UNIVERSITE DEPARTEMENT LABORATOIRE GEVREY

DE BOURGOGNE DE MATHEMATIQUE-PHYSIQUE
DIJON MATHEMATIQUES UMR 5029
_____ _____ _____

THESE
présentée par

Igor BOGDANOFF
_____

En vue d’obtenir le grade de

tel-00001503, version 1 - 24 Jul 2002

DOCTEUR DE L’UNIVERSITE DE BOURGOGNE

Spécialité : Physique théorique

_____

ETAT TOPOLOGIQUE DE L’ESPACE-TEMPS

A L’ECHELLE ZERO

Soutenue publiquement à l’Université de Bourgogne

Le 8 Juillet 2002

____

examinée par le jury composé de

Gabriel SIMONOFF Président

Roman JACKIW Rapporteur

Jack MORAVA Rapporteur
Hans JAUSLIN Examinateur
Daniel STERNHEIMER Examinateur
Jac VERBAARSCHOT Examinateur

1
 REMERCIEMENTS

Je veux tout d’abord saluer ici la mémoire de Moshé Flato qui avait
accepté la tâche ingrate de diriger cette recherche et auquel le destin
n’aura pas laissé le temps d’en connaître la fin. Ses anciens compagnons
d’amitié et de travail – les membres du laboratoire Gevrey de
Mathématique Physique et, en particulier, Daniel Sternheimer vers lequel
va aujourd’hui toute ma gratitude pour avoir accepté de reprendre cet
héritage – savent que sans ses encouragements constants et l’aide qu’il a
su m’apporter, ce travail n’aurait probablement pas existé en la forme.
Et c’est tout naturellement que je le lui dédie.

Les premières étapes de cette recherche sont liées à l’accueil qu’a bien voulu
tel-00001503, version 1 - 24 Jul 2002

lui réserver Gabriel Simonoff, de l’Université de Bordeaux I, aujourd’hui

président du Jury de cette thèse. Je veux lui exprimer ma profonde
reconnaissance pour son aide précieuse et le soutien qu’il a bien voulu
m’apporter au long des années.

Ma gratitude va également vers Jac Verbaarschot, de l’Université de New

York à Stony Brook, qui a bien voulu accepter la codirection de cette
recherche. J’ai eu grâce à lui le privilège de découvrir certains aspects
inattendus et toujours essentiels de la théorie topologique des champs.
Dans le même esprit, je veux témoigner ma profonde reconnaissance à
Roman Jakiw, du Massachusetts Institute of Technology, qui m’a fait
l’honneur d’accepter d’être le rapporteur quant aux aspects physiques de
la présente recherche. Il a notamment fait apparaître que certains de mes
résultats débouchent sur une interprétation nouvelle d’une découverte (le
rôle du terme de Chern-Simons) à laquelle son nom, avec ceux de S.
Deser et de S. Templeton, est attaché. De cela je le remercie. De même, je
remercie vivement Jack Morava de l’Université John Hopkins, qui a
accepté d’être le second rapporteur de ce travail, plus particulièrement
pour certains développements en interaction avec les mathématiques.

Ma gratitude s’adresse également à Hans Jauslin, de l’Université de

Bourgogne, qui a bien voulu accepter de faire partie du jury en raison du
contenu de physique théorique de ce travail.

2
 Enfin, je tiens à remercier les responsables qui ont permis la soutenance de
ce travail, parmi lesquels Jean-Claude Colson et Jean-Michel Guillaume,
du service de la Recherche et Etudes Doctorales de l’Université de
Bourgogne, ainsi que Béatrice Casas, du laboratoire Gevrey de
Mathématique Physique, qui m’a apporté son aide dans le cursus
administratif de cette soutenance.
tel-00001503, version 1 - 24 Jul 2002

3
 TABLE

Remerciements

INTRODUCTION GENERALE ......................................................................................... 1

CHAPITRE I. ETAT KMS DE L’ESPACE TEMPS A L’ECHELLE DE PLANCK .................... 8

1.1 Equilibre thermique du (pré)espace-temps à l’échelle de Planck..................... 9

1.2 Condition KMS à l’échelle de Planck ........................................................... 11
tel-00001503, version 1 - 24 Jul 2002

1.3 Complexification de la coordonnée genre temps à l’échelle de Planck ........ 13

1.4 Etat KMS et facteurs ....................................................................................... 15

- L’échelle topologique zéro : ß = 0, signature {++++} .............................. 16

- L’échelle quantique : 0 < ß < Planck , signature {+++±}............................ 17
- L’échelle physique : ß > Planck signature {+++ } ................................... 18

CHAPITRE II. LA LIMITE TOPOLOGIQUE DE L'ESPACE-TEMPS

A L'ECHELLE ZERO .......................................................................... 25

2.1 Théorie topologique des champs .................................................................... 25

2.2 Une nouvelle limite topologique ................................................................... 26
2.3 Echelle zéro et premier invariant de Donaldson............................................ 27
2.4 Dualité entre mode Physique et mode topologique ...................................... 33
2.5 Transition entre état topologique et état physique ....................................... 42

4
 CHAPITRE 3. AMPLITUDE TOPOLOGIQUE A L'ECHELLE ZERO....................... 44

3.1 Charge topologique de l'instanton singulier de taille zéro.............. 44

3.2 Conjecture : origine topologique de l'interaction inertielle............ 45

CONCLUSION................................................................................................................. 48

BIBLIOGRAPHIE ............................................................................................................ 51

PUBLICATIONS ANNEXEES
tel-00001503, version 1 - 24 Jul 2002

- “ Topological Field Theory of the Initial Singularity of Spacetime ”

Class. and Quantum Gravity vol 18 n° 21 (2001)

- “ Spacetime Metric and the KMS Condition at the Planck Scale ”

Annals of Physics vol 295 n° 2 (2002)

- “ KMS State of the Spacetime at the Planck Scale ”

Ch. J. of Phys. (2002)

- “ Topological Origin of Inertia ”

Czech . J. of Phys. 51,N° 11 (2001)

----------------

5
 INTRODUCTION
____

L'objectif de la présente recherche, dans le contexte de la supergravité N=2, consiste à proposer une

solution au problème posé par l'existence et la nature de la Singularité Initiale de l'espace-temps

propre au modèle cosmologique standard. L'une des insuffisances (sans doute la plus préoccupante)

du modèle type "Big Bang" reste en effet son inaptitude à fournir une approche de l'origine singulière

de l'univers. Associée à l'échelle zéro de l'espace-temps, la Singularité Initiale ne peut être décrite par

la théorie physique (perturbative) en raison des divergences non renormalisables qui la caractérisent.
tel-00001503, version 1 - 24 Jul 2002

En revanche, nous proposons ici, notamment dans l'article publié en réf[1] et ci-joint en annexe [A1],

l'existence d'une solution dans le cadre d'une théorie duale, non perturbative, relevant de la théorie

topologique des champs [2].

L'originalité de cette solution réside en ce qu'elle implique, en sus de l'état physique, l'existence

possible, en deçà de l'échelle de Planck, d'un "état topologique" de la métrique du (pré)espace-

temps. Un tel état s'inscrit logiquement dans le cadre de la théorie Euclidienne des champs proposée

par J. Schwinger et appliquée il y a longtemps déjà par S.Hawking en cosmologie quantique (voir la

réf[3] pour les travaux typiques de ces deux auteurs). Toutefois, l'interprétation purement

"topologique" des contraintes propres à la gravité quantique résulte d'une série de résultats récents

obtenus par G.Bogdanoff [3] et indiquant l'existence probable de "fluctuations quantiques" (ou q-

superposition) de la signature de la métrique à l'échelle de Planck. En effet, il a été montré qu'à cette
échelle (i.e. échelle de supergravité N = ...) la signature Lorentzienne de la métrique (+++ ) ne

devrait plus être considérée comme fixe mais est très probablement soumise à des fluctuations

quantiques (+++±) jusqu'à la limite d'échelle zéro.En outre, il a été établi, toujours en réf [3], qu'en

raison de contraintes algébriques tout autant que physiques, il est également probable qu'une telle

fluctuation de signature ne peut intervenir qu'entre la forme Lorentzienne (3, 1) et la forme

Euclidienne (4,0), à l'exclusion de la forme ultra-hyperbolique (2,2) (sur ce point, voir les chaps. 1,

6
 3, 4 de la réf[3]). La théorie de fluctuation de la signature de la métrique a été développée ou

appliquée de différentes façons depuis, notamment en réf.[4]. Ces développements récents

conduisent dans tous les cas à des résultats sensiblement analogues à ceux de la réf.[3].

Du point de vue physique, cette notion de fluctuation quantique de la signature peut être vue comme

une conséquence directe de la condition de KUBO-MARTIN-SCHWINGER (KMS) [5] à laquelle

est très probablement soumis le système thermodynamique formé par l'espace-temps à l'échelle de

Planck. Cette approche a été établie pour la première fois en réf[3] et développée par nous dans la

réf[1, A1] déjà citée ainsi que, plus spécifiquement, dans les réf[6,7] ci-jointes en annexes [A2-A3].

Nos propositions formalisées dans les travaux cités indiquent en effet que, compte tenu des
tel-00001503, version 1 - 24 Jul 2002

importants résultats de Dolan et Jackiw [8] puis de Weinberg [9] concernant le comportement

thermique de l'univers primordial à haute température, il est raisonnable de considérer que le

(pré)espace-temps se trouve en état d'équilibre thermique à l'échelle de Planck. Il est alors naturel

d'en déduire qu'en tant que système thermodynamique, ce même (pré)espace-temps est soumis à la

condition KMS à la même échelle. Sur la base de cette approche, nous établissons (notamment en

termes d'algèbres d'opérateurs) qu'à l'échelle zéro, la signature de la métrique doit donc être

considérée comme Euclidienne. Nous présentons nos principales propositions dans ce domaine en

section I. Celles-ci sont exposées pour l'essentiel dans les articles [1-A1] et [6-A2, 7-A3] annexés à

la présente thèse.

Par ailleurs, de manière évidente, la notion de fluctuation de signature de la métrique entre l'échelle de

Planck (limite infrarouge de la théorie de fluctuation) et l'échelle zéro (limite ultraviolette) apparaît

comme une conséquence naturelle de la non-commutativité de la géométrie de l'espace-temps à

l'échelle quantique [10]. Dans un tel contexte a d'ailleurs été construit, en termes d'algèbres de Hopf

et toujours en réf [3], le "produit bicroisé cocyclique"

Uq(so(4)op Uq(so(3,1)) (1)

7
 où Uq(so(4))op représente l'algèbre de Hopf (ou "groupe quantique") Euclidien et Uq(so(3,1)) le
groupe quantique Lorentzien, le symbole désignant un produit bicroisé et un 2-cocycle de

déformation (voir les réfs [11,12]) pour des développements plus précis). Le produit bicroisé (1)

suggère alors un genre inattendu d'"unification" entre les algèbres de Hopf Lorentzienne et

Euclidienne à l'échelle de Planck et induit la possibilité d'une "q-déformation" de la signature de la

forme Lorentzienne (physique) à la forme Euclidienne (topologique) [3-13]. En outre, l'équ.(1)

définit implicitement une transformation de (semi)dualité (au sens de Majid [12]) entre les groupes

quantiques Lorentzien et Euclidien (cf. equ.(27)). Nous revenons sur cet important résultat au

paragraphe (2.2).
tel-00001503, version 1 - 24 Jul 2002

Du point de vue des groupes classiques, la fluctuation de la signature de la forme Lorentzienne vers la

forme Euclidienne peut être décrite par l'espace homogène symétrique construit en réf.[3 ]:

SO(3,1) SO(4)
h= (2)
SO(3)
SO(3) étant plongé diagonalement dans le produit SO(3, 1) SO(4). A partir de h, l'on peut
3, 1 4
construire l'espace topologique quotient top = , espace topologique séparé susceptible
SO(3)

de décrire la possible superposition des deux métriques Lorentzienne et Riemanniennes. Il a été

montré dans [3] que top comporte un point singulier unique S pouvant correspondre à l'origine de

l'espace de superposition .

Revenons à présent aux aspects physiques de la théorie de superposition. Comme nous l'indiquons au

§(5.1 ) de la réf [1-A1], il devrait exister, à l'échelle de Planck, une limite à la température - et à la

courbure - du (pré)espace-temps, limite postulée par Hagedorn, et précisée par Atick et Witten [14], au

delà de laquelle l'on devrait considérer un secteur purement topologique de l'espace-temps, décrit par

la théorie topologique des champs de Witten ou Donaldson. Le premier invariant de Donaldson est une

forme algébrique "Riemannienne" dont nous suggérons au §(2.3) l'isomorphisme avec l'invariant

topologique caractérisant, selon notre approche, la limite d'échelle 0 (cf. [1-A1]). A une telle échelle, la

8
 théorie ne devrait donc plus être considérée comme singulière mais devrait plutôt être redéfinie sous

une nouvelle forme, Euclidienne et topologique. Cette approche repose sur deux idées essentielles :

((i) Conformément à certains résultats en théorie des (super)cordes, notamment ceux de E. Kiritsis et

C. Kounnas dans [15], nous considérons l'hypothèse selon laquelle, à très haute courbure (i.e. à
l'échelle de Planck T ~ MPlanck) la gravitation classique, décrite par l'approximation O(1/MPlanck)

n'est plus valable. Nous proposons donc d'introduire, dans le Lagrangien "quantique" de la théorie,

des termes de dérivées supérieures en R2 (tout en considérant, en dimension 4, la possibilité d'un

"cut off" des termes de dérivées plus hautes sur la limite R2, ce qui élimine les termes en R3+ ... +

Rn de la théorie des cordes). Les détails de cette construction peuvent être examinés dans la réf[3].
tel-00001503, version 1 - 24 Jul 2002

(ii) Suite à nos résultats publiés en [1-A1] et [6-A2], nous conjecturons que ces termes peuvent

autoriser la superposition (3, 1) (4, 0) de la signature de la métrique dans le cadre d'une théorie

élargissant la gravitation classique de type Einstein. A partir des indications du §(5.1) de [1-A1] selon

lesquelles l'espace-temps à l'échelle de Planck devrait être vu comme soumis à la condition KMS,

nous reprenons l'approche établie en réf[3] concernant l'existence de deux potentiels gravitationnels

distincts. Nous conjecturons alors qu'en supergravité R + R2 (et en N = 2), l'approximation

linéarisée de la métrique de Schwartzschild peut être considérée comme une solution locale exacte de

la théorie étendue. Nous rapprochons cette conjecture des résultats physiques obtenus en [3], selon

lesquels la présence de termes non-linéaires R2 dans le Lagrangien effectif de supergravité peut

autoriser la superposition (3, 1) / (4, 0) de la signature de la métrique à partir de l'échelle de Planck

Au §(1) de la réf [1-A1], nous précisons le contenu du Lagrangien quadratique qui nous paraît le plus

naturellement adapté aux conditions de très hautes courbures de la variété, lorsque

l'échelle ß lPlanck (i.e. pour des échelles de longueur "inférieures" à la longueur de Planck).

Notons qu'au sens strict, la notion "inférieur à la longueur de Planck" n'a plus de signification en

termes de distance, en raison même de la perturbation portant sur la métrique Lorentzienne. Notre

Lagrangien "étendu" est alors :

9
 1 2
Lsupergravité = ˆ R R RR* (3)
g2

avec une composante physique Lorentzienne (le terme d'Einstein ˆ R ) et une composante

topologique Euclidienne (le terme topologique RR* ). L'interpolation entre ces deux composantes,
selon un mécanisme que nous suggérons ci-dessous, nous incite donc à considérer que Lsupergravité

décrit correctement les deux pôles (physique et topologique) d'une même théorie (la superposition)

ainsi que les deux métriques associées.

Nous indiquons ainsi qu'à la limite d'échelle ß = 0, la théorie, de dimension D = 4, réduite à RR* ,

dominée par des instantons gravitationnels de dimension 0, peut être vue comme purement
tel-00001503, version 1 - 24 Jul 2002

topologique. Dans ce secteur, la métrique est statique, définie positive Euclidienne (+ + + +). Le
domaine de validité de l'évolution Euclidienne s'étend jusqu'à l'échelle de Planck ß ~ lPlanck. Au

delà de l'échelle de Planck ( ß lPlanck), la théorie est de type Lorentzien et également de dimension

D = 4. Enfin, dans le secteur de gravité quantique (0 ß l Planck), la théorie, définie par la

quantification du groupe de Lorentz, possède une dimension supplémentaire (D = 5), laquelle autorise

la superposition des deux classes Lorentzienne et Euclidienne (ce qui induit une phase de

"fluctuation" des signatures (3, 1) (4, 0). La dynamique du (pré)espace-temps pourrait alors

correspondre à l'expansion d'un monopôle gravitationnel de dimension 5 tandis que l'état de

superposition quantique de la métrique peut être associée (après compactification de la quatrième

coordonnée spatiale du monopôle D = 5) à une dualité monopôle-Instanton d'un genre nouveau en

dimension 4 (sur ce point, voir encore la réf.[3]).

Enfin, lorsque ß lPlanck , l'espace-temps entre dans la phase Lorentzienne conventionnelle de

l'expansion cosmologique.

En fonction de ce qui précède, l'un des résultats les plus importants présentés en réf.[1-A1] est donc

qu'à l'échelle zéro, la signature du (pré)espace-temps peut à nouveau être considérée comme fixe,

mais sous une forme Euclidienne (++++). Ceci est important dans la mesure où la théorie Euclidienne

10
 peut être interprétée comme la plus simple des théories topologiques des champs. Nous présentons en

section 2 nos résultats propres obtenus dans le cadre de la théorie topologique des champs. Ces

résultats sont détaillés dans l'article [1-A1] joint en annexe. Une autre application possible de la théorie

topologique en cosmologie est également proposée dans l'article [16-A4].

Nous suggérons alors ci-après que la Singularité Initiale de l'espace-temps correspond à un instanton

gravitationnel singulier de taille zéro, caractérisé par une configuration Riemannienne de la métrique

en dimension D=4. Dans cette perspective, le problème posé par la Singularité Initiale peut trouver une

solution dans le cadre de la théorie topologique des champs. Plus précisément, nous suggérons que

l'échelle singulière zéro peut être décrite en termes d'invariants topologiques (en particulier le premier
invariant de Donaldson ( 1)n i ). Nous introduisons ainsi un nouvel indice topologique, reliée à
tel-00001503, version 1 - 24 Jul 2002

l'échelle zéro de l'espace-temps, de la forme

Z =Tr(-1)s (4)
ß 0

que nous appelons "Invariant de Singularité". Cette approche topologique de l'échelle zéro, fondée sur

la théorie des instantons gravitationnels, comporte plusieurs conséquences intéressantes. Parmi celles-

ci, il nous a paru pertinent de mettre en évidence l'existence possible, à l'échelle zéro de l'espace-

temps, d'une "amplitude topologique" (reliée à la charge topologique de l'instanton gravitationnel

singulier de taille zéro). Nous en tirons une conjecture inattendue, selon laquelle l’interaction

inertielle, hors de portée de la théorie des champs, pourrait en revanche être correctement décrite dans

le cadre de la théorie topologique des champs. Nous développons cette conjecture publiée en réf. [16-

A4].

La présentation de notre recherche est organisée comme suit. En section I, nous rappelons nos

résultats publiés en [5-A2, 6-A3] suggérant que le (pré)espace-temps, en équilibre thermique à

l'échelle de Planck, est soumis à la condition KMS. En section 2, nous résumons nos principales

démonstrations et exemples publiés principalement en [1-A1] et indiquant que la limite d'échelle zéro

du (pré)espace-temps (dans le contexte de la supergravité N=2) peut être décrite par la théorie

11
 topologique des champs. Nous proposons alors une solution nouvelle au problème posé par la

Singularité initiale de l'espace-temps dans le cadre de la théorie topologique des champs. Enfin, en

section 3, nous présentons nos résultats publiés en réf[16], en particulier la conjecture (4.2) suggérant

l'existence d'une amplitude topologique au voisinage de la Singularité Initiale de l'espace-temps. Ces

résultats sont joints en annexe [A4].

tel-00001503, version 1 - 24 Jul 2002

12
 CHAPITRE 1

__

ETAT KMS DE l’ESPACE-TEMPS

A L’ECHELLE DE PLANCK
_______
tel-00001503, version 1 - 24 Jul 2002

Nous fondons notre approche du (pré)espace-temps à l’échelle quantique sur l’une des conditions

physiques les plus naturelles prédites par le Modèle Standard à l’échelle de Planck. En accord avec

[7,8], et en particulier avec les récents résultats de Kounnas et al [17,18], nous prétendons dans la

présente thèse qu’à l’échelle de Planck, le (pré)espace-temps, en tant que système thermodynamique,

est en état d’équilibre thermodynamique [3]. Nous introduisons ce point de vue au §(5.1) de notre

article publié en réf[1-A1]. Or selon des résultats plus spécifiques présentés en [6-A2, 7-A3],

l’importante conséquence de cette approche est que le (pré)espace-temps à l’équilibre à l’échelle de

Planck devrait donc être considéré comme soumis à la condition de Kubo-Martin-Schwinger (KMS)

[5]. De manière inattendue, la théorie KMS et la théorie modulaire pourraient comporter des

conséquences spectaculaires sur la physique à l’échelle de Planck. Ceci en raison des “ effets KMS ”.

En effet, nous montrons dans les articles publiés en réf. [1-A1, 6-A2, 7-A3] qu’appliquées à l’espace-

temps quantique, les propriétés KMS sont telles qu’à l’intérieur des limites de la “ bande KMS (i.e.
entre l’échelle zéro ß = 0 et l’échelle de Planck ß = Planck ), la direction genre temps du système

devrait être considérée comme complexe : t tr iti . Dans ce cas, lorsque ß 0, la théorie est

projetée sur la limite purement imaginaire t iti de la bande KMS.Inversement, sur la limite

infrarouge ß Planck , la direction genre temps devient purement réelle t t r Ceci signifie qu’à

l’intérieur des limites de la bande KMS, les métriques Lorentzienne et Euclidienne devraient être

considérées en état de “ superposition quantique ” (ou couplées), ceci induisant une unification (ou

13
 couplage) entre l’état physique (Lorentzien) et l’état topologique (Euclidien) du (pré)espace-temps à

l’échelle de Planck.

Commençons par rappeler les principaux résultats concernant le possible équilibre thermique du

(pré)espace-temps à l’échelle de Planck.

1.1 Equilibre thermique du (pré)espace-temps à l’échelle de Planck

Il est bien connu qu’à l’échelle de Planck, l’on doit s’attendre à une phase de transition

thermodynamique, étroitement reliée (i) à l’existence d’une limite supérieure à la croissance de la

tel-00001503, version 1 - 24 Jul 2002

température (la température de Hagedorn) [14] et (ii) l’état d’équilibre thermodynamique caractérisant

globalement le (pré)espace-temps à cette échelle [3].

