Gen eral Relativi ty an d Gravi tation , Vol. 31, No.

5, 1999
Absolute Spacetime: The Twentiet h Century Ether
Carl H. Brans
1
Recei ved October 26, 1998
All gauge theories need somet hing xed even as somet hing changes.
Underlying t he implement at ion of t hese ideas all maj or physical t heo-
ries make indispensable use of an elaborat ely designed spacet ime model
as t he somet hing xed, i.e., absolut e. This model must provide at
least t he following sequence of struct ures: point set , topological space,
smoot h manifold, geomet ric manifold, base for various bundles. The ne
st ruct ure of spacet ime inherent in t his sequence is of course empirically
unobservabl e direct ly, cert ainly when quant um mechanics is taken int o
account . This issue is at the basis of t he diculties in quant izing gen-
eral relat ivity and has been approached in many dierent ways. Here
we review an approach t aking into account the non-Boolean propert ies
of quantum logic when forming a spacet ime model. Finally, we recall
how t he fundament al gauge of dieomorphisms ( t he issue of general co-
variance vs. coordinat e condit ions) raised deep concept ual problems for
Einstein in his early development of general relat ivity. This is clearly
illustrat ed in t he notorious hole argument . This scenario, which does
not seem to be widely known t o pract icing relat ivist s, is nevert heless still
int erest ing in t erms of its impact for fundament al gauge issues.
1. INTRODUCTION
Gauge theories, to which Friedrich Hehl has contribut ed so much, explore
the mysterious fundamental role which symmetries play in our underst and-
ing of the physical world. To have a symmetry we need two part s: some-
thing xed while something else changes. Much of t he progress of modern
physical theories has come as a result of studying t his xed/ changing
dichot omy, analyzing it and suggest ing new paradigms. For most of the
1
Loyola University, New Orleans, Louisiana 70118, USA. E-mail: brans@loyno.edu
597
0 0 0 1 - 7 7 0 1 / 9 9 / 0 5 0 0 - 0 5 9 7 $ 1 6 . 0 0 / 0
c
1 9 9 9 P l e n u m P u b l i s h i n g C o r p o r a t i o n
59 8 B r an s
hist ory of physics, space, and more recently spacetime, has in some sense
or other been t he underlying xed object on which theories are writ ten and
in terms of which experiment al result s are report ed. The active, changing,
part of the symmet ry, the gauge group, consist s of coordinat e changes, or
to use more contemporary terminology, dieomorphisms. Yet, in spit e of
its uncont rovert ible central role, a thorough underst anding of spacet ime
models is still one of the most elusive goals of modern physics.
In this paper, I would like to review quest ions relat ed to these issues
using the now discredit ed ether models of the eight eenth and ninet eent h
centuries for comparison. In earlier times, the ether was some not -directly-
observable subst rat um t hought t o be needed by certain theories. For ex-
ample, in t he Newtonian gravit ational precursor to eld theory, Newton
thought that action at a dist ance was ... so great an absurdity , that ...
no man, who has in philosophical matters a competent faculty for think-
ing, can ever fall int o it (Ref. 1,vol. 1, page 68) . He speculat ed that in the
case of gravit ation, the force may be produced by varying densit ies of the
mechanical ether in the presence of gravit ating masses. In lat er periods it
played a more passive role, providing a xed, preferred reference syst em
relat ive to which velocit ies should be measured for calculat ions of Lorentz
forces and current sources in Maxwell equat ions. We will not be concerned
with t he actual details of these old ether models here, but only use t hem
as a backdrop to consider contemporary quest ions revolving around ob-
servability issues in physical models. The int erest ed reader is invit ed to
consult the massive two volume work on the ether by Whit taker [1].
Those of us who work day by day in t heoretical physics and especially
relat ivity may tend t o take for grant ed the huge package of assumptions
that we impose on our spacet ime models, most of which cannot be sup-
port ed by direct experiment. It is t his tacit acceptance of unobservable
propert ies of our model t hat mot ivat es this paper. Of course, these as-
sumptions have been quest ioned by many workers from the time of Greek
physics to t he present. Max Jammer has given us an excellent review
of this subject [2]. Without any claim to completeness we can also not e
more cont emporary work, for example, Connes [3], Rovelli [4], Madore and
Saeger [5], Heller and Sasin [6], Brans [7]. Other part icipant s of this meet-
ing have contribut ed to t his eld, including Rosenbaum [8] in his lecture
to this meeting, and L ammerzahl and Macias [9].
Many of these quest ions border on the philosophical, and philosophers
and hist orians of science have certainly made their contribut ions. Again
the lit erat ure is t oo huge to survey here. I mention only the work of
Gr unbaum [10] and Earman [11].
A b s o l u t e Sp ac e t i m e : T h e T w e n t i e t h C e n t u r y E t h e r 599
2. CONTEMPORARY SPACETIME MODELS
In almost all theories a model of spacet ime, say M , is required wit h
at least the following propert ies:
Poin t set . That is, M contains at omic elements, p 2 M , representing
idealized point event s. The existence and ident ity of this structure is
absolut ely necessary for all of t he following ones.
Top ol ogi cal m an ifol d . M must have a topology such that it is
locally Euclidean. That is, each p must lie in a neighborhood homeo-
morphic to
n
for some int eger n. Again, this property is needed for
the others.
Sm oot h m an ifold . M , in addit ion t o being locally homeomorphic
to
n
, must be locally dieomorphic to it. The local dieomorphisms
const itute t he local coordinat es needed to express smoot h funct ions
and t o operat e on them dierentially.
Geom et r ic m an i fold . M must carry a smoot h met ric, connect ion
and perhaps, as Hehl has taught us, torsion.
B u n d le st ru ct u r es. Finally, addit ional gauge structures and elds
require local pat ching t oget her of product s of M wit h models of the
eld/ gauge group space.
The advent of supersymmetry, etc. surely makes this list incomplet e,
but it does provide some idea of the extensive and detailed structure that
spacet ime models use. Here we want to point out that this involved logical
const ruction is built from bricks that are as essent ially unobservable as
were the vort ices and atoms of the mechanist ic ether of earlier times. It
is in t his sense that the title of this talk compares spacet ime to the ether.
But before getting int o the details of these foundat ional lacunae, let us
look at t he basic gauge group of spacet ime relat ivity, start ing from its
hist orical root s.
3. THE CLASSICAL ETHER
Traut man [12] has suggest ed a helpful way to look at the progress of
relat ivity in spacet ime physics using modern terminology. For more philo-
sophical and hist orical details along this pat h see Jammer [2]. Starting
with Aristotelian physics we can consider t he basic spacetime model, M ,
as a simple direct product of space with time, each having some int rinsic,
absolut e propert ies,
M = space time =
3

