features of C-space are:

(i) There is no need for extra dimensions of spacetime. Extra dimensions are in C-space.

(ii) There is no need to compactify the “extra dimensions”. The extra dimensions of
C-space, namely ω, xµν , xµνρ , xµνρσ sample the extended objects, therefore they are
physical dimensions.

(iii) The number of components GµM̄ , µ = 0, 1, 2, 3, M̄ 6= µ, is 12, which is the same as

the number of the gauge fields in the standard model.

5 Spinors as members of left ideals of Clifford algebra

and the generalized Dirac equation
Let us consider a polyvector valued field Φ(X) on a curved C-space manifold. At every
point X ∈ C a field Φ can be expanded, e.g., in terms of the orthonormal basis according
to
Φ = φA γ A , A = 1, 2, ..., 16 (32)
where φA are complex valued scalar components.
Alternatively, we can use another basis with elements ξà ≡ ξαi ∈ IiL , α = 1, 2, 3, 4; i =
1, 2, 3, 4, where IiL is the i-th left minimal ideal of Clifford algebra3 . Expansion of a
3
For a more detailed description see [13].

8
polyvector field then reads [12, 13]

Φ ≡ Ψ = ψ à ξà (33)

Such object is the sum of four independent 4-component spinors, each in a different ideal
IiL .
By assumption a field Ψ has to satisfy

∂Ψ ≡ γ M ∂M Ψ = 0 (34)

which generalizes the Dirac equation to C-space. The above equation is covariant, be-
cause the derivative ∂M , if acting on a generalized spinorial basis elements ξà , gives the
generalized spin connection:
∂M ξà = ΓM B̃ à ξB̃ (35)
Using the latter relation, we can write eq. (34) in the form

γ M ∂M (ψ à ξà ) = γ M (∂M ψ à + ΓM à B̃ ψ B̃ )ξà ≡ γ M (DM ψ à ) ξà = 0 (36)

An action which leads to eq. (34) is [13]:

Z q Z q
n n
I[Ψ, Ψ‡ ] = d2 x |G| iΨ‡ ∂Ψ = d2 x |G| iψ ∗ B̃ ξB̃‡ γ M ξà DM ψ à (37)

n
q
where d2 x |G| is the invariant volume element of the 2n -dimensional C-space, G ≡
det GM N . We take n = 4.
A generic transformation in the tangent C-space TX C which maps a polyvector Ψ (i.e.,
a generalized spinor) into another polyvector Ψ′ is given by

Ψ′ = RΨS (38)
1 AB 1 AB
where R = e 4 ΣAB α and S = e 4 ΣAB β , with αAB and β AB being parameters of the left
and right transformations, respectively. The generators ΣAB are defined as

γA γB , if A < B


ΣAB = −ΣBA =  −γA γB , if A > B (39)

0, if A = B

Transformation (38) can be written in the form of a 16 × 16 matrix:

ψ ′Ã = U Ã B̃ ψ B̃ or ψ ′ = Uψ , U = R ⊗ ST (40)

where R and S are 4 × 4 matrices representing the Clifford numbers R and S. We see that
U is the direct product of R and the transpose ST of S.

9
 Under the transformation (40) the generalized spin connection transforms as
D̃
ΓM Ã B̃ = UD̃ B̃ U C̃ Ã Γ′M C̃ + ∂M U D̃ Ã UD̃ B̃ , i.e., ΓM = U Γ′M U−1 + U ∂M U−1 (41)

while the covariant derivative transforms as

D′M ψ ′Ã = U Ã B̃ DM ψ B̃ , i.e., D′M ψ ′ = U DM ψ (42)

where D′M ψ ′Ã = ∂M ′

ψ ′Ã + Γ′M Ã B̃ ψ ′B̃ and DM ψ Ã = ∂M ψ Ã + ΓM Ã B̃ ψ B̃ . We see that ΓM
transforms as a non abelian gauge field.
The generally covariant equation in 16-dimensional C-space contains the coupling of
spinor fields ψ Ã to non abelian gauge fields Γ′M Ã B̃ which altogether form components of
connection in the generalized spinor basis.