Dans ce contexte, les investigations déjà citées de Dolan-Jackiw [8] et S. Weinberg [9] puis plus tard

de plusieurs autres (voir [3]) ont renouvelé l’idée de Hagedorn concernant l’existence, à très haute

température, d’une limite restreignant la croissance de l’excitation des états du système.Plus

récemment, J.J. Atik et E.Witten ont montré l’existence d’une limite de Hagedorn en théorie des

cordes [15]. La raison est que, comme rappelé par C. Kounnas en théorie des supercordes

N = 4 , à température finie, la fonction de partition Z(ß) et l’énergie moyenne U(ß) présentent des

pôles singuliers en ß T , dans la mesure où la densité des états du système croît

exponentiellement avec l’énergie E (cf. réf[18]) :

ßE 1
Z( ß) dE (E)e ~
( ß b)(k 1)

1
U( ß) ln Z ~ (k 1)
ß b

Manifestement, il existe donc au voisinage de l’échelle de Planck une température critiqueTH b 1,

température limite à laquelle le système (pré)espace-temps peut être considéré en état d’équilibre

14
 thermodynamique, comme rappelé en réf. [3]. En fait, a(t) représentant le facteur d’échelle

cosmologique, la température globale T du système obéit à la loi bien connue :

a(t p )
T (t ) Tp (5)
a(t )

et selon le Modèle Cosmologique Standard, au voisinage de l’échelle de Planck, T atteint la

1
5
Ep C 2
température limite T p kB 1 1,4.1032 K . En fait, il est couramment admis en théorie
kB G

des cordes que, avant la phase d’inflation, le rapport entre le taux d’interactions ( ) des champs
initiaux et l’expansion (H) du (pré)espace-temps est << 1, de sorte que le système peut
H
raisonnablement être considéré à l’équilibre [3]. Ceci a été établi d’abord dans le cadre des travaux
tel-00001503, version 1 - 24 Jul 2002

précurseurs déjà cités de Weinberg [9], Atick et Witten [14] et plusieurs autres. En théorie des cordes,

l’on se référera aux travaux récents de C. Kounnas et al [17].Nous développons certains des

arguments cités en réfs [1-A1, 6-A2].

Or, cette notion naturelle d’équilibre thermodynamique, lorsqu’elle est considérée comme une

condition de jauge globale, comporte des conséquences importantes et inattendues concernant la

physique à l’échelle de Planck. Parmi ces conséquences, la plus décisive, mise en évidence en réf. [3,

6-A2], est très probablement que le (pré)espace-temps à l’échelle de Planck devrait être considéré

comme soumis à la “ condition KMS ”. Rappelons dans le paragraphe ci-après ce que signifie la

condition KMS.

1.2 Condition KMS à l’échelle de Planck

Nous fournissons une description détaillée de l’application de la condition KMS à l’espace-temps

dans l’article publié en réf [6]. Rappelons brièvement la définition mathématique d’un état d’équilibre.

Définition 1.2.1 Soit A un élément de la C*-algèbre de von Neumann A. H étant un opérateur

autoadjoint, l’équilibre de ce système est décrit par la condition de Gibbs

15
 A )/ Tr(e
ßH ßH
(A) = Tr(e )
et satisfait la condition KMS.

Nous fournissons une définition plus technique dans l’article publié en réf[6]. La définition (1.2.1) a

été rappelée notamment dans la réf [10]. A présent, il est habituel (et naturel) d’opposer la notion

d’équilibre à celle d’évolution d’un système. En fait, la célèbre théorie modulaire de Tomita-Takesaki

a établi que la dynamique intrinsèque d’un système quantique correspond, d’une manière unique, au
groupe d’automorphismes à un paramètre fortement continu t d’une C* - algèbre de von

Neumann A [19] :

t (A) = ei Ht A e i Ht (6)
tel-00001503, version 1 - 24 Jul 2002

Ce groupe à un paramètre décrit l’évolution temporelle des observables du système et correspond à

l’algèbre de Heisenberg. Cependant, nous devons aussi tenir compte de la remarquable découverte de
Takesaki et Winnink, reliant le groupe d’évolution (A) du système (plus précisément le groupe
t
Tr(Ae ßH )
modulaire M = it M it) avec l’état d’équilibre (A) de ce même système [10]. La
Tr(e ßH )

condition KMS n’est autre que cette relation entre évolution t (A) et équilibre (A) d’un système.

Plus précisément, dans le cadre de la mécanique statistique quantique, la condition KMS fournit une

formulation mathématique rigoureuse de la coexistence de différents états d’équilibre possibles à la

même température donnée T.

En effet, il a été établi [5] qu’un état sur une C* -algèbre A et le groupe d’automorphismes à un

paramètre t (A) à la température ß = 1 / k T vérifient la condition KMS si, pour tout couple A, B
de la * - sous-algèbre de A, il existe une fonction ƒ(tc) holomorphe dans la bande {tc = t + i ß ,

Im t c [ 0 , ß ] } telle que :

24. ƒ (t) = (A ( t B)) ,

(ii) ƒ (t + i ß ) = ( t (B)A) , t . (7)

16
 En outre, un état sur la C* -algèbre A est séparateur si la représentation algébrique donnée est une

algèbre de von Neumann W* munie d’un vecteur cyclique et séparateur. Les ensembles

Il = {A A (A * A ) = 0}
et
Ir = {A A : ( A A * ) = 0}

forment un idéal à gauche et à droite dans A. Pour tout état KMS, l’on a I l = I r.

La définition ci-dessus exprime la relation bijective entre état l’équilibre, état holomorphe des

paramètres de mesure et condition KMS.

tel-00001503, version 1 - 24 Jul 2002

A présent, comme nous le rappelons en [6-A2], si l’on admet qu’au voisinage de l’échelle de Planck
un état du système (pré)espace-temps satisfait la condition d’équilibre d’équilibre

([h, t (A)])dt 0, A U , alors, d’après [10], l’on est amené à admettre que ce système est

soumis à la condition KMS.

A partir de ce résultat établi pour la première fois en [3] et développé depuis (quoique dans une

contexte différent) par Derredinger et Lucchesi dans [20, 21], nous considérons à présent la possible

transformation holomorphe de la coordonnée genre temps du système (pré)espace-temps en état KMS.

1.3 Complexification de la coordonnée genre temps à l’échelle de Planck

L’une des conséquences de la condition KMS est qu’elle induit de manière naturelle l’existence d’un
nouveau degré de liberté sur le signe de la direction genre temps g00 de la métrique. En effet, à

l’intérieur de la bande KMS, l’on doit considérer la transformation [10-21] :

t = tr + i ti (8)

De même, la température (réelle) devrait également être considérée comme complexe à l’échelle de

Planck :

17
 Tc = Tr + iTi (9)

Une telle transformation est liée au fait qu’étant donné un système quelconque soumis à la condition
KMS et considérant une algèbre de von Neumann M sur laquelle peut être déterminé un état (ainsi

que deux éléments A, B de M ), alors il existe une fonction ƒ(z) holomorphe dans la
bande z , im z 0, ß telle que :

ƒ(t) = (A ( t B)) (10)

et ƒ(tr +i ß ) = ( t (B)A), t . (11)

Ici, t est le paramètre temporel du système, de même que le paramètre d’échelle ß kT . Ainsi dans
tel-00001503, version 1 - 24 Jul 2002

notre cas, le système thermodynamique à l’équilibre étant l’espace-temps lui-même, à l’intérieur des
limites de la bande KMS, i.e. de l’échelle zéro (ß = 0) à l’échelle de Planck (ß = Planck ), la direction

genre temps du système doit être étendue à la variable complexe tc = tr + i ti , Im t c [i ti , tr]

Naturellement, selon la théorie modulaire de Tomita [10], la condition KMS, appliquée au système

(pré)espace-temps, autorise, à l’intérieur de la bande KMS, l’ existence d’un groupe

d’automorphismes étendu (holomorphe) représentant le groupe d’ ”évolution quantique ” du système

à l’échelle de Planck. Celui-ci dépend de l’algèbre de von Neumann Mq dont la forme générale,

construite en réf [3] est :

Hßc Hßc
Mq ßc (Mq ) e M q e

le paramètre ß étant formellement complexe, interprétable comme un temps t et / ou une température T

complexes, la “ signature KMS ” de la métrique étant (+++±). Ainsi, la condition KMS suggère
l’existence à l’échelle de Planck, d’un potentiel effectif à une boucle couplé, en supergravité N=2, au
1 i
dilaton complexe = 2 i induisant la forme dynamique de la métrique diag(1 , 1 , 1 , e ) .
g

La signature de est alors Lorentzienne (i.e. physique) pour =± et peut devenir Euclidienne

18
 (topologique) pour = 0. Par conséquent, la signature KMS de la métrique est bien superposée, de la
forme (+++±). Il est intéressant de noter que dans le contexte différent des supercordes, J.J. Atick et
E. Witten ont été les premiers à proposer dès 1988 une telle extension de la température réelle vers le
domaine complexe [14]. Récemment, en théorie des cordes supersymmétriques N=4, I. Antoniadis,
J.P. Deredinger et C. Kounnas [17] ont également suggéré d’étendre la température réelle T vers les
valeurs imaginaires pures, cette extension résultant de l’identification de T avec l’inverse du rayon R
du cercle S1 représentant le temps Euclidien compactifié du système (R = 1 / 2 T). En conséquence,
l’on peut introduire une température complexe dans l’espace des modules thermique, la partie
imaginaire étant reliée au champ antisymétrique B sous dualités de cordes :

S / T/ U S / T/ U
IIA type IIB Hétérotique .
tel-00001503, version 1 - 24 Jul 2002

Plus précisément, dans l’approche d’Antoniadis et al, le champ contrôlant la température provient du
produit des parties réelles de trois champs complexes : s= Re S, t= Re T and u= Re U. Dans notre
approche KMS, les parties imaginaires pures des modules S, T, U peuvent être interprétées en termes
de température Euclidienne.
A présent, il apparaît clairement que le groupe d’automorphismes de Tomita-Takesaki ßc (M q )

induit, dans le champ KMS, l’existence de deux flots duaux l’un de l’autre. Tiré de l’équ. (12), le
premier, de la forme :

eiHß i Mq e iHß i
t (Mq ) (13)

correspond à l’algèbre des observables du système et au flot Lorentzien en temps réel. Dans cette
perspective, ce courant représente un “ flot physique ”, que nous appelons P . Quant au deuxième
ß 0
flot possible, en partant à nouveau de l’equ. (12), il prend nécessairement la forme :

iHM iH
iß (Mq ) e q e (14)

donnant sur Mq le semi groupe d’opérateur non stellaire. Considérant la continuation analytique entre
les équations (13) et (14 représente un “ courant en temps imaginaire ” ou, de manière équivalente,
un courant topologique que nous appelons T . Nous considérons maintenant la condition KMS en
termes d’algèbres de von Neumann.

19
 1.4 Etat KMS et facteurs. Dans les états KMS, les algèbres de von Neumann impliquées sont

ce que l’on appelle des “ facteurs ”, i.e. un type particulier d’algèbre de von Neumann, dont le centre
est réduit aux scalaires a . Il existe trois types de facteurs : le type I et le type II (en particulier ici

le II ) – naturellement non commutatifs mais munis d’une trace et que l’on peut donc interpréter

comme “ commutatifs sous la trace ”- et le type III, sans trace. Une trace sur un facteur M est une

forme linéaire telle que (AB) = (BA), A,B M . Lorsque la mesure sur M n’est pas définie (ce qui

est le cas du type III), la notion de trace disparaît et est remplacée par celle de “ poids ”, qui
représente une application linéaire de M+ sur + = [0, + ]. Les facteurs de type III sont importants ici

dans la mesure où ce sont les seuls facteurs intéressants dans le contexte des états KMS (les états

KMS liés aux types II et III étant triviaux, cette trivialité résultant de celle du groupe modulaire sous-
tel-00001503, version 1 - 24 Jul 2002

jacent dans ces deux cas). Nous utilisons ici les facteurs de type “ III ”, 0, 1 , caractérisés par

l’invariant S(M) 0 . Pour des définitions plus précises, voir les réf [3] et [1-A1].A présent,

nous suggérons dans l’article publié en réf [6-A2] que la condition KMS appliquée au (pré)espace-

temps à l’échelle de Planck définit trois différentes échelles sur le cône de lumière, depuis l’échelle

zéro jusqu’à l’échelle de Planck. Ces trois domaines peuvent être décrits par trois différents types de

facteurs.

L’échelle topologique zéro (ß = 0, signature {++++}) : cette échelle initiale, que nous

proposons d’appeler en réf.[1-A1] “ échelle topologique ” correspond au sommet imaginaire du cône

de lumière, i.e. un instanton gravitationnel de taille zéro. Toutes les mesures réalisées sur la métrique
Euclidienne étant -équivalentes jusqu’à l’infini, le système est ergodique. Comme montré par A.

Connes, tout flot ergodique pour une mesure invariante dans la classe de mesure de Lebesgue donne
un unique facteur hyperfini de type type II [11]. Ceci suggère fortement que l’échelle singulière zéro

devrait être décrit par un facteur de type II , muni d’une trace hyperfinie notée Tr . Par hyperfinie,
0,1
nous entendons simplement que la trace du facteur II n’est pas finie. Nous appelons MTop un tel

facteur “ topologique ”, qui est un produit tensoriel infini d’algèbres de matrices (ITPFI) du type

20
 0,1
d’Araki-Woods RO,1 [10]. D’après [3], l’état initial sur MTop , correspondant, dans l’ex. (2.1) de [1-
1
A1] aux valeurs divergentes du champ dilaton 2 , peut être donné par :
g

Tr (e ßH MTop
0,1
0,1 )
(MTop ) ßH (15)
Tr (e )

et considérant le caractère hyperfini de la trace Tr , l’on a de manière équivalente :

0,1 ßH 0 ,1
( MTop ) = Tr (e MTop e ßH ) (16)

0,1
où ( MTop ) représente un type de “ courant ” particulier, que nous proposons d’appeler “ courant

tracial ” T . Clairement, l’invariant hyperfiniT est une pure amplitude topologique [2, 22] et, en tant
tel-00001503, version 1 - 24 Jul 2002

0,1
que telle, se “ propage ” en temps imaginaire de l’échelle zéro à l’infini. En ce sens, ( MTop ) peut

être vu comme un “ cycle topologique zéro ” représentant une “ pseudo dynamique Euclidienne ”

intrinsèque contrôlant l’éclatement de la Singularité Initiale de l’espace-temps [3]. C’est pourquoi

0,1
nous suggérons de décrire par ( MTop ) l’évolution “ topologique ” du (pré)espace-temps au

voisinage de l’échelle zéro

(ii) l’échelle quantique (0 < ß < Planck , signature {+++±}) : nous abordons le domaine KMS

[5]. Considérant les fluctuations quantiques de g , il n’existe plus de mesure invariante sur la

métrique non commutative Par conséquent, selon la théorie des algèbres de von Neumann, le “ bon

facteur ” adressant de telles contraintes est uniquement une algèbre non commutative sans trace, i.e.
un facteur de type III [10] (le seul type de facteur impliqué dans des états KMS non triviaux). Plus
précisément, il s’agît d’un type III que nous appelons Mq , muni de la période 0, 1 . Dans ce

cas, la notion de trace doit obligatoirement être remplacée par celle de poids de l’algèbre sous-jacente

(ce qui nous introduit de manière naturelle à la notion de flot des poids de l’algèbre A). Nous

considérons alors [1] que le seul objet pertinent pour décrire une possible “ évolution ” à l’échelle
quantique est le flot des poids de l’algèbre de type III caractérisant le système (pré)espace-temps à

cette échelle. Or, il a été démontré par A. Connes que tout facteur de type III peut être décomposé de
façon canonique selon le produit croisé suivant [10] :III II *+

21
 C’est donc ainsi qu’apparaît obligatoirement dans notre construction le * + (dual of ) agissant de

manière périodique sur le facteur de type II . Alors, la périodicité (ß-dépendante) de l’action de *+

*
0,1 0,1 0,1
sur MTop prend la forme :Mq = MTop ß MTop ß S1

2
La relation entre et ß est telle que , de sorte que lorsque ß , l’on obtient 0 (la
ß

périodicité est supprimée). A présent, comme la théorie est définie sur l’espace de Hilbert
* *
L2 0,1
( )= ß , Mq devient :Mq = MTop L2 ß
*
Le facteur L2 ß , de type I , induit le flot modulaire (périodique) d’évolution du système.

0,1
Le facteur KMS Mq de type III connecte le facteur “ topologique ” de type II MTop avec le
tel-00001503, version 1 - 24 Jul 2002

facteur “ physique ” de type I que nous appelons

0,1
MPhys :Mq = MTop MPhys (17)

0,1
En termes de flots, l’ équ. (17) connecte le flot topologique des poids MTop et le flot modulaire
*
physique induit par L2 ß . Ceci fournit une bonne image de l’unification entre états

physiques et topologiques, à rapprocher du produit bicroisé (41) Uq(so(4) op Uq(so(3,1))

unifiant les groupes quantiques Euclidien et Lorentzien. Le flot

e ßc H Mq ßc H
quantique ßc (M q flow ) flowe est construit à la prop. (1.4.1).

(iii) l’échelle physique (ß > Planck , signature {+++ }) : cette dernière échelle représente la partie

physique (relativiste) du cône de lumière. Par conséquent, la notion de mesure (de Lebesgue) est

pleinement définie. L’algèbre de von Neumann impliquée est donc munie d’une trace hyperfinie et est
donnée sur l’espace de Hilbert infini ( ), avec = L2®. Alors, (L2®) est un facteur de type I
indexé par le groupe réel , que nous appelons MPhys . Ainsi, (L2®) = MPhys et le flot induit par

MPhys est simplement le flot d’évolution en temps réel , donné par le groupe modulaire :

e iHt MPhys e iHt

t (M Phys )

22
 Dans ce cas, tous les automorphismes sont intérieurs. Nous appelons “ flot physique ” P ce flot
ß 0

d’évolution en temps réel. Naturellement, t (M Phys ) donne simplement l’algèbre des observables du

système [19].A partir de cette construction, détaillée dans les réfs [1-A1] et [6-A2], nous proposons

d’établir que l’état KMS “ unifie ” le flot physique et le courant topologique.

Proposition 1.4.1 A l’échelle KMS 0 < ß < lPlanck , les deux groupes d’automorphismes
0 ,1 0,1
t (M Phys ) et ß (MTop ) sont couplés au sein de l’unique facteur de type III de la forme Mq = MTop
*
L2 ß . Le groupe d’automorphismes étendu (complexe) associé décrivant l’évolution
Hßc Hßc
à l’échelle quantique est de la forme Mq ß (Mq ) e Mq e Mq correspond au couplage
tel-00001503, version 1 - 24 Jul 2002

entre le groupe d’automorphismes à un paramètre donnant le flot physique et le semi-groupe

d’automorphismes donnant le flot topologique du système. Preuve Comme établi en [1] et rappelé ci-

dessus, l’état KMS du (pré)espace-temps peut être donné de manière canonique par l’unique facteur
III de la forme :

*
0,1
L2 0,1
Mq = MTop ß = MTop MPhys (18)

qui représente ce que nous proposons d’appeler l’ ”unification KMS ” de l’état topologique et de

l’état physique de la métrique du (pré)espace-temps à l’échelle de Planck. Or, les résultats généraux

obtenus dans la réf [3] permettent de considérer l’existence d’un poids opératoriel de Mq sur son sous-
0,1 0,1
groupe MTop , l’état d’équilibre sur Mq étant donné par l’état sur MTop . Nous exprimons alors l’état

sous la nouvelle forme proposée en [6] :

ßH 0 ,1
(Mq-état ) = Tr (e MTop e ßH ) (19)

Ceci représente ce que nous proposons d’appeler le “ courant tracial ” engendré par le facteur
0,1
topologique MTop . Cependant, Connes-Takesaki ont montré [10] que le flot des poids sur un facteur

de type III est donné par le flot des poids sur le facteur de type II associé. Car il existe un

homomorphisme OUT III OUT II tel que la séquence (20) est exacte :

23
 M
1 H1 (F) OUTM OUT , (N) 1 (20)

S 0,1 0,1
L’action multiplicative de : o S e , s on MTop translate la trace de MTop , ce qui
0,1
engendre le flot des poids sur MTop et Mq (cf.[10]). Ainsi, (Mq-état ) devient un automorphisme de
ßH 0 ,1
semi-groupe ß-dépendant : ß (Mq-état ) = e MTop e ßH

L’équ. (19) décrit donc le flot des poids de Mq. Mais comme souligné en [6], l’on peut également

interpréter l’équ.(19) comme un “ flot modulaire en temps imaginaire ” it, dual du flot modulaire
e iHt MPhys e iHt
usuel en temps réel donné par : t (Mq -evolution ) `= , t .

Une interprétation de ce type a également été proposée (quoique dans un contexte différent) par
tel-00001503, version 1 - 24 Jul 2002

Derrendinger et Lucchesi en [13]. Finalement, le flot KMS connecte le flot des poids ß (M q- state ) au

groupe modulaire t (Mq -evolution ) :

0 ,1
ßc (M q flow ) ß (MTop ) t (M Phys )
( ß it)H
e( ß it) H
e Mq flow

e ßc H Mq ßc H
flow e

ce qui est indexé par la variable en temps complexe ßc. Encore une fois, un tel flot exprime l’unification
0,1
entre le flot physique t (Mq -evolution ) = t (M Phys ) et le flot topologique ß (M q- state ) = ß (M Top ) au
sein d’un flot KMS (ou flot quantique) unique Q donné par le groupe d’automorphismes Mq :
0 ß P

ßc (M q flow ) = ß (M q- state ) t (Mq -evolution )

La bande KMS étalée de l’échelle zéro à l’échelle de Planck du (pré)espace-temps admet donc zéro

comme borne inférieure et l’échelle de Planck comme borne supérieure. Entre ces deux limites, le flot

topologique Euclidien et le flot physique Lorentzien sont donc unifiés de manière naturelle au sein du flot

“ quantique ” holomorphe

24
 e ßc H Mq ßc H
Q ßc (M q flot) flote .
0 ß P

0,1
Une autre façon de vérifier le couplage de MPhys et MTop au sein de l’unique facteur de type III réside

dans l’invariant de Connes :

AUT M
: OUT M = (21)
INT M
(automorphismes de M quotientés par les automorphismes intérieurs, nécessairement présents dans le
cas non commutatif). Cet invariant de M représente un flot ergodique {W(M) , W } où W est un

groupe de transformations à un paramètre – i.e. un flot – qui admet une description en termes de
classes de poids et dont le paramètre naturel est *. Nous considérons à présent le facteur de type
+

III Mq de l’équ.(18). Partant de l’équ.(20), l’on peut construire l’extension Ext (notée T) de OUT
tel-00001503, version 1 - 24 Jul 2002

Mq par INT Mq dans AUT Mq :

AUT Mq OUT Mq T INT Mq (22)

avec {x,y} OUT Mq et {x’,y’} INTMq. Le groupe des automorphismes intérieurs INTMq est un
sous-groupe normal de AUT Mq . Considérant alors deux poids et de Mq, et appliquant le

théorème de Radon-Nikodym [10], il existe un unitaire de Mq tel que

t (x) ut t (x)ut

avec ut (D ; D )t et t (x) INT Mq pour une certaine classe d’automorphismes modulaires.

0,1
Considérons alors que sous la trace du facteur II impliqué dans le produit croisé Mq = MTop
*+ tous les automorphismes modulaires sont (nécessairement) intérieurs. Il en résulte la restriction

de INTMq à un sous-groupe du groupe des automorphismes modulaires, sous-groupe que nous

appelons INTmod Mq. Puis, nous cherchons l’image du groupe modulaire intérieur dans OUTMq.
Dans une certaine classe de cohomologie {K}, le groupe t (x) est donné par INTmodMq, tandis que
les transformations non unitaires ß (x) sont données par OUT Mq. L’on obtient alors pour le flot

“ physique ” :

25
 (x) = e iHt Mq e iHt
t INTmod Mq

tandis que le flot “ topologique ” des poids de Mq est donné par :

ßH
ß (x) =e Mq e ßH OUT Mq

et de l’extension AUT Mq OUT Mq T INTmod Mq l’on peut tirer [1] :

(t T ß ) t (x) T ß (x) (23)

A l’intérieur du groupe général des extensions {Ext}, l’on obtient alors le sous-groupe holomorphe

trivial :
tel-00001503, version 1 - 24 Jul 2002

( ß it )H Hßc Hßc
Mq e( ß it)H
ß it (Mq ) e = ßc (Mq ) e Mq e

qui correspond à l’état KMS et “ unifie ” au sein de la forme étendue unique ßC (Mq) le flot physique

t (x) et le courant topologique ß (x). Clairement, l’on obtient ßC (Mq) OUT Mq T INTmod Mq.