1
. (1)
60 0 B r an s
In this cont ext, what passed for dynamics was dened by ascribing of
nat ural places t o specic types of matt er, earthy things to the earth,
et c.
Skipping centuries of int eresting int ellectual hist ory, we arrive at the
relat ivity of Galilei and Newton. In actuality, as Jammer point s out , New-
ton was strongly at tached to the idea of some absolut e nat ure for space,
something more along t he lines of (1). Nevertheless the formalism of his
mechanics logically leads to a replacement of ( 1) with something like
M = Bundle =
space =
3

p
1
= time
(2)
Thus, spacet ime, M is a bundle over t ime,
1
, wit h space as ber,
3
.
What dist inguishes bundles from ordinary product s as in ( 1) is the ab-
sence of any natural ident icat ion of the ber over one base point wit h
that over some ot her base point . Each ber, space at a given time, is
isomorphic t o
3
, but in no nat ural or canonical way. In classical mechan-
ics, t he ber group, relat ivity gauge group G, is Galilean group. This is
essentially t he real linear ane group of dimension three. We can complet e
this pict ure t o describe classical mechanics in modern t erms by adding a
preferred at Euclidean metric on the ber, three-space, and a preferred
linear structure on the base space, time. Finally, we need a corresponding
bundle connect ion whose geodesics provide the pat hs of free part icles.
Thus, the formalism of Newtonian mechanics is more nat urally asso-
ciat ed with a bundle st ructure as in (2), rather than with absolut e space
st ructure, (1). Nevertheless, Newton felt strongly drawn to the lat ter,
perhaps in part because of his diculties wit h action-at -a-dist ance, as il-
lust rated in the quot ation in the Introduct ion. Thus, to Newton, space
needed some mechanical propert ies to enable it to transfer force and en-
ergy over dist ances in eld theories. This mechanical struct ure must be
associat ed with some subst ance, the ether, which incident ally provides
an absolut e rest frame, or the reduction of the bundle (2) to the trivial
product , (1).
The attractiveness of such an absolut e space model was reinforced
with the advent of t he unied eld t heory of electromagnet ism. In this
theory velocity appears twice. First in the Lorentz force law,
F = q(E + v B ), (3)
where v is the velocity of t he charge q being acted on by t he electro-
magnet ic eld, ( E,B ) . The quest ion left hanging is: velocity relat ive to
A bs ol u t e Sp ac e t i m e : T he T we nt i e t h C e nt u r y E t h e r 601
what reference frame? Secondly, the velocity v appears in t he source eld
equat ion