6 Conserved charges and isometries

In curved space in general there are no conserved quantities, unless there exist isometries
which are described by Killing vector fields. Suppose we have a curved Clifford space
which admits K Clifford numbers k α = kM α
γ M , α = 1, 2, ..., K; M = 1, 2, ..., 16, where
the components satisfy the condition for isometry, namely

α α
D N kM + DM kN =0 (43)

the covariant derivative being defined in eq. (26). We assume that such C-space with
isometries is not given ad hoc, but is a solution to the generalized Einstein equations that
arise from the action which contains a “matter” term, such as (37), and a field term defined
by means of the C-space curvature (see [13]).
Taking a coordinate system in which k αµ = 0, k αM̄ 6= 0, µ = 0, 1, 2, 3, M̄ 6= µ, the
metric and the vielbein can be written as4
Gµν GµM̄ ea µ ea M̄
   
GM N = , eA M = (44)
GM̄ ν GM̄ N̄ eĀ µ eĀ N̄

Here
ea M̄ = 0 (45)
whilst the components eĀ µ can be written in terms of Killing vectors and gauge fields
Wµα (xµ ):
eĀ µ = eĀ M̄ k αM̄ Wµα , ∂M̄ Wµα = 0 (46)
4
This is a C-space analog of the Kaluza-Klein splitting usually performed in the literature. See, e.g.,
[22, 23].

10
If we set the C-space torsion to zero and calculate the connection ΩABM by using eqs. (44)–
(46), we obtain an analogous result as given, e.g., in ref. [22]:
1
ΩM̄ N̄ µ = k[αM̄,N̄ ] Wµα (47)
2
where k[αM̄ ,N̄] ≡ ∂N̄ kM̄
α α
− ∂M̄ kN̄
Let us now rewrite the C-space Dirac equation by using eqs. (43)–(47). Omitting the
terms due to the C-space torsion, we obtain
1
   
(4) µ
γ ∂µ − Ωab µ [γ a , γ b ] − q α Wµα + ... + γ M̄ ∂M̄ + ... ψ = 0 (48)
8
µ
where γ (4) = ea µ γ a are 4-dimensional coordinate basis vectors, and
1 α
q α = k α M̄ ∂M̄ + k e M̄ e N̄ ΣĀB̄ (49)
8 [M̄ ,N̄] Ā B̄
are the charges, conserved due to the presence of isometries k α M̄ . They are the sum of the
coordinate part and the contribution of the spin angular momentum in the “internal” space,
spanned by γ M̄ . The coordinate part is the projection of the linear momentum onto the
Killing (poly)vectors, and can in particular be just the orbital angular momentum of the
“internal” part of C-space. The first term that contributes to the charge q α comes from the
vielbein according to eq. (46), whilst the second term comes from the connection according
to eq. (47). Both terms couple to the same 4-dimensional gauge fields Wµα , where the index
α denotes which gauge field (which Killing polyvector), and should not be confused with
the spinorial index, used in Dirac matrices.
In eq. (48) we explicitly displayed only the the most relevant terms which contain the
the ordinary vierbein ea µ and spin connection Ωab µ (describing gravity and torsion), and
also Yang-Mills gauge fields Wµα which, as shown in eqs. (46),(47), occur in the C-space
vielbein and in the C-space “spin” connection. We omitted the terms arising form the
C-space torsion

7 Discussion
Clifford space provides a promising approach to the unification of fundamental interactions.
At first sight one might think that the signature (p, p) brings ghosts into the description.
This is not the case, if we adopt the Jackiw et al. definition of vacuum [18, 19, 17, 9]. Then
such pseudoeuclidean signature turns out to be welcome for string theory [9] and for the
resolution of the cosmological constant problem [17].
Another possible obstacle could be seen in the Coleman-Mandula theorem [24] which
forbids nontrivial mixing of spacetime and internal symmetries. But all such theorems