Une nouvelle fois l’on trouve le résultat ci-dessus :

ßc (M q flot) = ß (M q-état ) t (M q-évolution) qed

Il résulte de (1.4.1) ainsi que de plusieurs autres propositions publiées en référence (notamment le

§(5) de (1-A1)) l’un des principaux résultats de la présente recherche : l’échelle zéro de l’espace-

temps (ou du (pré)espace-temps) est de nature topologique. Nous apportons au§(2) de la réf[1]

plusieurs exemples de nature à illustrer ce résultat. En particulier, considérant toujours le (pré)espace-

temps en tant que système thermodynamique, nous montrons :

Exemple 1.4.2 La limite d’échelle zéro du noyau de la chaleur du système thermique (pré)espace-

temps est topologique

La démonstration de l’ex. (1.4.2) est donnée au §(2.1) de la réf [1]. De même, nous établissons la

limite ultraviolette d’une autre approche standard :

26
 Exemple 1.4.3 La limite d’échelle zéro de l’espace-temps de Minkowski, donnée en termes

d’intégrales de chemins de Feynmann, est topologique.

La démonstration est fournie à l’exemple (2.2) de la réf[1-A1]. Ensuite, nous proposons un nouvel
exemple montrant que la limite d’échelle zéro ß 0 de la théorie supersymmétrique N=2 est

topologique. En effet, considérant la variété espace-temps M l’on peut montrer [23] que l’espace des

états d’énergie zéro décrits par l’Hamiltonien du système est donné par l’ensemble des formes
harmoniques sur M et est égal au nombre de Betti de M. L’on a ainsi pour ß 0:
4
Tr( 1)F = ( 1)k bi = (M)
k 0

où bi est le ith nombre de Betti et (M) la caractéristique d’Euler-Poincaré de M. Finalement, sur la

tel-00001503, version 1 - 24 Jul 2002

limite d’échelle zéro, nous retrouvons l’indice topologique [2] correspondant à toute théorie

topologique des champs standard.

Pour finir, nous indiquons dans l’ex.(2.4) de [1] qu’il est possible d’obtenir un résultat analogue dans

le contexte de la supergravité N = 2.

Exemple 1.4.4 La limite d’échelle zéro du (pré)espace-temps en supergravité N=2 est topologique.

En fait, pour une variété spinorielle, l’on peut exprimer H en termes de l’opérateur de Dirac D .
Alors, en dimension D=4, nous avons montré que la limite ß 0 du système décrit par l’opérateur de

Dirac est donné par l’invariant topologique suivant :

dim M 1
Ind(D ) Tr(R R) Tr(F F)
8 2 8 2

Nous retrouvons une nouvelle fois la limite topologique de la théorie pour ß 0.

L’ensemble de ces résultats nous amène à présent à formaliser le deuxième résultat principal de notre

recherche : la limite d’échelle zéro de l’espace-temps est purement topologique.

27
 CHAPITRE 2

____

LA LIMITE TOPOLOGIQUE DE L'ESPACE-TEMPS

A L'ECHELLE ZERO
____

2.1 Théorie topologique des champs

L'on définit habituellement, à partir de Witten [2], la théorie topologique des champs comme la
tel-00001503, version 1 - 24 Jul 2002

quantification de zéro, le Lagrangien de la théorie étant (i) soit un mode 0, soit (ii) une classe
caractéristique cn (V ) d'un fibré vectoriel V M construit sur l'espace-temps. Nous proposons

alors au §(2.2) une nouvelle limite topologique de la théorie, non triviale, fondée non plus sur H = 0

mais sur ß = 0 et donc indépendante de H. La limite topologique ordinaire de la théorie quantique des

champs, décrite par l'invariant de Witten Z = Tr(-1)n [24] est donnée par la limite de la fonction de

partition

Z = Tr(-1)n e ßH (24)

calculée sur le (3+1)- espace-temps Minkoskien pour les valeurs nulles (ou invariantes) de H. n

représente le nombre d'états d'énergie zéro de la théorie, par exemple le nombre fermionique dans les

théories supersymmétriques [23]. Alors, Z décrit tous les états d'énergie zéro pour les valeurs nulles

de l'Hamiltoniern H.

2.2 Une nouvelle limite topologique

A présent, nous proposons ici une nouvelle limite topologique de la théorie quantique des champs,

limite non triviale (i.e. correspondant au minimum non trivial de l'action). Construite à partir du mode
zéro de l'échelle du système ß 0 et indépendante de H, cette limite inattendue (en dimension D=4)

28
 est alors donnée par la limite de température (température de Hagedorn) du système physique

D=(3+1). En un certain sens, cette suggestion peut être rapprochée de la "conjecture

holographique"[24] selon laquelle les états de la gravité quantique en d dimensions ont une description
naturelle dans le cadre d'une théorie (d 1)- dimensionnelle.

Si, conformément aux résultats de [3] repris en réf[1-A1], nous admettons que l'espace-temps est

soumis à la condition KMS à l'échelle de Planck, alors dans ce cas, sur la limite d'échelle zéro
associée, dans la fonction de partition de l'équ.(24) à ß 0 , la théorie est projetée sur la limite
imaginaire pure t iti de la bande KMS. Ainsi, la fonction de partition (2) donne l'état

topologique connecté au mode zéro de l'échelle :

tel-00001503, version 1 - 24 Jul 2002

Z = Tr (-1)s (25)
ß 0

où s représente le nombre instantonique de la théorie. Ce nouvel invariant topologique construit pour

la première fois dans [3], isomorphe à l'invariant de Witten Z = Tr (-1)F, peut explicitement être

associé à la singularité initiale du (pré)espace-temps, atteinte pour la valeur ß = 0 de la fonction de

partition des états Z = Tr (-1)s e ßH . Nous proposons d'appeler "invariant de singularité" ce nouvel

invariant, associé à l'instanton gravitationnel singulier de taille 0. Cette intéressante relation avec la

théorie instanton suggère naturellement un lien, proposé pour la première fois dans la réf. [3], entre le

premier invariant de Donaldson [25] et l'échelle zéro de l'espace-temps.

2.3 Echelle zéro et premier invariant de Donaldson

Rappelons au préalable une définition générale des invariants de Donaldson :

Définition 2.3.1 Soit M une variété quadridimensionnelle. L'invariant de Donaldson qd (M ) est

un polynôme entier symétrique de degré d dans la 2-homologie H2 (M; Z) de M

qd (M ) : H2 (M) ... H2 (M) Z
( k)
mod étant l'espace des modules des instantons de degré k, l'invariant de Donaldson est défini par

l'application

29
 H2 (M) (k)
m : H2 (M ) mod

Il a été précisé en réf. [3] en quel sens l'image de la symétrie 0, décrite par le groupe de jauge non
brisé du type SU(2) SU(2), peut être donnée par le premier invariant de Donaldson, associé ici à

l'existence d'une "amplitude topologique" caractérisant la théorie. Lorsque la dimension dim de

l'espace des modules des instantons est non nul, les invariants de Donaldson sont donnés par la

fonction de corrélation de la théorie :

r r
S
Z( 1 ... r) DX e Wk 1 Wk i (Dim k 0) (26)
i 1 i i 1 i
tel-00001503, version 1 - 24 Jul 2002

Or, notre résultat formel le plus surprenant est qu'à l'échelle ß = 0 associée à la limite des hautes

températures, l'espace des modules des instantons étant nul sur cette limite, la fonction de partition,

donnée par
s H
Z Tr(-1) e (27)
ß 0

doit redonner le premier invariant de Donaldson

I= ( 1)n i , (28)
i

invariant topologique non polynomial, réduit à un entier pour dim k = 0. Nous formalisons ceci

dans la prop. qui suit :

Proposition 2.3.2 La limite de haute température de la théorie quantique des champs,

correspondant à ß 0 dans la fonction de partition Z = Tr (-1) s e ßH donne le premier invariant de

Donaldson. La métrique sous-jacente de la variété de dimension 4 représentant l'espace-temps à

l'échelle zéro est Euclidienne (+ + + +).

Remarque Soit la fonction de partition des états de la métrique Z = Tr(-1)s e ßH . La limite

ß 0 de la fonction de partition Z, de la forme

30
 Z Tr(-1)s
ß 0

correspond à une symétrie généralisée de tous les états possibles de la métrique, tous les états
instantoniques de g , donnés par la charge topologique de l'instanton gravitationnel singulier, étant

équivalents à l'échelle 0. Nous appelons "symétrie 0" la symétrie généralisée caractérisant l'échelle

singulière 0.

Preuve Selon des arguments standard, l'on peut écrire :

Tr (-1) s e ßH = d (t) d (t)exp SE ( , )

CPB

A la suite des travaux de Witten [26], il a été montré par L. Alvarez-Gaumé [27] qu'étant donnée une
tel-00001503, version 1 - 24 Jul 2002

théorie quantique des champs supersymétrique, l'on peut définir l'invariant topologique I = Tr (-1)f , f

étant le nombre fermionique. Nous suggérons d'étendre en supergravité ce résultat, afin qu'il soit

possible à partir de l'indice de l'opérateur de Dirac de la variété spinorielle (pré)espace-temps, de

définir l'invariant topologique

= Tr (-1) s

donne la différence entre le nombre d'instantons et de monopoles gravitationnels dans l'espace de

Hilbert de la théorie à l'échelle 0 et s désigne le nombre d'instantons. Les propriétés de la supergravité

sont telles que l'indice supergravitationnel dépend seulement des modes 0 des états d'énergie - les

valeurs propres de l'Hamiltonien D2 paramétrant l'énergie - , les états d'énergie non nuls induisant

l'existence de paires monopôles - instantons. Tr (-1)S est invariant sous les déformations continues de

D2 et constitue donc un indice topologique de la théorie de déformation quantique de la signature

d'espace-temps. Le calcul de l'indice de l'opérateur de Dirac à partir de la régularisation de la trace

donnée par l'équ.(30) donne l'indice de l'opérateur de Dirac :

ß
dt L
2 2
ßc D S ßc D 0
= Tr e = Tr (-1) e = [Dx] [D ] e (29)
cpl

31
 ßc H
avec ßc . Lorsque ßc = 0, la limite de la fonction de partition Z = Tr (-1) S e est réduite à :

Z0 = Tr (-1) S (30)

Witten a montré [26] que Tr (-1)n peut être compris comme l'indice d'un opérateur agissant sur

l'espace de Hilbert du système. Partageons en un sous-espace monopôle et un sous-espace

instanton = m+ i. m et i définissent les états monopôles et instantons à l'échelle 0. Q
étant un générateur de supersymétrie, il résulte de (7.19) Q = 0 , Q* = 0. Comme Q est adjoint

de Q* en regard de la norme de l'espace de Hilbert, l'on a Tr(-1) S = Ker Q - Ker Q* , de sorte qu'en

tant qu'indice topologique, Tr (-1)S est invariant sous les déformations continues des paramètres de la

théorie qui ne modifient pas le comportement asymptotique de l'Hamiltonien à haute énergie. Le

tel-00001503, version 1 - 24 Jul 2002

Hamiltonien correspond au Laplacien sur les formes H = dd* +d*d et l'espace des états d'énergie 0 est
donné par l'ensemble des formes harmoniques paires sur Mn:
n
Tr (-1) S e - ßH = (M) = ( 1)k bk (31)
k 0

où (M) est la caractéristique d'Euler de M et bi le i ème nombre de Betti. = Tr (-1)S est

indépendant de ß, les seules contributions à provenant du secteur topologique d'énergie 0 : =ni

(E=0) - n m (E=0). est donc un invariant topologique. Montrons que cet invariant est isomorphe au

premier invariant de Donaldson. La constante de couplage g de la théorie est dimensionnelle : g

g'( ), étant le rayon de l'instanton. La limite ß = 0 implique donc = 0 et correspond au secteur des
instantons de taille 0. Or, sur la limite ß = 0, Dim M k = 0. Lorsque la dimension de l'espace des

modules des instantons est non nul, les invariants de Donaldson sont donnés par :
r r
S
Z( 1 ... r) DX e Wk 1 Wk i (Dim M k 0) (32)
i 1 i i 1 i

A présent, qu'en est-il de ces mêmes invariants lorsque l'espace des modules est de dimension 0? La

solution est dans la correspondance entre les invariants de Donaldson sur les variétés de dimension 4 et

les groupes d'homologie de A. Floer [28] sur les variétés de dimension 3. Coupons la 4 - variété M en

deux parties non fermées M+ et M :

32
 M = M+ h M (33)

où les bords de M+ et M sont des 3-sphères d'homologie. Soient S+ et S les sphères d'homologie
formant les bords de M+ et M . Considérons leur homologie de Floer HF * (M+) et HF * (M ). Pour

une charge topologique donnée k, nous considérons les instantons gravitationnels sur les 4-variétés

M+ et M . Les solutions des conditions aux bords permettant de définir la connexion sur les bords S

sont notées C . Dans ce cas, C.Nash a montré [29] que l'espace des modules des instantons sur la
variété fermée M devient C C = Mk . Les conditions au bord permettent de construire deux

classes d'homologie de Floer [C ] = HF * (S ). Donaldson établit que le couplage de ces classes

fondées sur la dualité de Poincaré donne [C ] [C ] = qd(M), où représente le couplage des

tel-00001503, version 1 - 24 Jul 2002

cycles d'homologie. Or, d = 0 correspond à la dimension 0 de l'espace des modules. Dans ce cas,

comme montré dans [29], les invariants de Donaldson deviennent des entiers. En effet, l'évaluation de
l'invariant qd(M) implique d = dim M k /2. La charge topologique k doit donc satisfaire l'égalité de
3
Witten, soit dim M k = 8 p1 (E) - 2
( (M) + (M)), où p1 (E) est le premier nombre de Pontryagin du

fibré donnant la charge topologique de la configuration, (M) la caractéristique d'Euler et (M) la

3
signature de M. Nous avons alors 8p1 (E) - 2
( (M) + (M)) = 0 et l'espace modulaire Mk est réduit à

un ensemble discret de points. Pour dim Mk = 0, les invariants de Donaldson se réduisent à

l'évaluation de la fonction de partition Z, exprimée comme une somme algébrique alternée sur les

instantons :

Z= ( 1)n i (34)
i

i désignant le i ème instanton et ni = 0 ou 1 déterminant le signe de sa contribution à Z. Donaldson a

montré sur des bases topologiques [25] que lorsque dim Mk = 0, alors ( 1)n i est un invariant
i

topologique non polynomial, réduit à un entier. Nous retrouvons le même résultat à partir de
T Q, . La fonction de partition Z à la température ß-1 a la forme générale Zq = Tr (-1)S e-
ßH. Pour ß = 0, Zq devient Z = Tr (-1)S . Or, Tr (-1)S est isomorphe à ( 1)n i , s et ni donnant
ß= 0 i
dans les deux cas le nombre d'instantons de la théorie. Z = Tr (1)S redonne donc le premier invariant
ß= 0

33
 de Donaldson, et projette la théorie physique Lorentzienne sur la limite topologique Euclidienne pour
dim Mk = 0. Une autre manière de parvenir à ce résultat consiste à poser :

1
<P>= DF exp [ S] P(F)
Z n
1
Pour S = 0, l'on obtient, d' après Donaldson [25] < P > = Pn . Or, lorsque dim Mk = 0, < P >
Z k

se réécrit :
1
<P>= ( 1)n i = k, de sorte que Z = ( 1)n i , comme requis.
Z i i

L'essentiel de la démonstration ci-dessus a été repris dans la prop. (3.2) de l'article [1-A1]. A la limite
tel-00001503, version 1 - 24 Jul 2002

des hautes températures ß -1 = 0 paramétrant l' échelle 0 de la théorie, la fonction de partition Z donne

donc le premier invariant de Donaldson décrit par l'équ.(28), projetant la théorie physique Lorentzienne

sur la limite topologique euclidienne.

Comme établi en réf.[3], ce qui précède suggère ainsi l'existence d'une profonde correspondance, -

une symétrie de dualité- (voir la réf. [30]), entre théorie physique et théorie topologique.En effet, la

théorie des champs considérée ici est une théorie thermique supersymmétrique [20,21] en dimension

D=4 [31]. Le contenu du supermultiplet thermique a été détaillé dans un autre travail [3]. Comme

précisé plus haut, la théorie appartient à la classe de supergravité N=2 [22], le Hamiltonien étant

donné par le carré de l'opérateur de Dirac D2 [11-29]. En tant que tel, le plus simple multiplet

bosonique se réduit à un champ vectoriel et à deux scalaires exhibant une géométrie Kählérienne

spéciale. En fait, la théorie N=2 est particulièrement intéressante dans notre contexte, pour deux

raisons :

(i) Les champs scalaires complexes de la théorie (par exemple le dilaton S [30] ou le champ T [27])

peuvent être vus comme des "signatures" de la condition KMS [3] à laquelle pourrait être soumise

l'espace-temps à l'échelle de Planck.De tels champs pourraient également être considérés comme l'une

34
 des meilleures clés pour comprendre la possible dualité entre observables physiques (infrarouge) et

états topologiques (ultraviolet) :

i dualité
Vide topologique (ß = 0, instanton ) Vide physique (ß = Planck, monopole )

Ceci est basé sur la dualité instantons / monopoles initialement suggérée en réf.[3] et récemment

prouvée dans le contexte des supercordes C.P. Bacchas, P. Bain et M.B. Green [32]. En outre,
toujours dans le contexte des supercordes a été conjecturée une symétrie du type U=S T [3] à partir

de laquelle l'on peut inférer la dualité ci-dessus entre observables (physique) et cycles (topologique)
tel-00001503, version 1 - 24 Jul 2002

sur une 4-variété M :

Si à présent l'on associe, de manière naturelle, l'état "physique" de l'espace-temps à la forme

Lorentzienne de la métrique (échelle de Planck) et l'état topologique à la forme Euclidienne (échelle

zéro), alors il est également naturel d'en déduire qu'entre l'échelle zéro et l'échelle de Planck, il devrait

exister une superposition (+++±) entre les structures métriques (et algébriques) Lorentzienne

(physique) and Euclidienne (topologique).

2.4 Dualité entre mode Physique et mode topologique

Afin d'établir d'un point de vue algébrique les hypothèses de superposition et de dualité évoquées ci-

dessus, nous suggérons à présent d'adopter la démarche proposée en réf[3], consistant à considérer la

métrique d'espace-temps comme soumise à une q-déformation (i.e. déformation quantique au sens de

la réf[13]) à l'échelle de Planck. Cet état quantique de la métrique donne lieu à une nouvelle

description algébrique non plus en termes de groupes classiques mais de groupes quantiques. Dans ce

sens, nous utilisons ici une application du résultat général obtenu en [3] sous la forme du produit

bicroisé cocyclique :

M (H) = Hop H

35
 où H est une algèbre de Hopf, un produit bicroisé (i.e. un type spécial de produit croisé, défini
en réf[12]) et un 2-cocycle ou"twist" au sens de Drinfeld [33]. Cette construction est inspirée par

l'idée d'unifier deux groupes quantiques différents au sein d'une structure algébrique unique. Le point

intéressant est que les deux groupes quantiques impliqués sont reliés dans l'équ.(35) par une relation

de dualité, plus exactement de "semidualité", en un sens expliqué dans la réf[3]. Pour mettre en

évidence cette propriété qui fournit un cadre algébrique à la dualité annoncée entre mode physique et

mode topologique, nous proposons à présent de construire la semidualisation des données

correspondant à

H A
tel-00001503, version 1 - 24 Jul 2002

La forme exacte de l'objet résultant A* H a été conjecturée dans [3] mais demeure non explicite à

certains égards, malgré les progrès effectués dans [34]. Alors, en rapportant les conditions ci-dessus

aux éléments de A* , nous avons :

Proposition 2.4.1 Le produit bicroisé H A admet la semidualisation suivante :

(i) Considérant deux bigèbres X et H, X est un H-module cogébrique à gauche, i.e.

(h x) h(1) x(1) h(2) x (2) et (h x) (h) (x)

(ii) H est un X-module cogébrique cocyclique à droite dans le sens nouveau :

(h x) h(1) x (1) h(2) x(2) et (h x) (h) (x)

(h(1) x(1) ) y(1) (h(2) , x(2) ,y(2) ) (h(1) , x(1) , y(1) )h(2) (x(2) y(2) )

où

(h(1) x(1) ,y(1) , z(1) ) (h(2) ,x (2) , y(2) z(2) ) (h(1) ,x (1) , y(1) ) (h(2) , x(2) y(2 ) ,z(2) )

(h,1, x) (h, x,1) (h) (x)

(iii) les deux algèbres de Hopf sont compatibles dans le sens

36
 h 1 (h),1 1 x (x), (1, x, y) (x) (y)

et

(A) (h(1) ,x (1) , y(1) )h(2) (x (2) y(2) ) h(1) x(1) (h(2) x(2) ) y

(B) (hg) x h (g(1) x(1) ) g(2) x (2)

(C) h(2 ) x (2 ) h(1) x (1) h(1) x (1) h(2) x(2)

(D) (hg, x, y) (h(1) , g(1) x(1) ,(g(2 ) x(2) ) y(1) ) (g(3) , x (3) , y(2) )

Remarque Il résulte de ces données l'existence d'un certain type de double produit croisé cocyclique
tel-00001503, version 1 - 24 Jul 2002

de la forme

X H , dont la structure, d'abord conjecturée dans [3], a été précisée dans [34]. A partir de (ii), il

est clair qu'il s'agît d'une forme de quasi-algèbre de Hopf duale, où le produit serait associatif sous
conjugaison par une fonctionnelle construite à partir de .

Démonstration Pour réaliser la semidualisation, l'on suppose que A est de dimension finie. Soit

X = A* . Les conditions résultantes conservent leur signification pour tout X. Le fait que A soit un H-

module algébrique à droite implique que X est un H-module cogébrique à gauche, en accord avec

a h, x a,h x x A*

Ensuite, nous définissons sur H X X

(h, x,y) x y, (h)

et l'on peut vérifier que H devient un X-module cogébrique à droite, comme annoncé. L'action de X

est donnée ici par la coaction de A selon :

h x x,h( 1 ) h(2 ) h H

37
 Finalement, l'on parvient à semidualiser les conditions de compatibilité (A) et (D). Concernant (B) et

(C), ils sont dualisés selon les formules de la réf [12] pour les produits bicroisés usuels (le cocycle
n'intervient pas). Pour (A), nous considérons x y pour obtenir

(h(1) ,x (1) , y(1) ) x (2) y(2) , a h(2) = x (1) , a(1) h(1) y, a(2) (h(2) x(2) )

ou encore, en utilisant les définitions ci-dessus :

(h(1) ,x (1) , y(1) ) h(2) (x(2) y(2 ) ), a h(1) x (1) ,a(1) (h(2) x(2) ) y, a(2)

(h(1) x (1) )(h(2) x (2) ) y, a

tel-00001503, version 1 - 24 Jul 2002

pour tout a A, qui est la condition (A)- énoncée. De même pour (D).