B = m
0
rv +
1
c
2
E
t
, (4)
along wit h the new quant ity of dimension speed, c 1/

e
0
m
0
. Again, the
quest ion of what reference frame should be used t o measure the source
current velocity arises. Furthermore the new, unexpect ed speed, c, reap-
pears as t he speed of electromagnet ic waves in one of t he consequences of
the vacuum eld equat ions,

1
c
2

2
t
2
E
B
= 0 . (5)
So, Maxwells unied eld theory of electromagnet ism leaves us with three
speeds, that of the source of the elds, (4) , t hat of the object on which the
elds act, (3), and the eld waves themselves, (5). The Galilean relat ivity
in ( 2) must then break down, since the presence of such speeds breaks its
invariance, and we return to some absolut e space model, ( 1), where space
is now the luminiferous et her, with spacetime
M = ether time. (6)
Of course, there were strong voices in opposit ion to the not ion of ab-
solut e space, most not ably Bishop Berkeley (Ref. 2, p.108) and lat er Mach
who referred to the not ion of absolut e space as a concept ual monst rosity
(Ref. 2, p.143) . Einst ein claimed that such ideas were inst rumental in the
evolut ion of his thinking about relat ivity.
At this point it might be appropriat e to recall all of the eort that
was put int o the design of mechanical or pseudo-mechanical models of such
an ether [2]. It is nat ural to wonder how all of t he work of contemporary
physics involving elaborat e spacet ime st ructures and superst ructures may
likewise appear in the next century.
4. END OF THE CLASSICAL ETHER: SPECIAL RELATIVITY
But of course some hundred years ago the MichelsonMorley results
forced serious ret hinking of the classical et her-space model, (6). While
Lorentz and others attempted to preserve the ether by proposing lengt h
contractions and clock dilat ions as a result of mot ion through it, Ein-
st ein cut to the heart of the matter in his principle of special relat ivity,
closely tied t o the principle of operationalism which informally claims that
60 2 B r an s
if somet hing conspires successfully against its observat ion, then its exis-
tence should not be used as part of a physical theory. Thus, if the ethers
only claim t o exist ence is as an absolut e rest reference frame, and its prop-
erties make motion t hrough it unobservable, then from the viewpoint of
physics it doesnt exist.
In fact, Einst ein t aught us to t hink in terms of a unied spacet ime
model, with no preferred a priori split ting (apart from the qualit ative
space-like, time-like, light like ones) of space from time. The transfor-
mation group preserving t hese spacetime propert ies, the gauge group of
special relat ivity is of course the Poincar e group, that is, the homogenous
Lorentz group plus translat ions. Friedrich Hehl and his colleagues have
been leaders in emphasizi ng the import ance of this group especially in the
context of general relat ivity and its generalizat ions [1315].
With the many successes of special relat ivity, it seems that the ether
has nally been put to rest. Indeed it has in t his classical sense. If you
cant observe it, it doesnt exist, is a standard motto. Or t o paraphrase an
old axiom: No st u has exist ence unt il it is observed to have existence.
But should we not apply this to stu = manifold propert ies? So, is
spacet ime the new ether? Clearly, it does not play the same mechanical
role of transmit ter of forces, as the vort ex const ituted stu of the old me-
chanical one. Also, it clearly does not provide an absolut e rest reference
frame. But it does have other, similar propert ies. It provides, in opera-
tionally unobservable ways, t he subst ratum to carry the many st ruct ures
used by modern t heories, and it is t he point of this paper that spacet ime
st ructures in modern t heories comprise a replacement for it and so have
become a new ether.
5. GENERAL COVARIANCE: EINSTEINS RELATIVITY
In addit ion to the Principle of Equivalence, which we will not con-
sider, the Principle of General Relativity and Machs Principle, are gen-
erally taken as foundat ions in models of how Einst ein arrived at General
Relativity. Of course, the act ual hist ory is more complicat ed and int erest-
ing, and the reader can consult volume one of the Einst ein Studies [16] for
a deep and accurat e account of the story.
For our purposes, it is sucient to point out t hat Einst ein was aware
of the rigid struct ure st ill remaining on the spacetime of special relat ivity
by the Lorentz metric and the associat ed preferred set of inert ial refer-
ence frames. Machs Principle addresses the issue of why t he xed stars
have const ant velocity in the inert ial frames, while the Principle of Gen-
eral Relativity proposes extending the physically acceptable frames beyond
A bs ol u t e Sp ac e t i m e : T he T we nt i e t h C e nt u r y E t h e r 603
this restricted set. In other words, while special relat ivity had weakened
the assumption of a preferred (zero) absolut e-velocity-dening ether, it re-
placed it by a preferred (zero) absolut e-acceleration-dening one. So, in
the spirit of this paper, t he next step toward generally covariant theories
was a result of re-examining and loosening previous rigid structures. John
Norton [17] has given us a thorough and highly int erest ing analysis of how
Einst ein arrived at his equat ions of General Relativity. Here we will only
skim over the issue of the ident ity of spacetime point s as illust rated in
Einst eins hole dilemma (see also Ref. 4) .
Consider a model universe, with mat ter and metric elds, T, g on a
manifold containing a region U, which will be the hole. Einst ein was
thinking in coordinat es, so let T(x), g(x) be expression of solut ion eld
equat ions in terms of coordinat es (global) x.
Now re-coordinat e, x