11
are based on certain, often tacit, assumptions. One such tacit assumption (technically
expressed in terms of certain properties of S-matrix) in the derivation of the Coleman-
Mandula theorem is that probability conservation refers to spacetime, and not to the
“internal” space. In other words, the internal space was not taken on equal footing as
spacetime at the very beginning. However, it was shown by Pelc and Horwitz [25] that
Coleman-Mandula theorem can be extended to a higher dimensional space. In general,
in Kaluza-Klein approach conserved generators associated with gauge charges come from
the extra dimensions. Extra dimensions have in principle the same role as the ordinary
four dimensions, and there are the transformations that transform one into the other.
Thus internal symmetries, associated with extra dimensions, can in general nontrivially
mix with four spacetime dimensions. Analogously holds for Clifford space. In particular,
however, a curved higher dimensional space can admit isometries, and can then be written
as the direct product of 4-dimensional spacetime and the internal space. We have shown
that the conserved charges, which are due to the isometries of curved Clifford space, have
two contributions: one from the “orbital” angular momentum in the “internal” part of
C-space, and the other from the “internal” spin. This could have some important physical
consequences that need to be explored further.

References
[1] Perez A 2003 Class. Quant. Grav. 20 R43

[2] Pezzzaglia W 1999 Dimensionally democratic calculus and principles of polydimen-

sional physics Preprint gr-qc/9912025

[3] Castro C 1999 Chaos, Solitons and Fractals 10 295

[4] Pavšič M 2001 The Landscape of Theoretical Physics: A Global View; From Point
Particles to the Brane World and Beyond, in Search of Unifying Principle (Dordrecht:
Kluwer).

[5] Pavšič M 2001 Found. Phys. 31 1185 (Preprint hep-th/0011216)

[6] Pavšič M 2003 Found. Phys. 33 1277 (Preprint gr-qc/0211085)

[7] Castro C and Pavšič M 2002 Phys. Lett. B 539 133 (Preprint hep-th/0110079)

[8] Castro C and Pavšič M 2005 Prog. Phys. 1 31

[9] Pavšič M 2005 Found. Phys. 35 1617 (Preprint hep-th/0501222)

12
[10] Pavšič M 2004 Clifford space as a generalization of spacetime: Prospects for unification
in physics Preprint hep-th/0411053

[11] Castro C 2005 Found. Phys. 35 971

[12] Pavšič M 2005 Phys. Lett. B 614 85 (Preprint hep-th/0412255)

[13] Pavšič M 2005 Spin gauge theory of gravity in Clifford space: A realization of Kaluza-
Klein theory in 4-dimensional spacetime Preprint gr-qc/0507053

[14] Hestenes D 1966 Space-Time Algebra (New York: Gordon and Breach); Hestenes D
and Sobcyk G 1984 Clifford Algebra to Geometric Calculus (Dordrecht: D. Reidel)

[15] Ansoldi S, Aurilia A, Castro C and Spallucci E 2001 Phys. Rev. D 64 026003 (Preprint
hep-th/0105027)

[16] Aurilia A, Ansoldi S and Spallucci E 2002 Class. Quant. Grav. 19 3207 (Preprint
hep-th/0205028)

[17] Pavšič M 1999 Phys. Lett. A 254 119 (Preprint hep-th/9812123)

[18] Cangemi D, Jackiw R, and Zwiebach B 1996 Annals of Physics 245 408

[19] Benedict E, Jackiw R, and Lee H J 1996 Phys. Rev. D 54 6213

[20] M. Riesz 1947 Sur certaines notions fondamentales en théorie quantiques relativiste,
in Dixième Congrès Math. des Pays Scandinaves (Copenhagen: Jul.Gjellerups Forlag)
pp. 123-148

[21] Teitler S 1965 Supplemento al Nuovo Cimento III 1

[22] Luciani J F 1978 Nucl. Phys. B135 111

[23] Witten E 1981 Nucl. Phys. B186 412

[24] Coleman S and Mandula J 1967 Phys. Rev. 159 1251

[25] Pelc O and Horwitz L P 1997 J. Math. Phys. 38 139 (Preprint hep-th/9605147)

13