A présent, appliquons la construction générale obtenue ci-dessus et considérons les structures

algébriques Lorentzienne et Euclidienne.Alors, nous suggérons la proposition suivante (voir la

prop(4.1) de la réf[1]) :

Proposition 2.4.2 Les algèbres de Hopf Euclidienne et Lorentzienne sont reliées par le produit
bicroisé cocyclique de la forme Uq(so(4)) op Uq(so(3, 1))

Preuve Considérant l'approche en termes d'algèbres enveloppantes, à partir de l'algèbre de Hopf

Euclidienne H = Uq(so(4)), nous avons la décomposition bien connue H = Uq(su(2)) Uq(su(2))

ainsi que l'algèbre "opposée" Hop = Uq(su(2))op Uq(su(2))op, tandis que la forme Lorentzienne est
A = H = Uq(su(2)) Uq(su(2)) Uq(so(3, 1)). Comme expliqué en [3], le cocycle de

déformation est = 23 . Alors, l'action et la coaction sont :

(a b) (h g) h(1) aSh(2) g(1)bSg(2)

ß(h g) (h(1) g(1) ).(Sh(3) Sg(3) ) h(2) g(2)

h(1) Sh(3) g(1) Sg(3) h(2) g(2) (36)

38
 où l'on retrouve la structure de produit tensoriel de l'action et de la coaction pour chaque copie
Uq(su(2)). D'autre part, la cocycle h,g Uq(su(2)) est :
(1) (1)
(h g) (h(1) g(1) )(1 )(Sh(4) Sg(4) )(1 )
(2) (2)
(h(2) g(2 ) )( 1)(Sh(3) Sg(3) )( 1)

où le produit est en H = Uq(su(2)) Uq(su(2)). Ceci donne :

(1) (1) (2) (2)

(h g) h(1) Sh(4 ) g(1) Sg(4) h(2) Sh(3) g(2) Sg(3)

(1) (1) (2) (2)

h(1) Sh( 4) g(1) Sg(2) h(2) Sh(3) 1 (37)

pour les structures explicites de produit bicroisé. qed

tel-00001503, version 1 - 24 Jul 2002

Clairement, la prop.(2.4.1) prouve la possible "unification" (à l'échelle de Planck dans notre modèle)

entre les algèbres de Hopf q-Lorentzienne et q-Euclidienne.

Par ailleurs, le résultat ci-dessus suggère un certain type de dualité entre les groupes quantiques

Lorentzien et Euclidien. Pour mettre en évidence cette dualité, l'étape suivante consiste à montrer

l'existence d'une intéressante relation de "semidualité" (proposée dans le cas général par S. Majid

[12]) entre algèbres de Hopf Lorentzienne et Euclidienne. Mieux, une telle dualité fournit une

description de la transition entre le groupe quantique q-Euclidien et le groupe quantique q-Lorentzien

[3]. D'où la proposition :

Proposition 2.4.3 Le groupe quantique Euclidien Uq-1(su(2)) Uq(su(2))

Uq (su(2)) op Uq (su(2)) est connecté par semidualité au groupe quantique Lorentzien Uq(su(2))

Uq(su(2))op* (Uq(su(2))). Alors, la semidualité relie une version de Uq(so(4)) (Euclidien) à

une version de Uq(so(3, 1)) (Lorentzien).

Il existe dans la réf. [3] une démonstration complète de la prop. (2.4.2), basée sur les propriétés du
"double de Drinfeld" (Uq(su(2))). Alors, utilisant la construction en termes de cocycle M (H), nous

obtenons la relation :

39
 semidualisation
Uq(su(2)) Uq(su(2)) Uq(so(4)) Uq(su(2))* Uq(su(2)) ~ Uq(so(3,1))

(38)

La "q-déformation" de l'algèbre de Hopf q-Euclidienne vers l'algèbre de Hopf q-Lorentzienne

correspond à une transformation de dualité et induit l'existence d'un 2-cocycle de déformation. De

même, le produit bicroisé cocyclique

Uq(so(4) op Uq(so(3, 1)) (39)

définit implicitement la nouvelle transformation de (semi)dualité :

tel-00001503, version 1 - 24 Jul 2002

semidualisation
Uq(so(4))op Uq(so(3, 1)) Uq(so(4)) SOq(3,1) Uq(so(4))op

où est construit à partir de celui-ci étant dérivé de la structure quasitriangulaire de Uq(su(2)).

Naturellement, l'on peut également semidualiser à partir des autres facteurs pour construire certains
types de quasi-algèbres de Hopf A H* , associé à H A. Cette fois, la coaction cocyclique

de A sur H est dualisée en une coaction cocyclique de A sur H* tandis que l'action de H sur A est

remplacée par une coaction de H* sur A. La construction devient alors générale, de la forme A Y

(où Y joue le rôle de H* ). L'on obtient ainsi des exemples du type Uq(su(2)) Uq(su(2)) * ,

Uq(so(3, 1)) SOq(4)cop etc., par semidualisation de cette forme. Ceux-ci sont duaux des

constructions précédentes.

A présent, une conséquence intéressante de ces résultats concerne les propriétés de dualité au niveau

de la q-déformation de l'espace-temps lui-même. En effet, à la lumière des constructions précédentes,

l'on parvient à l'importante observation qui suit concernant la transition de la métrique q-Euclidienne à

la métrique q-Lorentzienne:

40
 Corollaire 3.4.4 Pour q 1, la transition de la métrique q-Euclidienne à la métrique q-Lorentzienne
au rayon unité est une dualité de * -algèbre de Hopf Uq(su(2)) SUq(2).

Démonstration Selon une construction introduite par Majid [12,13] et appliquée dans la réf[3], l'on
décrit le q-espace-temps 3,1 par l'algèbre tressée BMq(2), Mq(2) ayant une * - structure unitaire

correspondant à SUq(2) (par * - structure, nous entendons "structures réelles", au sens précisé dans

la réf[12]). Ici, BMq(2) admet une description en tant que matrices hermitiennes tressées. L'on écrit
3, 1
alors q BMq(2). Par ailleurs, l'on note Mq (2) la structure algébrique résultat du twist (au sens
4
de la réf[13]) de Mq(2). Enfin, il a été montré [3] que la - structure de q = Mq (2) est donnée par

la * - structure unitaire de Mq(2) qui, au rayon = 1, donne le * - groupe quantique SUq(2),

tel-00001503, version 1 - 24 Jul 2002

construction duale de celle associée à la * - algèbre de Hopf Uq(su(2)). Explicitement, la -

a b d q 1c
structure de 4q Mq (2) est et coïncide avec celle de la * - structure
c d qb a

unitaire de Mq(2) sur l'identification des deux espaces vectoriels. Or, Uq(su(2)) B(Uq(su(2))) en
tant que * - algèbre sous transmutation [12]. Cette transformation, combinée avec l'auto-dualité de ces

groupes de tresse, donne l'isomorphisme de * - algèbre Uq(su(2)) BSUq(2) comme expliqué ci-
dessus. Il en résulte (cf. [3]) que les structures naturelles de l'espace q-Euclidien 4q et celle du q-

Lorentzien 3, 1
q , covariantes sous U q(so(4)) et U q(so(3, 1)) [32] sont reliées comme suit :

dualité de -algèbres de Hopf 4

U q (su(2)) SU q (2) ~ q
/ 1
Transmutation q - changement de signature

3, 1
BU q (su( 2)) BSUq (2) = / 1
autodualité de groupes - tressés q

(40)

et cette construction rend explicite le changement de signature comme équivalent à une dualité de * -
algèbre de Hopf.

4 3, 1
L'on remarquera que nous obtenons une relation de dualité entre les q-espaces q et q comme une

sorte de T-dualité [27]. Cette interprétation est possible seulement lorsque q 1 - i.e. à l'échelle de

41
 Planck -. L'on peut étendre ces résultats, obtenus à partir des groupes quantiques Lorentzien et

Euclidien, au groupe de q-Poincaré

3, 1
~
Uq (so (3 , 1)) (41)
q

vu, bien sûr, comme dual du groupe de q-Poincaré Euclidien

4
~
Uq (so (4)) (42)
q

En revenant sur la remarque précédente, la dualité d'algèbre de Hopf a été récemment reliée à la T-

dualité en théorie des supercordes par C.Klimcik et P.Sevara [36]. De telles dualités en termes de

groupes quantiques ont également été proposées par S. Majid [13].

tel-00001503, version 1 - 24 Jul 2002

De manière intéressante, les résultats ci-dessus fournissent ainsi certaines indications sur l'origine

algébrique de la fluctuation de signature à l'échelle de Planck, considérée comme transformation de

dualité. Une remarque importante est que certains des isomorphismes ci-dessus sont valides seulement

lorsque q 1, i.e. en théorie non classique. Notons également que la dualité d'algèbres de Hopf au

niveau semi-classique est une dualité de bigèbres de Lie et a été comprise physiquement comme une T-
dualité non abélienne pour des modèles sur G, G* [36], de sorte que la dualité mise en évidence ici

est reliée à d'autres types de dualités en physique.

A présent, appliquons les résultats algébriques obtenus précédemment dans un contexte plus

physique. En effet, la dualité entre les groupes quantiques Lorentzien et Euclidien peut être étendue à

une dualité entre secteur "physique" (Lorentzien) et secteur "topologique" (Euclidien) de la théorie.

D'où la proposition formulée au § (4.3) de l'article [1-A1] :

Proposition 2.4.4 Il existe, à l'échelle de Planck, une symétrie de dualité entre l'anneau de

cohomologie BRST (secteur physique de la théorie) et l'anneau de cohomologie de l'espace des

modules des instantons (secteur topologique)

Ainsi, partant de la forme générique des groupes de cohomologie BRST(cf. réf. [22]), soit

42
 (g)
ker QBRST
H(g)
BRST = (43)
imQ(g 1)
BRST

nous montrons à la prop. (4.3) de (1-A1) que la théorie topologique réalise l'injection d'anneaux :

Uk dk
HgBRST (k) (i) (k)
HBRST g 0 H mod i 0H mod (44)

l'équ. (44) fournissant un chemin injectif du mode physique dans le mode topologique. En termes
d'observables O i et de cycles d'homologie Hi M mod dans l'espace des modules M mod des

configurations du type instantons gravitationnels [ (x)] sur les champs gravitationnels de la

tel-00001503, version 1 - 24 Jul 2002

théorie, nous relevons l'équivalence :

O1O 2 ... O n # (H1 H2 ... Hn )

où le secteur physique de la théorie est décrit par les observables O i et le secteur dual, de type

topologique, par les cycles d'homologie Hi M mod . L'oscillation de signature entre secteur
physique et secteur topologique est alors induite par la divergence Uk j d 4 x du courant-

fantôme j [22]. Lorsque U = 0, comme il n'existe pas d'espace de plongement pour l'espace des

modules, nous suggérons que la théorie est alors projetée dans la branche de Coulomb, à l'origine
de M mod , sur un instanton singulier de taille 0 [37] que nous identifions à l'espace-temps à l'échelle

0. La théorie est ramifiée sur le secteur purement topologique Hi , la signature correspondant à ce

secteur étant Euclidienne (+ + + +).

2.5 Transition entre état topologique et état physique

Nous concluons cette section par une question importante : comment peut-on expliquer, d'un point de

vue cosmologique, la transition de l'état topologique à l'état physique de l'espace-temps? Nous

suggérons une approche dans le cadre de la théorie KMS dans le § (5.2) de l'article [1-A1]. Cette

approche est formalisée dans la conjecture suivante :

43
 Conjecture 2.5.1 A l'échelle infrarouge ß Planck , la brisure de l'état KMS du (pré)espace-

temps induit le découplage entre le flot topologique et le flot physique de la théorie.

Rappelons que nous avons montré en [1-A1] que l'état KMS du (pré)espace-temps à l'échelle de

Planck induit de manière naturelle le couplage entre l'état physique et l'état topologique de la métrique.

Par conséquent, l'hypothèse nouvelle développée en [1-A1] est que la transition entre état topologique

(à l'échelle zéro) et état physique de l'espace-temps (au delà de l'échelle de Planck) peut être décrite en

termes de brisure de l'état KMS. Une telle approche est d'ailleurs renforcée par le fait que la brisure

d'état KMS au delà de l'échelle de Planck pourrait elle-même être liée à la brisure de supersymétrie

attendue à la même échelle. Cette importante relation est d'ailleurs annoncée par les intéressants
tel-00001503, version 1 - 24 Jul 2002

résultats de Deredinger et Luchiesi dans la réf [20], résultats que nous détaillons au §(5.2.1) de [1-

A1].

En fait, Derrendinger et Lucchesi ont clairement confirmé l'existence d'une étroite relation entre la

supersymétrie thermique et la condition KMS. Cette relation est réalisée au niveau des coordonnées de

Grassman thermiques, en raison d'une condition d'(anti)périodicité décrite par les équations (45) et

(46) :

ˆ (t iß) ˆ (t) (45)

ˆ ˙ (t ˆ ˙ (t)
iß) (46)

Les auteurs ont prouvé de manière convaincante qu'en l'absence de la corrélation supersymétrie/état

KMS au niveau de la métrique d'espace-temps, les bosons (périodiques) et les fermions

(antipériodiques) ne peuvent pas appartenir au même multiplet de supersymétrie. A présent, que se

passe-t-il au delà de l'échelle de Planck, lorsque l'état KMS est brisé? Dans ce cas, un point X du

superespace est à nouveau muni des coordonnées de Grassmann usuelles

X (x , , )

44
 Cette condition est équivalente à la supersymétrie à température nulle, pour laquelle les paramètres de

transformation (i.e. les Grassmanniennes and ) sont des constantes. Or précisément, le résultat

principal des réfs [20,21] établit qu'à température finie, il est impossible d'utiliser des paramètres

constants dans les règles de transformations de supersymétrie. Les paramètres de supersymétrie

doivent être des variables dépendant du temps, (anti)périodiques en temps imaginaire. Ainsi, d'une
manière naturelle, les coordonnées thermiques Grassmanniennes Xˆ (x , ˆ (t), ˆ (t)) doivent être
˙

"translatées" en temps imaginaire et sont par conséquent soumises aux conditions

d'antipériodicité ˆ (t iß) ˆ (t) et ˆ ˙ (t iß) ˆ ˙ (t) des équ.(45) et (46). Comme montré en

[1-A1], l'application de ces conditions implique que globalement, le système espace-temps doit être

soumis à la condition KMS à l'échelle de Planck. Cet important résultat peut être considéré comme
tel-00001503, version 1 - 24 Jul 2002

une confirmation de la conjecture (5.5) proposée en réf. [1-A1].

Nous concluons notre étude de l'échelle zéro de l'espace-temps par la recherche d'une
possible amplitude topologique caractérisant l'échelle zéro.

45
 CHAPITRE 3
___

AMPLITUDE TOPOLOGIQUE
A L'ECHELLE ZERO
____

Nous avons suggéré en [1-A1] que la Singularité Initiale de l'espace-temps, non réductible dans le
tel-00001503, version 1 - 24 Jul 2002

cadre habituel de la théorie quantique des champs, peut être décrite dans le cadre de la théorie

topologique des champs (voir section 2). De ce point de vue, l'échelle zéro de l'espace-temps peut

être identifiée à un instanton gravitationnel singulier de taille zéro [3, 37]. Nous poussons ici l'une

des conséquences de cette identification, relevant de l'existence d'une "amplitude topologique" à

l'échelle zéro de l'espace-temps.

3.1 Charge topologique de l'instanton singulier de taille zéro

Nos principaux résultats publiés en [16-A4] suggèrent que la géométrie de l'instanton peut être

identifiée à celle de la boule B4 dont le bord est la sphère S3. L'on peut alors montrer (voir encore

[16-A4]) que le bord de l'espace-temps peut être identifié au bord S3 de l'instanton gravitationnel

singulier B4. Dans un tel contexte, l'échelle zéro de l'espace-temps (i.e. la Singularité Initiale) peut
être entièrement caractérisée par la charge topologique Qs de l'instanton gravitationnel singulier B4,

soit

QS d 4 x R R˜ (47)

Par construction, Qs est un invariant topologique, indépendant de la taille de l'instanton et qui reste

donc défini à l'échelle ß = 0. De ce point de vue, la "propagation" (i.e. pseudo-dynamique

46
 Euclidienne au sens fixé en ([16]) de la singularité initiale est induite par l'existence de l'amplitude
topologique associée à Qs, détectable sur le bord S3 de l'instanton gravitationnel muni de la topologie

B4. Les pseudo-observables sont ici interprétées comme cocycles sur l'espace des modules des
instantons et sont associées aux cycles i de la 4-variété B4 (application de Donaldson). Considérant

un point X de B4, l'amplitude topologique assurant la propagation de la charge instantonique Qs

prend alors la forme:

OS 3 . O X # (S 3 , X) (48)

L'amplitude topologique de la théorie est donnée par les pseudo-observables du membre de

gauche, tandis que le membre de droite désigne le nombre d'intersections des i B4. La fonction

# (S 3 , X) est nulle si le point X est situé hors de la sphère S3 et vaut 1 si X est à l'intérieur de S3 (i.e.
tel-00001503, version 1 - 24 Jul 2002

si X B4), cas où il existe une amplitude topologique.

3.2 Conjecture : origine topologique de l'interaction inertielle

A titre d'application nous conjecturons une approche nouvelle, selon laquelle l’interaction inertielle

pourrait être correctement décrite dans le cadre de la théorie topologique des champs, proposée par

Edward Witten en 1988 [2]. Plus précisément, nous suggérons en réf. [16-A4] que le caractère non
local de la charge topologique Qs peut être relié à la nature non locale de l'interaction inertielle. Nous

conjecturons alors que cette propriété observable ne peut être expliquée en théorie quantique des

champs mais pourrait trouver une solution dans le cadre de la théorie topologique des champs. En

effet, l’évaluation de la contribution inertielle (ou potentiel inertiel) totale résultant de la somme des

masses de l’univers, de la forme :

GM
Uinertiel total = 2 1 (49)
univers c r

s’avère être un invariant pour chaque masse locale. Or, la charge topologique de l’instanton

gravitationnel singulier, de la forme

47
 1
Q 2 d 4 x R R˜ =1 (50)
32

représente également un invariant de type topologique. L’égalité entre la masse inertielle et la masse

gravitationnelle est ici expliquée en termes de quantification de la charge topologique de l’instanton

gravitationnel singulier. Nous en tirons un modèle de "pseudo-propagation" de l'amplitude

topologique Int top, susceptible d'être décrite par les transformations conformes Conf (S3) de la

sphère S3(voir, par exemple la réf. [38] pour rappel). Conf (S3) peut être décrit par le groupe de

Möbius Möb(3) [38], défini à partir de l' inversion de S3. D' où :

Proposition 3.2.1 Pour toute similitude h Sim ( 3), l'application générale définissant la
tel-00001503, version 1 - 24 Jul 2002

charge topologique de l'instanton, i.e. ƒ : S 3 S 3 , définie par ƒ(n) = n et ƒ = g-1 o h o g sur S 3

n appartient au groupe de Möbius Möb(3), groupe conforme de S 3.

La prop. (3.2.1) a été établie dans les réf.[3] et [16-A4].

L'on poursuit en suggérant dans la prop. (3.2.2) que Möb(3) est le groupe conforme Conf (S3) de

S3. Posons que Conf (S3) décrit l'invariance d'échelle (i.e. invariance conforme) de la sphère

identifiée ici, suivant l'inclusion S3 SL(2, C), à l' espace physique, compactifié de 3.

Proposition 3.2.2 Soit Möb± (3) = Conf ± (S3). le rayon r 0 de S 3 engendrant S3r 0, et

ƒ Möb(3), alors S3r 0 appartient au faisceau (S3)de sphères S 3. Réciproquement, une

bijection de S 3 vérifiant cette propriété appartient à Möb (3) . Le groupe Möb (3) présente un
4
isomorphisme naturel avec PO( ) de la quadrique d' équation q x2 x 25 .
i 1

La prop.(3.2.2) a été établie en réf[3]. Dans un tel cadre, la principale conjecture de l'article ci-joint

en annexe A4 est alors que le fondement sur lequel repose le "principe de Mach" [39] (tout comme

l'interaction inertielle) ne doit pas être considérée comme classiquement "physique" mais relève de la

théorie topologique des champs. Nous tirons de notre approche cf. [16-A4], l'existence d'un

"principe de Mach topologique", explicité dans la conj. (3.2.3).

48
 Conjecture 3.2.3 Les amplitudes topologiques associées à la propagation de la charge topologique

de l'instanton gravitationnel singulier de taille zéro correspondant à la Singularité Initiale de l'espace-

temps déterminent le comportement inertiel des masses locales.

A titre d'illustration de la conjecture 3.2.3, nous considérons l'expérience du pendule de Foucault ,

qui ne peut trouver d'explication satisfaisante en mécanique classique ou relativiste [40]. Rappelons

que le problème essentiel consiste en l'invariance angulaire du plan d'oscillation de . Alors, le

"principe de Mach topologique" énonce que l'interaction entre et l'espace-temps global E au sens

de Mach est de type topologique - ce qui pourrait expliquer les propriétés d'invariance globale du

système formé par la plan d'oscillation de et le reste de l'univers.

tel-00001503, version 1 - 24 Jul 2002

Finalement, les perspectives conjecturales introduites dans la réf [16-A4], confèrent une certaine

pertinence à la proposition de "principe de Mach topologique" et, plus généralement, à l'approche

topologique du problème de la singularité initiale. L'on peut espérer que cette approche préliminaire

permettra d'ouvrir des perspectives nouvelles sur l'origine de l'espace temps ainsi que sur plusieurs

autres questions non résolues (notamment celles concernant la nature de l'interaction inertielle).

49
 CONCLUSION
____

Nos principaux résultats, à travers notamment la théorie des groupes quantiques [12], les algèbres

d'opérateurs et la géométrie non commutative [10] et, enfin, la théorie topologique des champs [2]

suggèrent l'existence d'une possible transition de phase concernant la métrique de l'espace-temps

depuis l'échelle topologique zéro jusqu'à l'échelle physique (au delà de l'échelle dent Planck). Ces

résultats font suite à ceux obtenus par G.Bogdanoff dans [3] concernant la fluctuation attendue de la

signature de la métrique à l'échelle de Planck.

tel-00001503, version 1 - 24 Jul 2002

A ce stade, nous résumons nos résultats les plus significatifs :

(i) la métrique de l'espace-temps à l'échelle zéro peut être considérée comme Euclidienne (++++) i.e.

topologique ;

(ii) la singularité initiale de l'espace-temps pourrait être décrite par identification à un instanton

singulier de taille zéro ;

(iii) à partir de (i) et (ii), nous suggérons l'existence d'une symétrie de dualité (que nous appelons i -

dualité), entre état physique (échelle de Planck) et état topologique (échelle zéro) de l'espace-temps.

Une telle symétrie découle directement de la condition KMS à laquelle l'espace-temps devrait être

soumis à l'échelle de ¨Planck.