x9, with x = x9 in U, but not everywhere.
Then T9(x9), g9(x9) is also a solut ion, with
g
9
( x
9
) = g
9
( x) = g(x) (7)
within U. But matter and elds are dierent out side of U. So Einst ein
was deected from seeking a generally covariant theory since the following
fact would seem to be paradoxical: matter outside of U does not deter-
mine the elds inside of U uniquely (or vice versa). At this point we must
be caut ious about treating this as trivial, since we are so accustomed to
accepting general covariance as an obvious desiderat um. From the view-
point of development of the theory, there is more here than confusion about
coordinat ion.
In fact, it does seem on rst glance that Einst ein and Grossman were
confused about the expression of the same metric merely displayed in
dierent coordinat es. For example
ds
2
= dx
2
+ dy
2
, (8)
or
ds
9
2
= cosh
2
(x
9
)dx
9
2
+ dy
9
2
. (9)
Clearly (8) and ( 9) represent the same metric, and Einst ein was aware of
this. However changing the not ation in ( 9) results in
ds
9
2
= cosh
2
(x)dx
2
+ dy
2
. (10)
If we then identify the point s of the manifold with the pair (x, y), then
(8) and (10) are truly dierent metrics in some sense, alt hough they are
60 4 B r an s
dieomorphic (isometric). In fact it is possible to dene point = ordered
pair of numbers, not dieomorphism equivalence class of ordered pair of
numbers in each coordinat e system.
This discussion highlight s the dierence between the active and pas-
sive int erpretations of t he transformat ion (dieomorphism). Actually, as
discussed in detail by Norton [17], Einst eins fourt h present ation of this
argument shows t hat rather than being confused at the dierence between
(8) and (9), he was laying t he groundwork for the modern int erpretation
of dieomorphisms as physical gauge transformations.
What may be surprising about this is that it seems to rob the indi-
vidual point s of their ident ity, in the absence of a metric. In other words,
if P
1
, P
2
are two point s in a manifold, some dieomorphism maps one int o
the other. The geometry and elds around P
1
become those around P
2
, in
physically equivalent geometric manifolds, so P
1
cannot be dist inguished
from P
2
. Rovelli [4] discusses this subject in some detail, dist inguishing
spacet ime models as M
L
, local, with a part icular smoothness and at-
las as opposed to M
N
, non-local, which is the equivalence class of all
M
L
under dieomorphisms (gauge t ransformat ions) .
2
However, let us re-
call that this discussion concerns the mathematical model which is mapped
by some assumed dieomorphism ont o an absolut e point set, spacet ime.
In fact, without some underlying point set, there can be no not ion of dif-
feomorphism.
Nevertheless, the idea remains that the use of dieomorphism as phys-
ically unobservable gauge wipes out the individual ident ity of point s. In
fact, in discussing his nal generally covariant eld equat ions Einst ein said
in a letter to Schlick in 1915, thereby time and space lose t he last remnant
of physical reality. All that remains is that the world is to be conceived as a
four-dimensional (hyperbolic) continuum of 4 dimensions (Ref. 16, p.85) .
Our point here is that this continuum carries at least as much structure
as the replaced ether.
6. ABSOLUTE SPACETIME: QUANTUM THEORY
There still remains the list of absolut e spacet ime propert ies described
in the int roduct ion such as topology, smoothness, etc., which seem to be
arbit rarily chosen. This leads t o t he quest ion of regarding the role of
spacet ime as object or scratch pad. This is a quest ion certainly bordering
on philosophy, but also closely relat ed t o the operat ional foundat ions of
quant um theory.
2
Corresponding t o t his, mat hemat icians dist inguish smoot h st ruct ure from smoot h
manifold. We will discuss t his lat er.
A b s o l u t e Sp ac e t i m e : T h e T w e n t i e t h C e n t u r y E t h e r 605
The principal dist inguishing charact eristic of quant um logic as op-
posed to classical logic is that in quant um theory quest ions correspond
to project ion operat ors in Hilbert space. The logical operat ors, or,