Alors, la résolution possible de la Singularité Initiale dans le cadre de la théorie topologique nous

amène à envisager l'existence, avant l'échelle de Planck, d'une première phase d'expansion purement

topologique du (pré)espace-temps, paramétrée par la croissance de la dimension de l'espace des

modules dimM et décrite par la "pseudo-dynamique" Euclidienne

= e - ß H MTop
0,1 0,1
ß ( MTop ) eßH

50
 Ainsi, la chaîne d'évènements susceptibles d'expliquer la transition depuis la phase topologique zéro

jusqu'à la phase physique de l'espace-temps (au delà de l'échelle de Planck) pourrait prendre la forme

suivante :

En termes de C* -algèbres, les transformations ci-dessus sont données par :

brisure de Supersymmetrie
Flot topologique 0 ,1 ßH 0,1 ßH
brisure de l'équilibre thermodynamique T ß ( MTop ) e MTop e
Flot KMS Q ß 0
ß H
II *
+
0 ß P
ßc ( M
brisure e c KMS
deq ) l'état Mq e ßc H Flot physique
e iHt M Phys e iHt
P t ( MPhys)
découplage temps imaginaire / temps réel ß 0

Enfin, nous tirons des résultats de laétat

découplage présente recherche
topologique l'existence
/ état physiquede trois phases au cours de
l'expansion du (pré)espace-temps :

(i) - Echelle zéro (Singularité Initiale) : état topologique pur

tel-00001503, version 1 - 24 Jul 2002

(ii)- Echelle de Planck (état KMS) : état topologique + état physique

(iii)- Echelle classique (brisure de l'état KMS) : état physique pur

L'existence de la phase (ii) de fluctuation de la signature de la métrique de l'espace-temps à l'échelle

de Planck a été établie dans la réf.[3]. Les conclusions de [3] sont d'ailleurs renforcées par les
approches récentes de compactification 4D 3D de l'espace-temps à l'échelle quantique (sur ce

point, voir par exemple la réf[41]. En effet, il est généralement admis qu'à haute température

(température de Planck) une théorie dynamique (i.e. la théorie des champs) perd un degré de liberté

par réduction dimensionnelle de la direction genre temps. Dans ce nouveau contexte, il convient par

exemple de relever que la théorie physique de Yang et Mills est transformée en une théorie de Jauge à

trois dimensions. Le point intéressant est alors que le terme de Chern-Simon, introduit par Deser,

Templeton et Jackiw [42], devient naturellement relevant à l'échelle correspondant au couplage entre

théorie topologique à quatre dimensions et théorie physique. Or, le terme de Chern-Simon étant, à

l'échelle de Planck, associé à une 3-surface genre espace, sa présence dans la théorie apparaît ici

comme la condition de fluctuation de la signature de la métrique entre la direction genre temps et la

direction genre espace. Ceci simplement parce que le système devenant indépendant de la quatrième

51
 coordonnée, celle-ci peut être indifféremment considérée comme genre temps et/ou genre espace sans

que le terme de Chern-Simon ne soit perturbé.

Dans un travail ultérieur, nous proposons de développer l'idée selon laquelle à l'échelle zéro, la

dynamique Lorentzienne devrait être remplacée par une "dynamique Euclidienne" (ou pseudo-

dynamique) intrinsèque. Cette pseudo-dynamique Euclidienne, engendrée par les automorphismes

0,1
non-stellaires du facteur "topologique" MTop implique, suivant les résultats de la réf [7], une

"croissance spectrale" du diamètre de l'espace des états d( ) en temps Euclidien (dual de l'espace

des observables en temps Lorentzien). Cette pseudo-dynamique, liée aux automorphismes de semi-
0,1
groupe ß ( MTop ) peut être décrite de manière naturelle par le flot des poids (dans le sens de Connes-

Takesaki [43]) de l'algèbre Mq. Ceci achève de relier, à l'échelle zéro de l'espace-temps, le contenu
tel-00001503, version 1 - 24 Jul 2002

topologique de la Singularité Initiale à la première phase d'expansion du pré-univers, dont la source,

de manière inattendue, pourrait être également topologique.

BIBLIOGRAPHIE
____

1. BOGDANOFF G. BOGDANOFF I. "Topological Field Theory of the Initial Singularity of Spacetime" Class. and
Quantum Gravity vol 18 n° 21 (2001)

2. WITTEN E. "Topological Quantum Field Theory". Com. Math. Phys. 117 (1988)

3 BOGDANOFF G. "Fluctuations Quantiques de la Signature de la Métrique à l'Echelle de Planck" Th. Doctorat Univ.
de Bourgogne (1999)

4. STEWART J.M. "Signature Change, Mixed problems and Numerical Relativity " Class. and Qauntum Grav. Vol 18
n° 23 (2001)

5. HAAG. R. HUGENHOLZ N. WINNINK M. "On the Equilibrium States in Quantum Statistical Mechanics".
Commun. Math. Phys. 5, (1967)

6. BOGDANOFF G BOGDANOFF I. "Spacetime metric and the KMS condition at the Planck Scale" Annals of
Physics 295, (2002)

7. BOGDANOFF I. "KMS State of the Spacetime at the Planck Scale" : to be published in Ch. J. of Phys.
(2002)

8. DOLAN L JACKIW R. "Symmetry Behavior at Finite Temperature". Phys. Rev. D 9 (1974)

9. WEINBERG S. "Gauge and Global Symmetries at High Temperature". Phys. Rev. D 9 12 (1974)

10. CONNES A "Non Commutative Geometry ". Academic Press (1994)

11. DRINFELD V.G. "Quantum Groups". Proc. of the Int. Congress of Maths. Berkeley (1986)

52
 12. MAJID S. "Foundations of Quantum Groups" Cambridge University Press (1995)

13. MAJID S. "q-Euclidean Space and Quantum Wick Rotation by Twisting". J. Math. Phys. 35, N° 9 (1994)

14. ATICK J.J. WITTEN E. "The Hagedorn Transition and the Number of Degrees of Freedom in String Theory". Nucl.
Phys. B 310 (1988)

15. KIRITZIS E. KOUNAS C. "Dynamical Topology Change, Compactification and waves in String Cosmology"
Nucl. Phys. B 442 311-330 (1993)

16. BOGDANOFF I. "Topological Origin of Inertia" Czech.J.Phys. 51, N° 11, (2001)

17. ANTONIADIS I. DEREDINGER J.P. KOUNNAS, C. "Non-Perturbative Supersymmetry Breaking and Finite
Temperature Instabilities in N=4 Superstrings". hep-th 9908137 (1999)

18. KOUNNAS C. "Universal Thermal Instabilities and the High-Temperature Phase of the N=4 Superstrings". hep-th
9902072 (1999)

19. CONNES A. ROVELLI C. "von Neumann Algebra Automorphisms and Time Thermodynamics Relation in General
Covariant Quantum Theories". gr-qc 9406019 (1994)
tel-00001503, version 1 - 24 Jul 2002

20. DEREDINGER J.P. LUCCHESI C. "Realisations of Thermal Supersymmetry" Hep-Ph 9807403

21. LUCCHESI C. "Thermal Supersymmetry in Thermal Superspace" hep-ph 9808435

22. FRÉ P. SORIANI P. "The N=2 Wonderland. From Calabi-Yau Manifolds to Topological Field Theory". World
Scientific Publ. (1995)

23. ALVAREZ-GAUMÈ L. "Supersymmetry and the Atiyah-Singer Theorem".Commun. Math. Phys. 90, (1983)

24. WITTEN E. "Anti de Sitter Space and Holography" hep-th 9802150 (1998)

25. DONALSON S.K. "Polynomial Invariants for Smooth Four Manifolds". Topology 29, 3 (1990)

26. WITTEN E. "Constraints on Supersymmetric Breaking". Nucl. Phys. B 202 (1982)

27. ALVAREZ E. ALVAREZ-GAUMÈ L. LOZANO Y. "An Introduction to T-Duality in String Theory" Nucl.
Phys. B 451 (1995)

28. FLOER A. "An Instanton Invariant for 3-Manifolds". Commun. Math. Phys. 118, (1988)

29. NASH C. "Differential Topology and Quantum field Theory" . Academic Press (1996)

30. KAKU M. "Strings, Conformal fields and M-theory". Springer (2000)

31. WITTEN E. "Supersymmetric Yang-Mills Theory on a four Manifold". J. Math. Phys. 35 N°10 (1994)

32. BACCHAS C.P. BAIN P. GREEN M.B. "Curvature Terms in D-Branes Actions and their M-theory Origin". Hep-
th 990 3210 (1999)

33. DRINFELD V.G. "Quasi-Hopf Algebras". Leningrad Math. J. 1 (1990)

34. SCHAUENBURG P. "Hopf Algebra Extensions and Monoïdal Categories" Communicatin of the author.

35. CAROW-WATUMARA U. SCHLIEKER M. SCHOLL M. WATUMARA S. "Tensor Representation of the

Quantum Group SLq(2, C) and Quantum Minkowski Space". Z Phys. C 48 (1990)

53
 36. KLIMCIK C. SEVERA P. "Poisson Lie T-Duality and Loop groups of Drinfeld Doubles". Phys. Let. 372 (1996)

37. WITTEN E. "Small Instantons in String Theory". hep-th - 9511030 (1995)

38. BERGER M. "Géométrie" (1,2) Nathan, Paris (1990)

39. BARBOUR J.B. PFISTER H. "The Mach Principle" Birkhause, Boston. Einstein Studies, Vol 6 (1995)

40. VIGIER J.P. Foudations of Physics 25 1461 (1995)

41. SEIBERG N. WITTEN E. "Gauge Dynamics and Compactification to Three Dimensions". hep-th 9607163 (1996)

42. DESER S. JACKIW R. TEMPLETON S. "Topological Massive Gauge Theories" Ann. of Phys. 140 (1982)

43. CONNES A. TAKESAKI M. "The flow of weights on factors of type III" Tohoku Math. Journ. 29 (1977)
tel-00001503, version 1 - 24 Jul 2002

54
 ANNEXES

Note : Les tirés à part des articles originaux ont été annexés au manuscrit de la
thèse. Pour des raisons techniques, il n’a pas été possible de les faire figurer dans ce
dossier et ils ont été remplacés par leurs préprints en PDF.
tel-00001503, version 1 - 24 Jul 2002

PUBLICATIONS ANNEXEES

- “ Topological Field Theory of the Initial Singularity of Spacetime ”

Class. and Quantum Gravity vol 18 n° 21 (2001)

- “ Spacetime Metric and the KMS Condition at the Planck Scale ”

Annals of Physics vol 295 n° 2 (2002)

- “ KMS State of the Spacetime at the Planck Scale ”

Ch. J. of Phys. (2002)

- “ Topological Origin of Inertia ”

Czech . J. of Phys. 51,N° 11 (2001)

----------------

55
 Topological field theory of the initial singularity of spacetime

Grichka Bogdanoff and Igor Bogdanoff

Mathematical physics Laboratory CNRS UPRES A 5029 BOURGOGNE University, France

Received 28 november 2000, in final form 22 June 2001

Published 22 October 2001

Abstract

We suggest here a new solution of the initial space-time singularity. In this approach the initial singularity of space-
tel-00001503, version 1 - 24 Jul 2002

time corresponds to a zero size singular gravitational instanton characterized by a Riemannian metric configuration

(++++) in dimension D = 4. Connected with some unexpected topological datas corresponding to the zero scale of space-

time, the initial singularity is thus not considered in terms of divergences of physical fields but can be resolved in the

frame of topological field theory. Then it is suggested that the "zero scale singularity" can be understood in terms of
topological invariants (in particular the first Donaldson invariant ( 1)n i ). In this perspective we here introduce a
i
new topological index, connected with 0 scale, of the form Z = Tr (-1)s, which we call "singularity invariant".
ß 0

Interestingly, this invariant corresponds also to the invariant topological current yield by the hyperfinite II von

Neumann algebra describing the zero scale of space-time. Then we suggest that the (pre)space-time is in

thermodynamical equilibrium at the Planck scale state and is threrefore subject to the KMS condition. This might

correspond to a unification phase between "physical state" (Planck scale) and "topological state" (zero scale). Then we

conjecture that the transition from the topological phase of the space-time (around the scale zero) to the physical phase

observed beyond the Planck scale should be deeply connected to the supersymmetry breaking of the N=2 supergravity.

PACS number 0420D

INTRODUCTION

One of the limits of the standard space-time model remains its inability to provide a description of the

singular origin of space-time. Here we suggest, in the context of N=2 supergravity, that the initial

singularity, associated with zero scale of space-time, cannot be described by (perturbative) physical

56
 theory but might be resolved by a (non-perturbative) dual theory of topological type. Such an

approach is based on our recent results [6-7] concerning the quantum fluctuations (or q-

superposition) of the signature of the metric at the Planck scale. We have suggested that the signature
of the space-time metric (+++ ) is not anymore frozen at the Planck scale p and presents quantum

fluctuations (+++±) until zero scale where it becomes Euclidean (++++). Such a suggestion appears

as a natural consequence of the non-commutativity of the space-time geometry at the Planck scale

[11]. In this non-commutative setting, we have constructed (cf. 4.1) the "cocyle bicrossproduct" [6] :

Uq(so(4)op Uq(so(3, 1)) (1)

where Uq(so(4)op and Uq(so(3,1)) are Hopf algebras (or "quantum groups"[16]), the symbol a
tel-00001503, version 1 - 24 Jul 2002

(bi)crossproduct and a 2-cocycle of deformation (for more specific definitions, see ref[29]). The

bicrossproduct (1) suggests an unexpected kind of "unification" between the Lorentzian and the

Euclidean Hopf algebras at the Planck scale and yields the possibility of a "q-deformation" of the

signature from the Lorentzian (physical) mode to the Euclidean (topological) mode [6-30]. Moreover

equ.(1) defines implicitly a (semi)duality transformation between Lorentzian and Euclidean quantum

groups (see equ.(42)). This is important insofar we consider that the Euclidean theory is the simplest

topological field theory.

In other respect, it has been stated in string theory [25] that the behavior of string amplitudes at very

high temperature (Hagedorn limit) reveals the existence of a possible phase transition and the

restoration of large-scale symmetries of the system. In the context of this "unbroken phase", generally

expected at the Planck scale, the theory is characterized by a general covariance preserving the exact
symmetry of the system. The metric g is developed around zero and there exists at this level neither

light cone, wave propagation, nor movement. The exploration of this unbroken (and non-physical)

phase of the system is accessible only in the framework of a new kind of field theory proposed by E.

Witten under the name "topological field theory" [37].

57
 Topological field theory is usually defined as the quantization of zero, the Lagrangian of the theory
being either (i) a zero mode or (ii) a characteristic class cn (V ) of a vectorial bundle V M built
1
on space-time [31]. Starting from the Bianchi identity TrR R* = 30
TrF F* , our approach of 4D

supergravity leads us to describe the energy content of the pre-space-time system by the curvature R.
We therefore put ~ R R*. The value of the topological action Scl ass class c n (V ) k is
M M

then either zero or corresponds to an integer. The topological limit of quantum field theory, described

in particular by the Witten invariant Z = Tr (-1)n [36] is then given by the usual quantum statistical

partition function taken over the (3+1) Minkowskian space-time

Z = Tr (-1)n e ßH (2)
tel-00001503, version 1 - 24 Jul 2002

1
with ß = and n being the zero energy states number of the theory, for example the fermion number
kT
in supersymmetric theories [1]. Then Z describes all zero energy states for null values of the

Hamiltonian H.

Now, we propose here (§(1.2)) a new topological limit of quantum field theory, non-trivial (i.e.
corresponding to the non-trivial minimum of the action). Built from scale ß 0 and independent of

H, this unexpected topological limit (in 4D dimensions) is then given by the temperature limit

(Hagedorn temperature) of the physical system (3+1)D. In a way this can be derived from the

"holographic conjecture" [42] following which the states of quantum gravity in d dimensions have a
natural description in terms of a (d 1)-dimensional theory. In agreement with [4-34-39] and, in

particular, the recent results of C. Kounnas and al [3-27], we argue in §(5.1.1) that on the hereabove

limit (i.e. at the Planck scale), the "space-time system" is in a thermodynamical equilibrium state [34]

and, therefore, is subject to the Kubo-Martin-Schwinger (KMS) condition [24]. A similar point of

view has also been successfully developed in the context of thermal supersymmetry by Derendinger

and Lucchesi in [13-28]. Surprisingly, the KMS and modular theories [11] might have dramatic

consequences onto Planck scale physics. Indeed, when applied to quantum space-time, the KMS

properties are such that the time-like direction of the system, within the limits of the "KMS strip" (i.e.

58
 between the zero scale and the Planck scale) should be considered as complex : t tr iti . In this

case, on the ß 0 limit, the theory is projected onto the pure imaginary boundary t iti of the

KMS strip. Then the partition function (2) gives the pure topological state connected with the zero

mode of the scale :

Z = Tr (-1)s (3)
ß 0

where s represents the instantonic number. This new "singularity invariant" [6-7]), isomorphic to the

Witten index Z = Tr (-1)F, can be connected with the initial singularity of space-time, reached for ß =
0 in the partition function Z = Tr (-1)s e ßH . According to sec. 3, when ß 0, the partition function

Z gives the first Donaldson invariant

tel-00001503, version 1 - 24 Jul 2002

I= ( 1)n i (4)
i
( k) ( k)
a non-polynomial topological invariant, reduced to an integer for dim mod = 0 (dim mod being the

dimension of the instanton moduli space). This suggests that the (topological) origin of space-time

might be successfully represented by a singular zero size gravitational instanton [41]. A good image of

this euclidean point-like object is the "transitive point", whose orbits under the action of are dense
everywhere from zero to infinity. Then at zero scale, the observables O i should be replaced by the
( k)
homology cycles Hi mod in the moduli space of gravitational instantons. We get then a deep

correspondence -a symmetry of duality- [2-19-32], between physical theory and topological theory.

More precisely, it may exist, at the Planck scale, a duality transformation (which we call

"i-duality"[6]) between the BRST cohomology ring (physical mode) and the cohomology ring of

instanton moduli space (topological mode) [19]. In the context of quantum groups [16-17], we have

shown that transition from q-Euclidean to q-Lorentzian spaces [30-35] can also be viewed as a Hopf

algebra duality [29]. Interestingly, the Hopf algebra duality has been recently connected to

superstrings T-duality by C. Klimcik and P. Sevara [26].

The present article is organized as follows. In section 1 we define the topological field theory and
suggest that there exists at the scale limit ß 0 a non-trivial topological limit of quantum field theory,

59
 dual to the topological limit associated with ß . In section 2 we evidence that the ß 0 limit of

some standard theories is topological. We give several examples of such a topological limit. In section
3, we show that the high temperature limit of quantum field theory corresponding to ß 0 should

give the first Donaldson invariant. The signature of the metric of the underlying 4-dimensional

manifold is therefore expected to be Euclidean (++++) at the scale zero. In section 4, we emphasize,

in the quantum groups context, the existence of a symmetry of duality between the Planck scale

(physical sector of the theory) and zero scale (topological sector). In section 5, we discuss in the

framework of KMS state and von Neumann C* - algebras a way to understand the transition from the

topological (ultraviolet) phase of space-time to the standard physical (infrared) phase.

tel-00001503, version 1 - 24 Jul 2002

1. TOPOLOGICAL THEORY AT SCALE 0

1.1 Preliminaries

The field theory considered here is thermal supersymmetric [13-28] and in the context of D4 manifolds

[40]. We have detailed the content of the (thermal) supermultiplet in a previous work [6]. The theory

belongs to the class of N=2 supergravities [19], the Hamiltonian being given by the squared Dirac

operator D2 [11-31]. As such, the simplest bosonic multiplet reduces to a vector field plus two scalars

exhibiting a special Kähler geometry. Rightly, N=2 is here of a particular interest, for two main

reasons :

(i) the complex scalar fields of the theory (for example the dilaton S-field [32] or the T-field [2]) can

be seen as "signatures" of the KMS condition [11-25] to which the space-time might be subject at the

Planck scale. They might also be one of the best keys to understand the possible duality between

physical observables (infrared) and topological states (ultraviolet) :

i dualité
Topological vacuum (ß = 0, instanton ) Physical vacuum (ß = Planck, monopole )

This is based on the instantons / monopoles duality initially suggested by us in [6] and recently

proved in the superstrings context by C.P. Bacchas, P. Bain and M.B. Green [5]. Moreover, in

60
 string theory again, has been conjectured a U=S T-symmetry [25] from which we can infer the

hereabove duality between (physical) observables and (topological) cycles on a four-manifold M :

U duality
O1O 2 ... O n ( 1, 2 ,..., n)

Then the main contribution of the present article would be to emphasize that, as for conifolds cycles,

a zero topological cycle might control the blow up of the space-time Initial Singularity.

(ii) From another point of view, the S/T fields are closely related to the existence, in the Lagrangian,

of non-linear terms. As recalled by A. Gregori, C. Kounnas and P. M. Petropoulos [23], in the frame

of N=2 supergravity, the theory is generally inducing some non perturbative corrections and a BPS-

saturated coupling with higher derivative terms R2 + ... As our model is proposed in 4D dimensions,
tel-00001503, version 1 - 24 Jul 2002

the development of higher derivative terms can be limited in a natural way to the R2 term. Then the

Lagrangian usually considered in supergravity is :

L= d4 x g l2 ( R R R 2) R + LM (5)

from which we pull the simplified Lagrangian density that we use here :

ˆR 1 2
L supergravité = R RR* (6)
g2

This type of Lagrangian density is coupling the physical component (the Einstein term ˆ R) with the

topological term RR* . This is of crucial interest since, as observed in ex.(2.1), when ß 0, we are

only left with the topological term RR* (decoupled, on this limit, of the axion field ).

Now, let's begin with a brief reminder of topological field theory as originally introduced by E. Witten

in 1988 [37] :

Definition 1.1 Topological field theory is defined by a cohomological field such that a correlation
function of n physical observables O1 O2 ...... On can be interpreted as the number of intersections
O1 O 2 O n #(H1 H2 Hn )

61
 ( k) ( k)
of n cycles of homology Hi mod , in moduli space mod of configurations of the instanton type

[ (x)], on the fields of the theory.

The content of "cohomological fields" (for which the general covariance is exact) is given by the field

variations (which induce a Fadeev-Popov ghost contribution and gauge fixing part). The point,

however, is that the total gauge fixed action is a BRST commutator and the energy-momentum tensor

is BRST invariant [19-37]. In other words, the correlation functions of cohomological fields are

independent of the metric. Now, the topological field theory (for D = 4) is established when the

Hamiltonian (or the Lagrangian) of the system is H=0, such as the theory is independent of the

underlying metric. We propose to extend this definition, stating that a theory can also be topological if
tel-00001503, version 1 - 24 Jul 2002

it does not depend on the Hamiltonian H (or the Lagrangian L ) of the system.

Definition 1.2 A theory is topological if (the Lagrangian L being non-trivial) it does not depend

on L.

Def (1.2) means that L is a topological invariant of the form L = R R . Based on this definition,

we suggest that there exists a second topological limit of the theory, dual to that given by H = 0. In

this case, we can have H 0, but the theory is taken at the limit of scale zero associated with ß 0.

Then the minimum of the action is not zero (as it is in the trivial case) but has a non-trivial (invariant)

value.

We consider the possible existence of such a "topological field " at the high temperature limit of the

system.

1.2 A new topological limit

Proposition 1.3 There exists at the scale ß 0 a non-trivial topological limit of the theory, dual to
the topological limit corresponding to ß .

Proof The (thermo)dynamical content of the quantum field theory can be described by the partition

function :

62
 Z = Tr (-1)n e ßH (7)
where n is the "metric number" of the theory. When ß 0, the theory is no longer dependent on H.

On this limit, such that the temperature T THag(Hagedorn limit), equ.(7) becomes Z0 = Tr(-1)n, H

vanishing from the metric states partition function. ß plays the role of a coupling constant, such that it

exists an infinite number of states not interacting with each other and independent of H. The point is

that for ß = 0, the action S is projected onto a non trivial minimum, corresponding to the self-duality

condition R = ± R * . But in this case, the field configuration is necessarily Euclidean and defines a

gravitational instanton, i.e. a topological configuration [6]. We are therefore confronted to a 4D pure

topological theory , as described by the first Donaldson invariant [14] :

I = ( 1)n i
tel-00001503, version 1 - 24 Jul 2002

ni being the instanton number. The limit ß = 0 is here dual (in a sense precised in §(4)) to the usual

topological limit ß given by H = 0. The density operator of the (pre)space-time system is written

as :

H 0
=e

0 being (classically) a factor of re-normalization of the system. When ß = 0, the density operator is
0
thus reduced to =e , which is independent of H 0, characteristic of a second topological limit of

the theory.

Now we propose to show, through some very simple examples, that interesting contacts with

topological field theory can be made in taking the ß 0 limit of some established standard results. To

be as demonstrative as possible, we shall most often proceed in a heuristic way.

2. THE ß 0 LIMIT OF SOME STANDARD THEORIES

To warm up, we first consider the ß 0 topological limit of the standard (quantum) thermal field

theory.

63
 Example 2.1 The topological 0-scale limit of the heat kernel

(i) One famous mathematical proof of the Atiyah-Singer theorem (given, for example, by E.Getzler

[20]) lies in the heat equation [21-22]. Considering the heat operator e ßH acting on the differential
forms on a closed, oriented manifold X, the ß 0 limit of this operator corresponds to the local

curvature invariants of the manifold [31].

Let's consider a (quantum) thermal field theory on a system defined by the first order elliptic

differential operator P and it's adjoint P*. We put the laplacian = PP* et ' = P*P. For any ß > 0,
ß
we can evaluate the partition function K = Tr e giving the states of the metric of the system.
ß
Now, to get the asymptotic ß 0 limit, we take the symbol of Tr e (which can be expressed in
tel-00001503, version 1 - 24 Jul 2002

terms of ( ) and its derivatives) and we get :

ß ß
Tr e Tr (e )dxdk (8)
M

For ß 0, K degenerates on the Dirac mass and the right-hand side of (8) has an asymptotic

expansion such that

ß i n
Tr e t 2B
i
0

and as a result, we get the well known ß-independent topological index (in the Atiyah-Singer

sense [22]) :

Ind(P) Bn [ ] Bn [ ']

With this index we see in a simple way that the ß 0 limit of thermal field theory is topological.

Another important argument lies in the fact that at Planck , the (pre)space-time might enter a phase of

thermodynamical equilibrium (§(5.1.1)). Consequently (§(5.1.2)) it should be subject to the KMS

condition [24]. As evidenced in §(5.1.3) and in the ex.(5.2.1), this implies the holomorphicity of the
c
time-like direction, the real time-like and the real space-like directions given by g44 being

64
 compactified on the two circles S1t like and S1s like [6]. But one can easily see that this configuration is

equivalent to the dimensional reduction of the 4D Lorentzian theory onto a 3D theory. This type of

reduction has been described by Seiberg and Witten [33]. We then are left with three-manifold
invariants, in particular the Floer invariant of a supersymmetric non linear -model [18]. In this case,
the three dimensional pseudo-gravity (3) is coupled to the S,T complex scalar fields :

1
(1) S = ± i. i (axion) with S and S
g2

(2) T = g44 ± i g* i4 with T and T

Those scalar fields are propagating. Then the coupling of the S/T-fields with the 3D pseudo-gravity
tel-00001503, version 1 - 24 Jul 2002

is given by the extended -model :

SL(2, ) SL(2, )
= S O(3) (9)
SO(2) SO(1,1)

SL(2, ) SL(2, )
As the theory is independent of g44 , the 2D field in the Lorentzian case and in the
SO(1,1) SO(2)

Euclidean case can be viewed as equivalent. Thus the corresponding "superposition state" of the

signature (+++±) is able to be described by the symmetric homogeneous space

SO(3,1) SO(4)
h =
SO(3)

SO(3) being diagonally embedded in SO(3, 1) SO(4). Next step, as suggested in [6], a

"monopoles+instantons" configuration can be associated to this 5D metric configuration at the Planck

scale. Instantons and monopoles are here connected by a S-field. The form of the 5D metric induced
by the -model (9) and constructed in [6] is :

dw 2
ds2 a(w)2 d 2
(3) dt 2 (10)
g2

where the axion term is a = ƒ(w , t), the 3-geometry being d 2(3) (x, y, z). Clearly the expected
1
values of the running coupling constant (dilaton) are giving the two 4D limits of the 5D metric
g2

65
 dw 2
of equ.(10). Thus we get: - Infrared : ß . In this strong coupling sector we have 0 and
g2

the w direction of 5 is cancelled. So after dimensional reduction (D=5 D= 4) the metric on 5

becomes 4D Lorentzian :

ds2 a(w)2 d 2
(3) dt 2 (11)

The -model (9) is reduced to the usual Lorentzian symmetry :

infrared
SL(2, ) ß(g)
S O(3) SO(3, 1) (12)
SO(1,1)

Likewise, when g , the R2

term cancels in the 5D Lagrangian density
tel-00001503, version 1 - 24 Jul 2002

1 2
L supergravité = ˆ R R RR* , and, as R = R *, the topological term RR* is also suppressed.
g2

So, we get L = ˆ R .

Let's see now what happens on the (dual) ultraviolet limit, when ß 0.

- Ultraviolet : ß 0. We can construct a boundary of equ.(10), corresponding to the small

coupling constant sector of the coupled theory and we get divergent values for the real dilaton field
1
. Then naïvely, we can apply one of the results of [23] saying that the axion field is decoupled
g2

of the theory on this limit and we are left with the divergent dilaton field only. So, we have for the
metric on 5 the new Euclidean form :

dw 2
ds2 a(w)2 d 2
(3) (13)
g2

Therefore, in the ultraviolet, the -model (9) is reduced to the four-dimensional target space :
ultraviolet
SL(2, ) ß(g) 0
S O(3) SO(3) SO(3) = SO(4) (14)
SO(2)

and on this small coupling limit, the reduced theory becomes Euclidean, i.e. topological.. Again, it
appears reasonable to conclude that the ß 0 limit has a pure topological content. Incidentally, this

result could as well be understood in the frame of the isodimensional instanton-monopole duality

66
 proposed by us in [6] and proved in the string context by Bacchas and al [5]. Indeed, we have shown
that the q-deformed 5D theory is dominated by the (3+1)D monopoles in the infrared (ß Planck )

and by the 4D instantons in the ultraviolet ( ß 0) [6]. In this sense, the Euclidean signature (++++)

can be seen as i-dual to the Lorentzian one (+++ ). Likewise, the topological limit ß 0 should be
viewed as i-dual to the physical limit ß Planck . This might be an unexpected application of the

Seiberg-Witten S/T-duality [32].

At present, let's explore the ultraviolet limit of another standard result, i.e. the Feynmann Path integral

[39].

Example 2.2 The topological 0-scale limit of the Feynmann (3+1) path-integral approach
tel-00001503, version 1 - 24 Jul 2002

(i) It's well known that in quantized Minkowski space-time, the amplitude (g2 , 2 , 2 g1 , 1 , 1 ) is

given by :

(g2 , 2 , 2 g1 , 1 , 1 ) = D[ ] exp [i S( )]

To include the point-like (0-modes) configurations of g , we put Tr(-1)n in the integral and we get :

(g2 , 2 , 2 g1 , 1 , 1 ) = Tr(-1)nD[ ] exp [i S( )] (15)

So, the trivial {t=ß 0, S=0} Lorentzian vacuum is distinct of the "topological vacuum" connected
8 2
to the minimum of the Euclidean action SE = 2 . But it has been shown [15-37] that the zero modes
g

in the expansion about the minima of S are tangent to the instanton moduli space Mk, so the

topological vacuum should be viewed as the "true vacuum" of the theory. Then equ.(15) becomes for
ß 0:

(g0 , 0 , 0 g0 , 0 , 0 ) = I0 = Tr(-1)n D[ 0 ] (16)

To define I0 , one can assume that at zero scale, the measure D[ 0 ] is concentrated on one unique point

and becomes a pure state, i.e. a positive trace class operator with unit trace. Concerning the field

67
 content can be given by the non linear term R 2 , so that the ß-dependant typical form of the Lagrangian

density is, as seen in [6] :

1 2
L supergravité ˆR R RR* (17)
g2

Now, for g= ß 0, the Einstein term R is cancelled and as R =R* , the only remaining term in

equ.(17) is the topological invariant RR*(itself decoupled from the axion field ). So, equ.(15) takes

the new form :

(g0 , 0 , 0 g0 , 0 , 0 ) Tr R 2 = TrRR* = I0 (18)

and I0 becomes a topological invariant. As

tel-00001503, version 1 - 24 Jul 2002

Tr(R(A)2 ) 2
8 k(E)
X

and we apply the Gauss-Bonnet theorem to find :

1
(M) 2 abcd ab R Rcd (19)
32 X

Therefore, the ß 0 limit of the Feynmann path integral is giving the Euler Characteristic, i.e. the

"true vaccum" mentioned hereabove and corresponding to the topological pole of the theory.

Next, we provide a new example showing that the ß 0 limit of the N=2 supersymmetric theory is

topological.

Example 2.3 The topological 0-scale limit of the (supersymmetric) quantum field theory

We apply here a well known quantum mechanical account of Morse theory due to Witten [40]. First,
we start from the standard supersymmetry algebra Qi ,Qj Qi Qj Qj Qi 0. Next, we express this

superalgebra in terms of data provided only by the space-time manifold M. To do so, let's define a set
1
of coboundary operators, the conjugation of d by e ßH being parametrised by ß = :
kT

68
 d e ßH d e ßH

(20)
d *ß e ßH d* e ßH

for a Morse function H(x). Then the spectrum of the ß-dependant Hamiltonian is:

Hß d ß d *ß d*ß d ß (21)

Now, let's send ß onto zero. We get for the Hamiltonian the invariant value :

H0 = dd* + d* d = p (22)
p 0

But this invariant is nothing else than the Betti numbers of M, given by bp dim ker p , which is a

discrete function, independent of ß. Consequently, the space of zero energy states of H is given by the
set of even (odd) harmonic forms on M and equals the Betti number of M. So we have, for ß 0:
tel-00001503, version 1 - 24 Jul 2002

4
Tr( 1)F = ( 1)k bi = (M) (24)
k 0

where bi is the ith Betti number and (M) the Euler-Poincaré characteristic of M. Finally, on the zero

scale limit, we recover the topological index [37] corresponding to any standard topological field

theory.

To finish, we obtain in the last following example some analog results in the frame of full (N=2)

supergravity.

Example 2.4 The topological ß 0 limit of (N=2) supergravity

As a matter of fact, for a spin manifold, we can express H in terms of the Dirac operator D . Then in
dimension D=4, we can calculate on the ß 0 limit the index of the squared Dirac operator :

2
ß
Ind( D +) = lim Str e =
ß 0
2 1 1
R , (R R) , B
1 2 16
n Str e dx d
(2 )
T *M

By the Mehler formula, we find the Dirac index in function of the Dirac genus Â( M):

69
 ind ( D + ) = ch(B)Â( M) (25)
M

ch being the Chern character, B the curvature and Â( M) the Dirac genus of the auxillary fiber bundle.
Since the spinors are interacting with Yang-Mills fields, the Â( M) term is coming from the
B
gravitational part whereas the rest of equ.(25) comes from the gauge part. As Ch(B) = Tr e 2i ,
we get :
k xj / 2
Â( M) = (26)
j = 1sinh (x j / 2)

and we can express the complete Yang-Mills + gravity index through the following invariant :

dim M 1
Ind(D ) Tr(R R) Tr(F F) (27)
8 2 8 2

Finally, in Yang-Mills + gravity context, we obtain again a topological invariant on the ß 0 limit.
tel-00001503, version 1 - 24 Jul 2002

Now, to go further, the next step consists to detect, on the ß 0 limit, the nature of the topological

invariant involved . We shall discover that Donaldson invariants are playing a very important role on

this boundary.

3. ß 0 SCALE AND DONALDSON INVARIANTS

From a topological point of view, Donaldson invariants are obtained from characteristic classes of an

infinite dimensional bundle on the manifold equally infinite and canonically associated with a

4-dimensional manifold :

Definition 3.1 Let M be a 4-dimensional manifold . The Donaldson invariant qd (M ) is a

symmetric integer polynomial of degree d in the 2-homology H2 (M; Z) of M

qd (M ) : H2 (M) ... H2 (M) Z
( k)
mod being the instanton moduli space of degree k, the Donaldson invariant is defined by the map

H2 (M) (k)
m : H2 (M ) mod

Now, we suggest that on the ß 0 limit, the 4D field theory is projected onto the first Donaldson

invariant.

70
 Proposition 3.2 The high temperature limit of quantum field theory corresponding to ß 0 in the

partition function Z = Tr (-1) s e ßH gives the first Donaldson invariant. The signature of the metric

of the underlying 4-dimensional zero scale manifold is therefore Euclidean (+ + + +).

Proof Let the partition function Z = Tr(-1)s e ßH connected with a set described by the density

matrix :

Q = (-1) s e ßH (28)

According to standard arguments, we can write :

Tr (-1) s e ßH = d (t) d (t)exp SE ( , ) (29)

CPB

It has been shown [1-36] that given a supersymmetric QFT, one can define the invariant I = Tr (-1)f,
tel-00001503, version 1 - 24 Jul 2002

f being the fermionic number. We propose to extend equ.(29) to supergravity and to define the

topological invariant

= Tr (-1) S (30)

where S is the instanton number. So, the regularization of the trace (30) gives the index of the Dirac

operator :
ß
dt L
ßc D2 ßc D2
= Tr e = Tr (-1) S e = [Dx] [D ] e 0
(31)
cpl
ßc H
with ßc . Then for ßc = 0, the value of the partition function Z = Tr (-1) S e is :

Z0 = Tr (-1) S (32)

and Tr (-1)S can be seen as the index of an operator acting on the Hilbert space . Dividing in
monopole and instanton sub-spaces = + i and Q being a generator of supersymmetry, we

get :
Q = 0 , Q* =0 (33)

So Tr (-1)S = Ker Q - Ker Q* such that as topological index, Tr (-1)S is invariant under continuous

deformations of parameters which do not modify the asymptotic behavior of the Hamiltonian H at high

71
 energy. H is given by H = dd* +d*d, the space of zero energy states corresponding to the set of even
harmonic forms on Mn:
n
Tr (-1)S e ßH = (M) = ( 1)k bk (34)
k 0

= Tr (-1)S is independent of ß, the sole contributions to coming from the topological sector of zero
= niE 0
nmE 0.
On formal basis, niE 0
nmE 0
energy : can be seen as the trace of the operator (-

1)S . Then is a topological invariant, i.e. the first Donaldson invariant. The coupling constant g being
dimensional, the limit ß = 0 implies = 0 and corresponds to the sector of zero size instantons [41].
( k) ( k)
So, Dim mod = 0. When Dim mod 0, the Donaldson invariants are given by :
r r
S ( k)
Z( 1 ... r) DX e Wk 1 Wk i (Dim mod 0) (35)
i 1 i 1
tel-00001503, version 1 - 24 Jul 2002

i i

What happens then ? The solution is in the correspondence between the Donaldson invariants on 4D

manifolds and the Floer homology groups [18] on 3D manifolds. Indeed, Donaldson invariants

amount to the calculus of the partition function Z, expressed as an algebraic sum over the instantons

[15]:

Z Z = ( 1)n i (36)
(k)
mod 0 i

i indicating the i th instanton and ni = 0 or 1 determining the sign of its contribution to Z. Donaldson
has shown on topological grounds [14] that when dim (mod k)
= 0, then ( 1)n i is a non-polynomial
i

topological invariant, reduced to an integer. We find the same result starting from T Q, .

In fact, the partition function of the system at temperature ß-1 has the general form Zq = Tr (-1)S e ßH
. For ß = 0, Zq becomes Z = Tr(-1)S , which is isomorphic to ( 1)n i , s and ni giving in both
ß= 0 i

cases the instanton number of the theory.

This result strongly suggests that on the high temperature limit ß 0 parameterizing the 0 scale of the
theory, the partition function Z projects the Lorentzian physical theory onto the Euclidean
(k)
mod 0

topological limit.

72
 Now, starting from hereabove, we suggest the existence of a deep correspondence, of the duality

symmetry type, between physical sector ( Planck scale) and topological sector (0 scale) of the

(pre)space-time.

4. DUALITY SYMMETRY BETWEEN PHYSICAL AND TOPOLOGICAL STATES

1
Ideally, the duality we are looking for (which we call "i-duality" t [5], of the type i = S T)
it
should exchange real time in strong coupling / large radius with imaginary time in weak coupling /

small radius. In this sense, Planck (physical) scale should be i-dual to zero (topological) scale.

Let's first outline a few formal aspects of Lorentzian/Euclidean duality in terms of Hopf algebras.
tel-00001503, version 1 - 24 Jul 2002

4.1 Duality between q-Lorentzian and q-Euclidean Hopf algebras

Considering the non commutative constraints at the Planck scale, it appears interesting to adopt an

approach in terms of "quantum groups" at this scale. So we have shown that in D=4, it should exist a

superposition (+++±) between Lorentzian (physical) and Euclidean (topological) algebraic structures.

Then we have constructed, in the enveloping algebras setting, the q-deformation of the cocyle

bicrossproduct [6]:

M (H) = Hop H

where H is a Hopf algebra, a bicrossproduct (i.e. a special type of crossproduct, defined in [29])
and a 2-cocycle or"twist" in the Drinfeld sense [16-17]. This is inspired by the idea to unify two
different quantum groups within a unique algebraic structure. So, we propose the following :

Proposition 4.1 The Euclidean and the Lorentzian Hopf algebras are related by the cocycle

bicrossproduct

Uq(so(4)) op Uq(so(3, 1))

Proof Starting, in the setting of enveloping algebras, from the Euclidean Hopf algebra H =
Uq(so(4)), we have the well known decomposition H = Uq(su(2)) Uq(su(2)) and the "opposite"
Hop = Uq(su(2))op Uq(su(2))op, whereas the Lorentzian form is A = H = Uq(su(2))

73
 Uq(su(2)) Uq(so(3, 1)). As explained in [6], the cocycle of deformation is = 23 . Then the action

and the coaction are :

(a b) (h g) h(1) aSh(2) g(1)bSg(2)

ß(h g) (h(1) g(1) ).(Sh(3) Sg(3) ) h(2) g(2)

h(1) Sh(3) g(1) Sg(3) h(2) g(2) (38)

where we find the structure of tensor product of the action and the coaction for each Uq(su(2)) copy.
On the other hand, the cocycle for h,g Uq(su(2)) is :
(1) (1)
(h g) (h(1) g(1) )(1 )(Sh(4) Sg(4) )(1 )
tel-00001503, version 1 - 24 Jul 2002

(2) (2)
(h(2) g(2 ) )( 1)(Sh(3) Sg(3) )( 1)

where the product is in H = Uq(su(2)) Uq(su(2)). This gives:

(1) (1) (2) (2)

(h g) h(1) Sh(4 ) g(1) Sg(4) h(2) Sh(3) g(2) Sg(3)

(1) (1) (2) (2)

h(1) Sh( 4) g(1) Sg(2) h(2) Sh(3) 1 (39)

for the explicit bicrossproduct structures. qed

Clearly, prop.(4.1) proves the possible "unification" between the q-Lorentzian and the q-Euclidean

Hopf algebras at the Planck scale.We give a detailed demonstration of this proposition in [6]. But

also, the hereabove result suggests a certain type of "duality" between Lorentzian (physical) and

Euclidean (topological) quantum groups. To see this, the next step consists in showing the existence

of a very interesting "semidualisation" (proposed in the general case by S. Majid [29]) between

Lorentzian and Euclidean Hopf algebras. Better still, such a duality allows a description of the

transition from the q-Euclidean group to the q-Lorentzian group [30] :

Proposition 4.2 Uq-1(su(2)) Uq(su(2)) Uq (su(2)) op Uq (su(2)) is connected by

semidualisation to Uq(su(2)) Uq(su(2))op* (Uq(su(2))). Then the semidualisation connects a

version of Uq(so(4)) to a version of Uq(so(3, 1)).

74
 We have given in [6] a complete demonstration of prop. (4.2), based on the properties of the Drinfeld
double (Uq(su(2))). Then, using our general cocycle construction M (H), we get the interesting

relation :

semidualisation
Uq(su(2)) Uq(su(2)) Uq(so(4)) Uq(su(2))* Uq(su(2)) ~ Uq(so(3, 1))

(40)

The "q-deformation" from q-Euclidean to q-Lorentzian Hopf algebras corresponds to a duality

transformation and induces the existence of a 2-cocycle of deformation. Likewise, the cocycle

bicrossproduct

Uq(so(4) op Uq(so(3, 1)) (41)

tel-00001503, version 1 - 24 Jul 2002

defines implicitly the new (semi)duality transformation

semidualisation
Uq(so(4))op Uq(so(3, 1)) Uq(so(4)) SOq(3,1) Uq(so(4))op

where is constructed from this one being derived from the quasitriangular structure of

Uq(su(2)) [5].

Now, an interesting consequence of those results concerns some duality characteristics at the level of

q-deformation of space-time itself. We have shown [6] that the natural structures of the q-Euclidean
space 4q and of the q-Lorentzian space 3, 1
, covariant under U q(so(4)) and U q(so(3, 1)) [8] are
q

connected as follows :

-Hopf algebras duality 4

Uq (su(2)) SU q (2) ~ / 1
q
Transmutation q - signature change (42)

3, 1
BU q (su(2)) BSU q (2) = / 1
- braided groups autoduality q

4 3, 1
where we get a duality relation between q and q as a kind of T-duality [2]. This interpretation is

possible only when q 1 - i.e. at the Planck scale -. We can extend those results to q-Poincaré

groups

75
 3, 1
~
Uq (so (3 , 1)) (43)
q

seen as dual to the Euclidean q-Poincaré group

4
~
Uq (so (4)) (44)
q

Interestingly, the Hopf algebra duality has been recently related to superstrings T-duality by

C.Klimcik and P.Sevara [26]. Such dualities in terms of quantum groups have also been proposed by

S. Majid [29].

Now, we apply the hereabove results into a more physical context. So, we propose the following :

Proposition 4.3 There exists, at the Planck scale, a symmetry of duality between the BRST
tel-00001503, version 1 - 24 Jul 2002

cohomology ring (physical sector of the theory) and the cohomology ring of instanton moduli space

(topological sector).

Proof Let be, at the Planck scale, BRST cohomology groups, of which the generic form, reviewed

in [37], is :
(g)
(g) ker QBRST
H BRST = (45)
(g 1)
imQBRST

(g)
where QBRST is the BRST charge acting on operators of the ghost number g. From the theory of

Donaldson [14-15], we conclude the existence, at 0 scale of space-time, of cohomology groups

constructed by de Rham :

(i) (k ) ker d (i)

H (M mod ) (46)
im d ( i 1)

where d (i) represents the external derivative acting on the differential forms of degree i on ( k)
mod .