,
corresponds to span of vector space union, while and,

, corresponds
to int ersection. This results in a non-Boolean algebra,
a

(b

c) / = (a

b)

(a

c), (11)
By contrast classical quest ions concern set-inclusion, so or,

, becomes
set union, [, and and,

, becomes \, set int ersection. Thus, for point
sets,
a \ (b [ c) = (a \ b) [ (a \ c). (12)
In other words, quantum logic is not in general consistent with the logic of
set-inclusion, which is fundamental to poin t-set questions .
For further discussion of these quest ions, see Brans [7], and especially
Marlow [18]. Maybe there is not enough in this bare-bones quant um logic
approach to work with (produce a t heory) , but many others, too many to
mention here, have also looked int o t he inuence of quant um theory on
spacet ime point set propert ies in dierent ways. However the contribut ions
of Connes [3] and Madore [19] stand out as leading to much current work
in t his area.
7. ABSOLUTE SPACETIME: CHOICE OF SMOOTHNESS
Until recently this was thought to be trivially determined by the
topology, at least for relat ively simple manifolds such as
4
. However,
it is not trivial. In fact, some of the most excit ing development s of dif-
ferential topology recently have come as a result of what can be termed
exot ic smoothness on spaces of relat ively trivial topology, for example
4
. In many respects, the development of this subject parallels that of
non-Euclidean and then dierential geometry. Thus, for many years there
were conject ures about the uniqueness of Euclidean geomet ry, bot h math-
ematically and as physics. Similarly, but more recently, there have been
conject ures that there is essentially only one way to do calculus globally
on topologically simple manifolds. The phrase how t o do calculus glob-
ally corresponds to what mathematicians call a dierentiable or smoot h-
ness structure. Physically such a structure is a global system of reference
frame pat ches covering all of spacet ime smoot hly, that is, with smoot h
(C