Topological theory then brings about ring injection which follows:

Uk g (k ) dk (i ) (k)
H BRST g 0
HBRST H mod i 0
H mod (47)

76
 and which, according to conditions given in [19], becomes a ring isomorphism. There exists therefore
an injective path from the physical mode to the topological mode. Now let O i be the physical

observables considered, such that a correlation function of n observables is the number given by the
matrix of intersections Hi :

O1 O 2 .. . O n # (H1 H2 ... Hn ) (48)

( k)
number associated with n cycles of homology Hi M mod , in moduli space mod of configurations

of the gravitational instanton type [ (x)], on the gravitational fields of the theory. The physical

sector of the theory is described by the left hand side of equation (48) and the topological sector by the
right hand side. One observes that O1O 2 ... O n 0, i.e. the theory has a physical content if
tel-00001503, version 1 - 24 Jul 2002

Uk j d 4 x , with j being the "ghost flow" of degree k, U its integral anomaly and

di gh [O i ] the ghost number of O i . Moreover

(k )
dk dim mod (49)

is the dimension of moduli space of degree k. Following the theorem of Atiyah-Singer [21], one can
show that Uk = dk . From this point of view, the correlation functions of a set of local observables

G(x1 ... xn ) O(x1 ) ... O(xn ) (50)

amounts to the integral over moduli space of the number of cohomology classes of space. The

associated BRST charge Q is of the form Q = (-1)n . When the divergence of the ghost flow is non-
zero, i.e. jM 0, then the theory oscillates between ( O i ) and (Hi) - i.e. between the Coulomb

branch and the Higgs branch in metric superposition space - . For the 0 mode of the scale, j M = 0,

then

O1O 2 ... O n = 0 (51)

( k)
which suggests that on this limit, dim mod = 0. In fact, after functional integration over the empty

degrees of liberty of the theory, the physical observables are reduced to closed forms i of degree di,

which signifies :

77
 ( k)
U = dim mod

and when U = 0, there exists no embedding space for moduli space and the theory is projected into
( k)
the Coulomb branch, at the origin of mod , on a singular instanton of zero size, identified to space-

time at zero scale. The corresponding signature in this sector of the theory is therefore Euclidean

(+ + + +). qed

This result suggests once more that at zero scale, the theory is no longer physical but purely

topological.

Now, here is a critical question raised by this paper : how do we go from the topological state of the
tel-00001503, version 1 - 24 Jul 2002

(pre)space-time around the origin to the usual physical state ? In the last section, we shall try to answer

this question.

5. TRANSITION FROM INITIAL TOPOLOGICAL PHASE TO STANDARD

PHYSICAL PHASE

Considering all the preceding developments, it's of crucial interest to worry about how the initial

(generally covariant) topological phase possibly characterizing the (pre)space-time at the vicinity of the

Initial Singularity does break down to the universe we observe to day. We then propose some

(hopefully) stimulant tracks able to be worked out within some further researches.

On general basis, we claim hereafter that the transition Topological phase Physical phase might be

deeply related to the breaking of the N = 2 supergravity at the Planck scale. In other words,

supersymmetry breaking, as showed by C. Kounnas and al in superstrings context [3-27], is

characterized by the loss of the thermodynamical equilibrium of the system. To sum up, the D-

dimensional space-time supersymmetry is spontaneously broken in (D-1) dimensions by thermal

effects. For this reason, supersymmetry breaking might bring about the decoupling of the topological

and the physical states of the (pre)space-time system. How is it so ? To see it, according to [4-27],

78
 let's recall that at the Planck scale, the (pre)space-time is generally characterized by two fundamental

properties : (i) the thermodynamical equilibrium state [34] and (ii) the non-commutativity of the

underlying geometry [11]. Those two properties are very often considered, together or separately.

However, it is critical to realize that for any system, properties (i) and (ii) are inducing the famous

"Kubo-Martin-Schwinger"(KMS) condition [24]. Therefore, we propose now to consider that, most

likely, space-time, as a thermodynamical system, is subject to the KMS condition at the Planck scale
[6]. Consequently, in the interior of the "KMS strip", i.e. from ß = 0 to ß = Planck, the fourth

coordinate g44 should be considered as complex, the two real poles being ß =0 (topological pole) and

ß= Planck (physical pole). This is a direct (and standard) consequence of the KMS condition. So, we

suggest [6] that within the KMS strip, the Lorentzian and the Euclidean metric are in a "quantum
tel-00001503, version 1 - 24 Jul 2002

superposition state" (or coupled), this entailing a "unification" (or coupling) between the topological

(Euclidean) and the physical (Lorentzian) states of space-time. Conversely, the transition from the

topological state to the physical state of the space-time can be seen in terms of "KMS breaking" (cf.

conj. (5.2.5)).

Now, let's begin with the hypothesis of global thermodynamical equilibrium at the Planck scale.

5.1 Thermodynamical equilibrium and KMS state of the space-time at the Planck

scale

5.1.1 Thermodynamical equilibrium of space-time

From a thermodynamical point of view, it appears that the Planck temperature

1
EP c5 2
ß -1
planck Tp kB1 1,4 1032 K
kB G

represents the upper limit of the physical temperature of the system. Indeed, it is currently admitted

that, before the inflationary phase, the ratio between the interaction rate ( ) of the initial fields and
the (pre)space-time expansion (H) is >> 1, so that the system can reasonably be considered in
H
equilibrium state. This has been established a long time ago within some precursor works of S.

79
 Weinberg [34], E. Witten [4] and several others. It has recently been shown by C. Kounnas and al in

the superstrings context [27]. However, this natural notion of equilibrium, when viewed as a global

gauge condition, has dramatic consequences regarding physics at the Planck scale. Which kind of

consequences? To answer, let's see on formal basis what an equilibrium state is.

Definition 5.1 H being an autoadjoint operator and the Hilbert space of a finite system, the
Tr (e ßH A)
equilibrium state of this system is described by the Gibbs condition (A) and
Tr (e ßH )
satisfies the KMS condition.
1
Here, Tr is the usual trace, ß is the inverse of the temperature, H the Hamiltonian, i.e. the
kT
generator of the one parameter group of the system. Of course, A is a von Neumann C* - algebra
tel-00001503, version 1 - 24 Jul 2002

(see §(5.1.4) for definitions). The equilibrium state implies that ß must be seen as a periodic
(imaginary) time interval [0, ß = Planck ]. Now, the famous Tomita-Takesaki modular theory [10-11]

has established that to each state (A) of the system corresponds, in a unique manner, the strongly
continuous one parameter * - automorphisms group t :

t (A) = ei Ht A e i Ht (52)

with t . This one parameter group describes the time evolution of the observables and

corresponds to the well known Heisenberg algebra. At this stage, we are brought to find the
remarkable discovery of Takesaki and Winnink, connecting (i) the evolution group t (A) of a
Tr(Ae ßH )
system (i.e. the modular group M = it A it) with (ii) its equilibrium state (A)
Tr(e ßH )

[11]. The famous "KMS condition" [24] is nothing else than this relation between t (A) and (A),

the content of this relation being precised in (i) and (ii) of §(5.1).

Then we claim in a natural way that the space-time, in equilibrium state at the Planck scale, is therefore

subject to the KMS condition at this scale.

80
 5.1.2 The (pre)space-time in KMS state at the Planck scale

When viewed as a hyperfinite system at the Planck scale, the (pre)space-time may be described by a

von Neumann C* -algebra A (a von Neumann algebra is hyperfinite if it is generated by an increasing

sequence of finite dimensional sub-agebras). Now, let's see the essence of the KMS condition, given
by the Haag-Hugenholtz-Winnink theorem [23] : a state on the C* -algebra A and the continuous

one parameter automorphisms group of A at the temperature ß = 1 / k T verify the KMS condition if,
for any pair A, B of the * - sub-algebras of A, it exists a ƒ(tc) function holomorphic in the strip

{tc = t + i ß , Im t c [ 0 , ß ] } such that :

(i) ƒ (t) = (A ( t B)) ,

tel-00001503, version 1 - 24 Jul 2002

(ii) ƒ (t + i ß ) = ( t (B)A) , t . (53)

Then we observe with (i) and (ii), the two crucial properties of the KMS condition : the holomorphicity
of the KMS strip and of course, due to the cyclicity of the trace, the non commutativity (( t A)B) =

(B( t +i ßA)) characterizing any "KMS space" (in fact, the two boundaries of the strip do not

commute with each other).

Now, if we admit that around Planck , the hyperfinite (pre)space-time system is in a thermal

equilibrium state, then according to [24], we are also bound to admit that this system is in a KMS
state. Incidentally, another good reason to apply the KMS condition to the space-time at Planck is that

at such a scale, the notion of commutative geometry vanishes and should be replaced by non

commutative geometry [11]. In this new framework, the notion of "point" in the usual space collapses

and is replaced by the "algebra of functions" defined on a non commutative manifold. Non

commutative geometry and quantum groups theory [16-29] are addressing such non-commutative

constraints. But the non-commutativity induced by the KMS state is in natural correspondence with the

expected non commutativity of the space-time geometry at the Planck scale.

Next, let's push forwards the consequences raised by the holomorphicity of the KMS strip.

81
 5.1.3 Holomorphic time flow at the Planck scale

As a consequence of the application of the KMS condition to space-time itself, we are induced to
consider that the time-like coordinate g00 becomes holomorphic within the limits of the KMS strip. So

we should have [11-24] :

t = tr + i ti (54)

as showed in [6]. In the same way, the physical (real) temperature becomes also complex at the Planck

scale :

T Tc = Tr + iTi (55)

as proposed by Atick and Witten in another context [4]. So, the KMS condition suggests the existence
tel-00001503, version 1 - 24 Jul 2002

at the Planck scale, of an effective one loop potential coupled, in N = 2 supergravity, to the complex
1
dilaton + axion field = 2 i and yielding the following dynamical form of the metric
g
i
diag(1 , 1 , 1 , e ) (56)

The signature of (56) is Lorentzian (physical) for =± and can become Euclidean (topological) for

= 0. This unexpected effect is simply due to the fact that, within the boundaries of the analytic KMS

field -i.e. from the scale zero up to the Planck scale- the "time-like" direction is extended to the
complex variable tc = tr + i ti , Im t c [i ti , tr] , the function ƒ(t) being analytic within the limits

of the KMS field and continuous on the boundaries. What happens on the ß = 0 limit ? Applying the

KMS properties, we find that the time like direction t becomes pure imaginary so that the signature is
Euclidean (++++). Conversely, t is pure real for ß Planck (+++ ). So, according to Tomita's

modular theory [11], the KMS condition, when applied to the space-time, induces, within the KMS

strip, the existence of the "extended" (holomorphic) automorphisms group :

Hßc Hßc
Mq ßc (Mq ) e M q e

with the ß parameter being formally complex and able to be interpreted as a complex time t and / or

temperature T. It is interesting to remark that in the totally different context of superstrings, J.J.

Atick and E. Witten were the first to propose such an extension of the real temperature towards a

complex domain [4]. Recently, in N=4 supersymmetric string theory, I. Antoniadis, J.P. Deredinger

82
 and C. Kounnas [3] have also suggested to shift the real temperature to imaginary one by

identification with the inverse radius of a compactified Euclidean time on S1, with R = 1 / 2 T.

Consequently, one can introduce a complex temperature in the thermal moduli space, the imaginary
S / T/ U
part coming from the B antisymmetric field under type IIA type IIB
S / T/ U
Heterotic string-string dualities. More precisely, in Antoniadis and al approach, the field

controlling the temperature comes from the product of the real parts of three complex fields :

s= Re S, t= Re T and u= Re U. Within our KMS approach, the imaginary parts of the moduli S, T,

U can be interpreted in term of Euclidean temperature. Indeed, from our point of view, a good reason

to consider the temperature as complex at the Planck scale is that a system in thermodynamical

equilibrium state must be considered as subject to the KMS condition [24].

tel-00001503, version 1 - 24 Jul 2002

Now, let's step forward a more algebraic comprehension of KMS state, in terms of von Neumann

algebras.

5.1.4 KMS state in terms of von Neumann algebras

The von Neumann algebras are, naïvely speaking, the non commutative analogs of measure theory.

They have a critical importance in our understanding of non commutativity of space-time around the

Planck scale.

In the KMS state, the only von Neumann algebras involved are what is called "factors", i.e. a special
type of von Neumann algebra, whose the center is reduced to the scalars a . There exists three

types of factors : the type I and type II (in particular here II ) -which are commutative and endowed

with a trace- and the type III, non commutative and traceless. A trace on a factor M is a linear form

such that (AB) = (BA), A,B M . In this case, any measure on M is invariant. When the measure

on M is ill defined (which is the case of type III), the notion of trace vanishes and is replaced by the
one of "weight", which is a linear map from M+ to + = [0, + ]. The type III factors have no definite

trace. They are very important hereafter as far as they are the only one involved in KMS states. We
work here with "III " factors, 0, 1 , characterized by the invariant S(M) 0 .

83
 Rightly, the KMS condition, when applied to the (pre)space-time at the Planck scale, cuts up three

different scales on the (pre)light cone, which can be described by three different types of von

Neumann algebras (or "factors").

5.1.5 From the topological scale to the physical scale of the space-time

(i) the topological scale (ß = 0, signature {++++}) : this initial "topological" scale correspond to

the imaginary vertex of the light cone, i.e. a zero-size gravitational instanton. All the measures
performed on the Euclidean metric being -equivalent up to infinity, the system is ergodic. As shown

by A. Connes, any ergodic flow for an invariant measure in the Lebesgue measure class gives a
unique type II hyperfinite factor [11]. This strongly suggests that the singular 0-scale should be
tel-00001503, version 1 - 24 Jul 2002

described by a type II factor, endowed with a hyperfinite trace noted Tr . By hyperfinite, we simply
0,1
mean that the trace of the II factor is not finite. We call MTop such a "topological" factor, which is an

infinite tensor product of matrices algebra (ITPFI) of the RO,1 Araki-Woods type [11]. Now, the
0,1 1
initial state on MTop , corresponding in ex. (2.1) to the divergent values of the dilaton field 2 , is
g

given by :
Tr (e ßH MTop
0,1
0,1 )
(MTop ) (58)
Tr (e ßH )

and, considering the hyperfinite characteristic of the trace Tr , we have equivalently :

0,1 ßH 0 ,1
( MTop ) = Tr (e MTop e ßH ) (59)

0,1
where ( MTop ) represents a very special type of "current", that we propose to call 'trace current" T .

Clearly, the invariant hyperfinite II trace current T is a pure topological amplitude [19-37] and, as
0,1
such, "propagates" in imaginary time from zero to infinity. In this sense, ( MTop ) can be seen as a

"zero topological cycle" which represents an intrinsic "Euclidean dynamic" controlling the blow up

of the space-time Initial Singularity [6].

(ii) the quantum scale (0 < ß < Planck , signature {+++±}) : we reach the KMS domain [24].

Considering the quantum fluctuations of g , there is no more invariant measure on the non

84
 commutative metric Therefore, according to von Neumann algebra theory, the "good factor"

addressing those constraints is uniquely a non commutative traceless algebra, i.e. a type III factor [9]
(the only one able to be involved in KMS state). More precisely, it is a type III that we call Mq , with

the period 0, 1 . Important, it has been demonstrated that any type III factor can be

canonically decomposed into the following way [9] :

III II *+ (60)
* + (dual of ) acting periodically on the "II factor". Then the ß-dependant periodicity of the action
0,1
* + on MTop
of takes the form :
*
0,1 0,1
Mq = MTop ß MTop ß S1 (61)
2
The relation between and ß is such that , so that when ß , we get 0 (the
tel-00001503, version 1 - 24 Jul 2002

periodicity is suppressed). Now, the theory being given on the infinite Hilbert space ( ) =
*
L2 ß , Mq becomes :
*
0,1
Mq = MTop L2 ß (62)
*
The type I factor L2 ß yields the modular flow of (periodic) evolution of the system. So,

0,1
the KMS type III factor Mq connects the "topological" type II factor MTop with the "physical"

type I factor MPhys :

0,1
Mq = MTop MPhys (63)
0,1
In terms of "flows", Equ. (63) connects the topological flow of weights of MTop with the physical
*
modular flow raised by L2 ß . This furnishes a good image of the unification between

topological and physical states, to be compared to the bicrossproduct (41) Uq(so(4) op

Uq(so(3,1)) unifying Euclidean and Lorentzian q-groups. The quantum flow

ßc H ßc H
ßc (M q flow ) e Mq flowe is constructed in prop. (5.2).

(iii) the physical scale (ß > Planck , signature {+++ }) : this last scale represents the physical part

of the light cone and, consequently, the notion of (Lebesgue) measure is fully defined. Therefore, the

85
 (commutative) algebra involved is endowed with a hyperfinite trace and is given on the infinite Hilbert
space ( ), with = L2( ). Then (L2( )) is a type I factor, indexed by the real group , which
we call MPhys . So, (L2( )) = MPhys and the flow raised by MPhys is simply the (real) time

evolution, given by the modular group :

e iHt MPhys e iHt
t (M Phys ) (64)

In this case (type I factor) all the automorphisms are inner automorphisms. We call "physical flow"
P this evolution flow in real time. Of course, t (M Phys ) is simply giving the usual algebra of
ß 0

observables [12].

At present, we shall evidence that the KMS state "unifies" the physical flow and the topological
tel-00001503, version 1 - 24 Jul 2002

current.

Proposition 5.2 At the KMS scale 0 < ß < lPlanck , the two automorphisms groups t (M Phys ) and
0 ,1 0,1
ß (MTop ) are coupled up to Planck scale within a unique III factor of the form Mq = MTop
*
L2 ß . The corresponding extended one (complex) automorphisms group describing the

quantum evolution is
Hßc Hßc
Mq ß (Mq ) e Mq e

Mq corresponds to the coupling between the one parameter automorphisms group giving the physical

flow and the automorphisms semi-group giving the topological flow of the system.

Proof The KMS state of the (pre)space-time is yield by the unique III factor given by equ. (60) :

*
0,1
L2 0,1
Mq = MTop ß = MTop MPhys (65)

which represents the KMS "unification" of the topological state and the physical state of the

(pre)space-time at the Planck scale. Now, since it exists an operatorial weight of Mq on its sub-group
0,1 0,1
MTop , the equilibrium state on Mq is given by the state on MTop . We express the state under the

new form constructed in [6] :

86
 ßH 0 ,1
(Mq -state ) = Tr (e MTop e ßH )

0,1
This represents what we have called in (5.1.5) the "trace current" of the "topological factor" MTop .

However, Connes-Takesaki have shown [10] that the flow of weights on a factor is given by the flow
of weights on the associated II factor. For it exists an homomorphism OUT III OUT II such

that the sequence (66) is exact :

M
1 H1 (F) OUTM OUT , (N) 1 (66)
S 0,1 0,1
The multiplicative action of : o S e , s on MTop translates the trace of MTop , which
0,1
generates the flow of weights on MTop and Mq (cf.[10]). So, (Mq -state ) becomes a ß-dependant

automorphism (semi)group :
tel-00001503, version 1 - 24 Jul 2002

ßH 0 ,1
ß (Mq- state ) = e MTop e ßH (67)

Equ. (67) describes the flow of weights [10] of the type III factor Mq. But as pointed in [6], we can

also interpret equ.(67) as a " modular flow in imaginary time" it, dual to the modular flow in real time

given by :
e iHt MPhys e iHt
t (Mq -evolution ) `= , t .

An interpretation of this type has also been proposed (in a different context, however) by Derendinger
and Lucchesi in [13]. Finally, the KMS flow connects the flow of weights ß (M q- state ) to the modular

group t (Mq -evolution ) :

0 ,1
ßc (M q flow ) ß (MTop ) t (M Phys )
( ß it)H
e( ß it) H
e Mq flow

e ßc H Mq ßc H
flow e

which is indexed by the complex time variable ßc. Again, this flow is expressing the unification
between the physical flow t (Mq -evolution ) = t (M Phys ) and the topological flow ß (M q- state ) =
0,1
ß (M Top ) within the unique KMS (or quantum) flow Q given by the automorphisms group of
0 ß P

Mq :

ßc (M q flow ) = ß (M q- state ) t (Mq -evolution )

87
 The (pre)space-time KMS strip has zero as infimum and the Planck scale as supremum. So between

those bounds, the Euclidean topological flow and the Lorentzian physical flow are unified in a natural
way within the holomorphic "quantum flow" Q ßc (M q flow ) e ßc H Mq flowe ßc H .
0 ß P

0,1
Another way to verify the coupling of MPhys and MTop in the unique type III factor lies in the

Conne's invariant

AUT M
: OUT M = (68)
INT M
(automorphisms of M quotiented by the inner automorphisms, necessarily present in the non
commutative case). This M invariant represents an ergodic flow {W(M) , W } where W is a one

parameter group of transformations - i.e. a flow - which admits a description in terms of class of
tel-00001503, version 1 - 24 Jul 2002

weights and whose the natural parameter is *. We consider now the type III factor Mq of equ.(61).
+

Starting from equ. (68), we can construct the extension Ext (noted T) of OUT Mq by INT Mq in AUT

Mq :

AUT Mq OUT Mq T INT Mq (69)

with {x,y} OUT Mq and {x',y'} INTMq. The inner automorphisms group INTMq is a normal
sub-group of AUT Mq . Considering two weights et of Mq, and applying the Radon-Nikodym
theorem [10], it exists a unitary of Mq such that t (x) ut t (x)ut , with ut (D ; D )t and

t (x) INT Mq for a certain class of modular automorphisms. Considering the fact that under the
0,1 *+ all the modular
trace of the factor II involved in the crossproduct Mq = MTop

automorphisms are inner automorphisms, we restrict INTMq to the sub-group of the modular
automorphisms, which we call INTmod Mq. Then we look for the image of the inner modular group in
OUTMq. Within a certain cohomology class {K}, the group t (x) is given by INTmodMq, whereas
the non-unitary transformations ß (x) are given by OUT Mq. We get then for the "physical" flow :

(x) = e iHt Mq e iHt

t INTmod Mq (70)

whereas the "topological" flow of weights of Mq is given by :

ßH
ß (x) =e Mq e ßH OUT Mq (71)

88
 and the extension AUT Mq OUT Mq T INTmod Mq yields :

(t T ß ) t (x) T ß (x) (72)

Within the general group of extensions {Ext}, we get the trivial holomorphic sub-group :
( ß it )H Hßc Hßc
Mq e( ß it)H
ß it (Mq ) e = ßc (Mq ) e Mq e or

which corresponds to the KMS state and"unifies" within the unique extended form ßC (Mq) the

physical flow t (x) and the topological current ß (x). Clearly, we get ßC (Mq) OUT Mq T
INTmod Mq. Again we find : ßc (M q flow ) = ß (M q- state ) t (Mq -evolution ) qed

Now, let's get over the last step. Our aim is to explain the transition from the topological state to the

physical state (TP transition) of the space-time. We shall cope with this problem following two
tel-00001503, version 1 - 24 Jul 2002

different ways :

(i) we conjecture that such a transition could be related to the N=2 supergravity breaking beyond

Planck ;

(ii) likewise, TP transition could be explained in terms of "decoupling", beyond the Planck scale,
0,1
between the (Euclidean) "topological current" (raised by MTop ) and the (Lorentzian) physical flow

(yield by MPhys ).

5.2 TP transition, supersymmetry breaking and flows decoupling

First of all, let's put in evidence the link between KMS state and supersymmetry. To do this, we

propose hereafter a relevant example able to be seen as a good toy model expliciting the deep

correspondence between thermal states, supersymmetry and extended space-time (i.e. extension of the

time-like direction in the complex plane).