) coordinat e transformat ions in their overlaps. The phrase, essen-

tially only one means only one equivalence class under dieomorphisms
of t he manifold. This dierence between dierent smoot hness st ruct ures
60 6 B r an s
and non dieomorphic ones can be a slippery concept to master, but is
central to an underst anding of dierent ial topology. In a way, it is parallel
to that involved in Einst eins hole argument discussed above. Just as the
choice of dierent coordinat es may make the metric look dierent when the
underlying geometry is actually t he same, so will a recoordinat ion make
the dierent ial structure appear to be dierent, when in fact it is equiv-
alent (dieomorphic) . This equivalence inducing class of dieomorphisms
corresponds to the underlying principle of general relat ivity.
By direct calculat ion, it is possible t o show that , up to dieomor-
phisms, there is only one smoot hness structure on each
n
for n = 1, 2, 3.
For n > 4 the same result was obt ained lat er making use of cobordism
techniques. However, the case n = 4 remained an open one. Because of
the topological triviality it was nat ural t o conject ure that it too would
turn out to be trivial with respect to dierential topology. Thus, it was a
tremendous surprise when as a result of t he work of Donaldson, Freedman
and ot hers it was established that
Th eor em : ( Don ald son , Fr eed m an , et al.) There are an innity of
smoot h manifolds of t opology
4
, none of which are in the dieomorphism
class of the any other (including the standard one) .
Thus, the dieomorphism gauge does not cover the entire range of
physics on topologically trivial
4
! Do these provide new structures for
new physics? See Brans [20] for a general review of these topics, and
Asselmeyer [21] for a specic suggest ion of physical content .
8. CONCLUSIONS
In this paper we have tried to survey some of t he extensive st ruct ures
used on contemporary spacet ime models, not ing their direct physical un-
observability and reect on this rigid ne structure in the light of the
hist orical parallel of the luminiferous ether. A hundred or so years ago, it
was generally thought (apart perhaps from a few people like Gauss) that
geometry was pre-physics, a nat ural given. Now we t ake it for grant ed
that geometric struct ures carry physical elds.
Perhaps it is now appropriat e to speculat e that the mathematical
st ructures point set (e.g., Boolean, or non-Boolean...) , topological
(e.g., non-Euclidean...) , smoothness (e.g., exot ic...) , et c., might also
serve to carry physical propert ies in a manner analogous to t hat of ge-
ometry.
A b s o l u t e Sp ac e t i m e : T h e T w e n t i e t h C e n t u r y E t h e r 607
ACKNOWLEDGEMENTS
I am indebt ed to John Norton and Carlo Rovelli for very helpful com-
ments, suggest ions and insight s. Also, this work was part ially support ed
by a grant , LaSpace, R150253.
Finally, of course, we are all grat eful to Friedrich Hehl for his persist ent
clarication of t he role of various spacetime gauge structures.
REFERENCES
1. Whitt aker, E. ( 1951) . A Hi story of the Theor i es of Aether an d Electri ci ty ( 2nd. ed.,
Thomas Nelson and Sons, London) .
2. J ammer, M. ( 1993) . Concepts of Space ( 3rd. ed., Dover)
3. Connes, A. ( 1994) . Non commutative Geom etry ( English ed., Academic Press, New
York) .
4. Rovelli, C. ( 1991) . Class. Quan tum Grav. 8, 297.
5. Madore, J ., Saeger , L. A. (1997) . Topology at t he Planck Lengt h gr-qc/ 9708053.
6. Heller, M., Sasin, W . ( 1997) . Int. J. Theor . Phys . 36, 1441.
7. Brans, C. ( 1980) . In Quan tum Theor y an d Gravi tation , A. R. Marlow, ed. ( Academic
Press, New York), p. 27.
8. Rosenbaum, M. ( 1998) . Preprint .
9. L ammerzahl, C., Macias, A. ( 1993) . J. Math. Phys. 34 , 4540; ( 1995) . Addendum,
J. Math. Phys. 36 , 2131.
10. Gr unbaum, A. ( 1977) . In Foundati on s of Space Ti me Theori es , J . Earman, ed.
( University of Minnesot a Press, Minneapolis).
11. Earman, J . ( 1990) . Wor ld Enough an d Space- Ti me : Absolute Vs. Relati on al The-
or i es of Space an d Ti me (MIT Press, Bost on) .
12. Traut man, A. ( 1970) . Rep. Math. Phys. 1 , 29.
13. Hehl, F., Von der Heyde, P., Kerlick, G., Nest er, J . ( 1976) . Rev. Mod. Phys. 4 8,
393.
14. Mielke, E. ( 1987) . Geometrodyn ami cs of Gauge Fi elds (Akademie-Verlag, Berlin) .
15. Hehl, F., McCrea, J ., Mielke, E., Ne eman, Y. (1995) . Phys. Rep. 258 , 1.
16. Howard, D., St achel, J ., eds. ( 1989) . Ei nstei n an d the Hi story of Gen eral Relati vity
( Birkh auser, Bost on) .
17. Nort on, J . ( 1989) . In Ei nstei n an d the Hi story of General Relati vi ty , D. Howard, J .
Stachel, eds. ( Birkh auser, Bost on)
18. Marlow, A. ( 1980) . Int. J. Theor . Phys . 19, 515.
19. Madore, J . (1995) . An Introducti on to Non commutative Di erential Geometry an d
i ts Physi cal Appli cations (Cambridge University Press, Cambridge) .
20. Brans, C. ( 1994) . J. Math. Phys. 3 5, 5494.
21. Asselmeyer, T. ( 1997) . Class. Quan tum Grav. 1 4, 749.