Example 5.2.1 thermal states, supersymmetry and KMS condition

In the following, we shall focus on some important results recently obtained by J.P. Derendinger and

C. Lucchesi [13]. Interestingly, it has been demonstrated that thermal supersymmetry (as opposed to

T=0 supersymmetry) must be considered in the context of thermal (i.e. KMS) superspace. We remark

89
 here that the authors apply the KMS condition to the thermal superspace (i.e. the thermal

supersymmetric space) in a general setting. In our own approach, as suggested in ref. 6 and in the

present paper as well, we apply the KMS condition to the thermal (pre) space-time at the Planck scale.

Considering that in the standard “hot big-bang” theory the (pre) space-time is generally viewed as

supersymmetric, such an identification is natural. Namely, the authors have established that the

thermal supersymmetry parameters must be both time dependant and (anti)periodic in imaginary time

on the interval [0, ß], where ß = 1/T. In other words, focusing on field representations of the thermal

super-Poincaré algebra and on chiral supermultiplet, one can straightfully see that thermal superfields

are characterized by their time/ temperature periodicity properties. To explicit this, let's simply recall

that at zero temperature, supersymmetry can heuristically be represented as a set of "generalized

tel-00001503, version 1 - 24 Jul 2002

translations", including Grassmann variables that are translated by the supersymmetry generators.

Therefore, a "point" X in superspace has coordinates

X (x , , ) (73)

where and are the usual Grassmannian objets. Since at zero temperature the parameters of

supersymmetry transformations are constant, the zero-temperature superspace coordinates are also

space-time constants. In fact, at T=0, the (anticommuting) Grassmann coordinates simply turn

bosonic commutation relations into fermionic anticommutators and conversely. Now, what happens at

finite temperature (i.e. the case of primordial universe investigated here)? As a matter of fact, the

situation is not so simple, because fermion and boson statistics involve, in addition, the appropriate

statistical weight in field theory Green's functions. In such a context, as pointed in refs [13] and [28],

it is natural to require that the variables which are translated by the effect of thermal supersymmetry

transformation express the same properties as the thermal supersymmetry parameters. Therefore, the

construction of thermal supersymmetry requires that the Grassmann variables get promoted to be time-
dependant and (anti)periodic in imaginary time. To see this, let's precise that the thermal average ... ß

of a field operator O is, as usual, given by

90
 1 ßH
O ß Tr e O (74)
Z( ß)

with the lowest energy state being E0 0 , so that we have on the zero temperature limit :

O ß 0O 0

Now, at finite temperature, the Green's functions are necessarily subject to periodicity constraints in

imaginary time. However, as showed in [6], those constraints are exactly defining the KMS

condition. To verify this important point, we now review those conditions for bosonic and fermionic
fields. Let's first begin with a free scalar (i.e. bosonic) field at x = (t, x) whose evolution is such

that :
tel-00001503, version 1 - 24 Jul 2002

eiHt (0, x)e iHt

(x) (75)

where the time coordinate t is allowed to be complex. Then the n-point thermal Green's function GnC

of the system is :

GnC (x1 ,..., x n ) T C (x1 )... (x n ) ß

(76)

TC being the path-ordering operator and ... the canonical thermal average. Then the thermal path-

ordered propagator takes the form ( Dc being the thermal propagator of the theory) :

DC (x1, x 2 ) C (t1 t 2 )DC (x1 , x2 ) C (t 2 t1 )DC (x1 , x2 )... (xn ) (77)

where c is the path Heaviside function. Then the thermal bosonic two-point functions DC , DC are

defined as :

DC (x1, x 2 ) (x1 ) (x2 ) ß

(78)
DC (x1, x 2 ) (x 2 ) (x1 ) ß

ßH
At this stage, as proposed in [6], the Boltzman weight e can be seen as an evolution operator in

Euclidean time, so that after a translation in imaginary time we get the formula (79) :

91
 ßH
e (t, x)e ßH (t iß,x) (79)

which is exactly the KMS condition formulated in equ. (53). Then DC (x1, x 2 ) in equ.(78) becomes :

1 ßH
DC (x1, x 2 ) Tr e (x1 ) (x 2 ) (80)
Z ( ß)

Likewise for DC (x1, x 2 ) . So using the cyclicity of the thermal trace and the notion of evolution in

Euclidean time it, one can construct the "bosonic KMS condition" [13-28]. Interestingly, such a
condition relates DC and DC by a translation in Euclidean (imaginary) time :

DC (t1; x1 ,t 2 ;x 2 ) DC (t1 iß; x1 ,t2 ;x 2 ) (81)

tel-00001503, version 1 - 24 Jul 2002

Of course the same construction holds for fermions. Indeed, defining the fermionic two-point function
SCab and SCab , (with a,b = 1 ...4 for Dirac four-components spinors) as

SCab (x1 ,x 2 ) a (x1 ) b (x2 ) ß

SCab (x1 ,x 2 ) b (x2 ) a (x1 ) ß (82)

and as in the bosonic case, the fermionic KMS condition takes the form :

SCab (t1 ;x1,t 2 ;x 2 ) SCab (t1 iß;x1 ,t2 ;x2 ) (83)

which differs from the bosonic condition only by a relative sign. From the structure of equ.(81) and

equ.(83), we deduce that when the temperature of the supersymmetric system (here the (pre)space-

time) is not zero, then bosonic fields are periodic in imaginary time whereas fermionic fields are

antiperiodic. Let's remark that supersymmetry algebra is not sensible to this periodicity-antiperiodicity

distinction. If (as pointed in [13-28]) it is true that the supersymmetry breaking is "encoded" in this

difference, the breaking becomes effective only when the KMS state is cancelled. For this reason, as

demonstrated in the hereabove refs., the KMS condition must be applied to the superfields of the

theory. In [13-28], the superfields are superspace expansions which contain as components the

bosonic and fermionic degrees of freedom of supermultiplets. Such superfields are usually formulated

92
 ˙
using two-component Weyl spinors and , related to Dirac spinors through a ˙ . Then

the KMS condition for Dirac spinors can be extended to Weyl spinors and, in the same way, to

Majorana spinors. The fermionic KMS condition for majorana spinors takes the form :
˙ ˙
SC (x1 , x2 ) (x1 ) (x2 )
ß
(84)
˙ ˙
SC (x1 , x2 ) (x 2 ) (x1 )
ß

Now, one can realize that imposing the KMS condition to superfields components implies that one

must also allow Grassmann parameters to depend on imaginary time. In fact, in the context of
tel-00001503, version 1 - 24 Jul 2002

supersymmetry, the main question is the following : under thermal constraints, how do we

successfully achieve the transformation of periodic bosons into antiperiodic fermions and vice-versa?

The answer, developed in [13-28], consists in constructing the thermal superspace, i.e. in introducing

time dependant and antiperiodic space-time coordinates. Henceforth, a point in thermal superspace

has "KMS coordinates", given by a new set of Grassmannian variables:

Xˆ (x , ˆ (t), ˆ (t))
˙
(85)

where the symbol "^" denotes the thermal quantities and ˆ (t), ˆ (t) are subject to the antiperiodicity
˙

conditions

ˆ (t ˆ (t) , ˆ ˙ (t ˆ ˙ (t)
iß) iß) (86)

Consequently, the condition (86) induces a temperature-dependant constraint on the time-dependant

superspace Grassmann coordinates ˆ (t) and ˆ (t) . From equ. (85), we finally observe that the
˙

KMS condition must be applied to the space-time metric itself, as formulated in §(5.1.2). Among the

consequences, we are therefore induced to consider that the time-like direction must be extended in the

complex plan (see §(5.1.3)).

93
 Now, what does these results mean in the context of our research? As a matter of fact, Derendinger

and Lucchesi have clearly confirmed that there exists a deep relation between (thermal)

supersymmetry and KMS condition. This relation is implemented at the level of thermal Grasmann

coordinates, because of the (anti)periodicity conditions given by equ.(86). Indeed, it has been proved

by the authors that the only way to preserve supersymmetry in the thermal context is to consider that

the space-time metric itself must be subject to the KMS condition. Otherwise, the periodic bosons and

antiperiodic fermions could not be related by supersymmetry. Now, let's put this simple question :

what happens when the KMS state collapses? The analysis of the "KMS Grassmann coordinates", in

particular the equ.(86), clearly show that supersymmetry cannot be implemented without applying the

KMS condition to space-time coordinates. The reason of this is that when the space-time system is not
tel-00001503, version 1 - 24 Jul 2002

subject to the KMS state (e.g. non-equilibrium state), a point X of superspace is endowed again with

the usual Grassmann coordinates

X (x , , )

This is equivalent to T=0 supersymmetry, for which the parameters of transformation (i.e. the

Grassmannians and ) are space-time constants. But rightly, the main result of refs [13-28]

establishes without ambiguity that at finite temperature, one cannot make use of constant parameters in

supersymmetry transformations rules. The supersymmetry parameters must be time dependant

variables, (anti)periodic in imaginary time. So, in a natural way, the thermal Grassmann coordinates
Xˆ (x , ˆ (t), ˆ (t)) must be "translated" in imaginary time and are consequently subject to the
˙

antiperiodicity conditions ˆ (t iß) ˆ (t) and ˆ ˙ (t iß) ˆ ˙ (t) of equ.(86). Obviously, the

only way to implement such a condition is to consider that globally, the space-time system is in KMS

state at a given scale (i.e. in our case between the scale zero and the Planck scale). Incidentally, the
hereabove approach can be seen as a confirmation that the ß 0 limit is topological. As a matter of

fact, the ß 0 limit of equ.(79) is given by the scalar field (x), which, by construction, is a

topological configuration marking the origin of the imaginary time direction of the theory.

94
 From hereabove we can now conclude that (thermal) supersymmetry and KMS states are linked in

such a manner that the breaking of the KMS state beyond the Planck scale should induce the breaking

of supersymmetry at the same scale. Let's go further in the exploration of such a breaking. In a very

stimulating way, Derendinger and Lucchesi have emphasized the fact that the thermal field boundary

conditions characterizing KMS state carry information that is of global nature in space-time. By

construction, the supersymmetry algebra being a local structure is insensitive to this global

information. What is the nature of this "global information"? Indeed, the translation of Grassmannians

variables into imaginary (topological) time clearly indicates that the natural state of such a global

information is a topological state, correctly described by topological field theory (which is precisely a

non local theory). More exactly, the boundary conditions characterizing the euclidean time dependence
tel-00001503, version 1 - 24 Jul 2002

of the supersymmetry parameters can be seen as topological invariants. In this perspective,

supersymmetry breaking can then be investigated in terms of cancellation of such topological

invariants. Let's now explore this occurrence.

5.2.2 Supersymmetry and topological invariants

In a famous precursor paper [36], and in some others, E. Witten has clearly put in evidence that if we

want supersymmetry breaking to occur, the various four-manifolds invariants (such that the

Donaldson invariant, the Euler number, the Witten index etc..) must necessarily vanish. The outline of

the argument is that the canceling of the supersymmetry index Tr(-1)F is canceling the zero energy

modes, which consequently breaks the Bose-Fermi pairs [1]. At this stage, if we agree with

supersymmetry theory, a reasonable conclusion is that (N=2) supergravity breaking could be viewed

as related to the canceling of topological configurations. Let's now go further : can supersymmetry
breaking explain the Topological Physical transition? In a certain sense, the answer might be yes.

In fact, since the context of the theory is supergravity N= 2, we may precise the conditions of

topological modes canceling within supersymmetry breaking. So :

95
 Conjecture 5.2.3 On a D = 4 Riemannian (pre)space-time manifold, the N = 2 supergravity

breaking at the Planck scale is related to the canceling of the Euler characteristic and of the topological

mode of the manifold.

Let M be the four dimensional Riemannian N=2 supersymmetric (pre)space-time. The Euler

characteristic of M is

1
(M) R R
32 M

We have shown in prop. (3.2) that this invariant is given by Tr( 1)S . Now, according toWitten's

results [36], a discontinuous change of Tr( 1)S is possible, due to the asymptotic behavior of the
tel-00001503, version 1 - 24 Jul 2002

manifold, allowing, for large field strengths, some energy states to "move in from infinity". For

instance, let's consider the potential

2
V( ) (m g )

One can easily observe that arbitrarily small g 0 induces the existence of extra low-energy states at
~ m g which have no counterpart for the pure g = 0 value. Therefore, Tr( 1)F will change

discontinuously from g = 0 to g 0. The same result can be extended toTr( 1)S , when coupling the

instanton radius to g. In this case, we meet again the conclusions of (ii) in example (2.1) (i.e. the

instanton configuration is cancelled for large values of g).

Next, we have seen (5.1.2) that the (pre)space-time should be in KMS state at Planck , so that the time

like direction t becomes holomorphic within the KMS strip. The metric configuration is described by

the symmetric homogeneous space

SO(3,1) SO(4)
h = (87)
SO(3)

SO(3) being diagonally embedded in SO(3, 1) SO(4) [6]. To h corresponds, at the level of the
3, 1 4
underlying spaces involved, the topological quotient space top = from which, assuming
SO(3)

that the compact part of the 3-geometry is a sphere S3 , the topology of the five dimensional

(pre)space-time can be viewed as isomorphic to S 3 ( being the space-like direction and

96
 3, 1 4
the time-like direction, out of the orbit of the action of SO(3) on ). We then meet again

the equivalent form S3 of the five dimensional manifold described in (2.1). The point is
that allows us to define the boundary conditions of the (pre)space-time 5-geometry 5. Therefore,

the form of the 5D metric is [6] :

2 2 2 d 2
ds a( ) d (3) dt 2 (88)
g2
2
where the axion term is a = ƒ( , t), the 3-geometry d (3) (x, y, z). Then, as showed in (2.1), on

the (infrared) strong coupling bound (i.e. the Planck scale, in respect of the (ultraviolet) zero scale),
1
condition (i) imply 2 0 and the direction of 5 is cancelled. So, we get a dimensional reduction
g

(D=4 D=3) of the compact Riemannian 4-geometry embedded in the five dimensional (pre)space-
tel-00001503, version 1 - 24 Jul 2002

time manifold 5. We have for the metric :

w compactification on S 1 0 Dimensional reduction
(++++ ) (+++(0) +++

Obviously, the boundary condition ß gives rise to the asymptotic cancellation of the 4 Euler

characteristic:

1
(M) R R = Tr( 1)F = 0 (89)
32 M

Likewise, the asymptotic flatness condition [6] for ß gives R R 0, which implies that

the dimension D of the asymptotic manifold must be odd, so that, again, we get = 0 for the (3+1)

usual space-time. Therefore, according to ref. [36], the supersymmetry is broken. Simultaneously, the
topological state, given by even values of the Euler number vanishes, implying the "TP transition" :
TP transition
Topological mode Physical mode .

To finish, we meet a novel problem : could TP transition be, in some way, related to the breaking of

the KMS state described in (5.1)? This question is discussed in the last paragraph.

97
 5.2.4 TP transition and decoupling between topological flow and physical flow

In answer to the hereabove question, we now conjecture that for ß Planck , i.e. at the (semi-classical)

scale where supersymmetry is being broken, the topological flow (evolution in imaginary time)

corresponding to the zero topological pole of the theory is decoupled from the physical flow

(evolution in real time).

According to most of the models, supergravity is considered as broken for scales greater than the

Planck scale [25]. But thermal supersymmetry breaking is also closely connected to the cancellation of

the thermodynamical equilibrium state [27-28]. Indeed, as already pointed in this paper, I. Antoniadis

and al have recently demonstrated that a five-dimensional (N = 4) supersymmetry can effectively be

tel-00001503, version 1 - 24 Jul 2002

described by a four-dimensional theory in which supersymmetry is spontaneously broken by finite

thermal effects [3]. In a similar way, Derendinger and Lucchiesi have outlined the fact that thermal

supersymmetry is a global (i.e. topological) property of the space-time in KMS state [13-28]. In this

context, the cancellation of the thermodynamical equilibrium state necessarily cancels the KMS state

and, consequently, breaks the supersymmetry [6]. This scenario is typically the one characterizing our

setting. As a matter of fact, the five dimensional supersymmetric theory evoked hereabove correponds

to the five dimensional supersymmetric (pre)space-time in KMS state. Then the (thermal)
supersymmetry breaking is characterized in Kounnas approach, by a D=5 D=4 dimensional

reduction, which corresponds exactly, in our case, to the decoupling between imaginary time and real

time. Indeed, we could have :

3
3 ss breaking ( four dimensional topological space - tim
( five dimensional KMS space - time) 3
( four dimensional physical space - time

So, supersymmetry breaking, KMS breaking and topological physical transition appear as deeply

connected. To see this, let's come back to the KMS state. We call "KMS breaking" the end of the

KMS state beyond the Planck's scale. The observed cancellation of the thermodynamical equilibrium

beyond the Planck scale (which gives the inflationary phase and the beginning of the cosmological

expansion) is inducing KMS breaking (see ex. (5.2.1)). Such a breaking must be seen as the inverse

of the KMS coupling between equilibrium state and physical evolution of the system. And logically,

98
 such a breaking should bring about the transition from the pure (non perturbative) topological phase

around the Initial Singularity to the physical phase of the universe we can observe to day.

Now, here is our conjecture :

Conjecture 5.2.5 In the infrared ß Planck scale, KMS breaking is inducing the decoupling

between the topological flow and the physical flow of the theory.

Considering the KMS state of space-time at the Planck scale, the KMS flow, as shown in prop.

(5.2), is :

= e ßc H Mqe ßc H
ßc (M q ) = ß (M q- state ) t (Mq -evolution ) (90)
tel-00001503, version 1 - 24 Jul 2002

or
( ß it )H
Mq e ( ß it) H
ßc (M q ) =e (91)

*
0,1
L2 0,1
Now, starting from Mq = MTop ß = MTop [L2 (ßS1 )], we can say that ß >

Planck is equivalent to ß , with respect to the scale zero. So, when ß > Planck , the period of the

system is so large that we can consider it as supressed from equ.(62), whereas the circle S1 is
2
decompactified on the straight line R. Moreover, this limit corresponds to 0. So, on this
ß
* *
limit * +. But the suppression of the period is equivalent to the cancellation of the
ß ß

equilibrium state and therefore induces the breaking of the KMS state. To see this, we can write the

"extended" automorphisms group corresponding to the KMS state :

= e ßc H [ MTop
0,1 * +]e ßc H ( ß it )H 0,1 * +] e ( ß it) H
ßc (M q ) =e [ MTop (92)

Then for ß > > *+

Planck , we get so the corresponding weight on Mq is such that .

But, according to Connes-Takesaki [10], the infinite dominant weight on Mq is dual to the hyperfinite
0,1
trace on MTop . Therefore, the image of the "flow of infinite weights" on Mq becomes, under the

ergodic action of *+ :

ßH 0 ,1
(Mq- state ) = Tr (e MTop e ßH ) (93)

99
 0,1
where we meet again the topological "trace current" T of MTop , independent of ß. But the

independence of T with respect to ß implies in the same way that T is also independent of * + on
0,1
* + must be decoupled of MTop
this limit. So, , which means that the modular evolution group
it it
t (M Phys) MPhys is itself decoupled from the crossed product (65). Moreover, since the

hyperfinite trace (93) is independent of ß, we are left with the "topological" state :

0,1
(Mq- state ) Tr ( MTop )

which is equivalent to say that the only value of ß contributing to equ. (79) is ß = 0. So, on this
boundary, (see equ.(64)), ßc (M q ) is reduced to the real pole, so that :
tel-00001503, version 1 - 24 Jul 2002

e iHt Mq -evolutione iHt

ßc (M q ) t (Mq -evolution ) `

But of course, in this case Mq, as type III algebra, is also suppressed. This is simply because, on the
0,1 0,1
infinite limit of the action of on MTop , the infinite trace Tr on MTop , dual to the dominant weight

on Mq, is left invariant. Applying a result of [11] on infinite weights, one can find that the infinite
weight on Mq is invariant under the inner automorphisms of Mq. Therefore, is a trace, which

is a sufficient condition to cancel Mq as a III factor. But this is equivalent to say that on this limit, the
0,1
action of is decoupled of MTop . Therefore, the crossed product (65) is broken into its two
0,1
subgroups MTop and (L2( * +)). This is as it should be, since beyond the Planck scale, i.e. at the

classical scale, the KMS state is broken and the measure space on the metric is again well defined, so

that the underlying algebra must be endowed with a trace. Consequently, it cannot be Mq anymore.
So, the new algebra involved should be a type I sub-algebra of Mq. Considering the decomposition
0,1
MTop (L2( * +)), this sub-algebra is necessarily (L2( * +)) = MPhys . Then t (Mq -evolution )

becomes simply :

e iHt MPhys e iHt

t (M Phys ) (94)

This corresponds to the usual modular group giving the physical evolution of the space-time. So the
0,1
product MTop MPhys shrinks onto MPhys so that we finally get Mq = MPhys in the infrared.
ß Planck

100
 In the same way, applying the result of prop. 5.2, we see that the breaking of KMS state implies
AUT Mq OUT Mq T INTmod Mq

reduces to the well known case of a factor I, where all the automorphisms of the algebra are inner

automorphisms:

AUT Mq INT Mq
So obviously, this transition causes the decoupling between OUTMq and INTmod Mq , i.e. between
the topological current ß (x) and the physical flow t (x) .

As a result of prop.(5.2.5), we finally can conclude that the breaking of the KMS state beyond the
e iHt MPhys e iHt
Planck scale induces the decoupling between the physical flow t (M Phys ) and the
tel-00001503, version 1 - 24 Jul 2002

zero topological current

= Tr (e - ß H MTop
0,1 0,1
ß ( MTop ) eßH ):

ßH
e ßH )
0,1
Tr (e MTop
KMS breaking
= e ßc H MTop
0,1 ßc H
ßc (M q ) e (95)
iHt iHt
e MPhyse

0,1
At the level of the von Neumann algebras, starting from the KMS algebra Mq = MTop MPhys ,
0,1
the KMS breaking can be seen as the decoupling between MTop and MPhys . This decoupling

describes the transition from the topological phase (zero scale) to the physical phase ( beyond the

Planck scale).

6. CONCLUSION

Even though certain of the hereabove results might seem mysterious, their interest is to outline,

through quantum groups theory and non commutative geometry, a possible phase transition from the

topological zero scale to the physical Planck scale. We describe with more details in a forthcoming

paper the unexpected "algebraic blow up" of the topological initial singularity. At this stage, we

propose to draw the following main ideas :

101
 (i) the metric, onto the zero scale, might be considered as Euclidean (++++) i.e. topological ;

(ii) the Initial Singularity of space-time could be understood as a 0-size singular gravitational

instanton;

(iii) From (i) and (ii), we suggest the existence of a deep symmetry, of the duality type (i - duality),

between physical state (Planck scale) and topological state (zero scale).

Then the possible resolution of the initial singularity in the framework of topological theory allows us

to envisage the existence, before the Planck scale, of a purely topological first phase of expansion of

space-time, parameterized by the growth of the dimension of moduli space dimM and described by
tel-00001503, version 1 - 24 Jul 2002

the Euclidean "pseudo-dynamic" :

= e - ß H MTop
0,1 0,1
ß ( MTop ) eßH

So, the chain of events able to explain the transition from the zero topological phase to the physical

phase of the space-time might be the following :

Supersymmetry breaking thermodynamical equilibrium breaking KMS state breaking

imaginary time / real time decoupling topological state / physical state decoupling

In terms of C*-algebras, the hereabove transformations are given by :

Topological flow 0,1 ßH 0,1 ßH
KMS flow Q
T ß( MTop ) e MTop e
ßc H ß 0
II *
+
0 ß P
ßc ( Mq ) e Mq e ßc H Physical flow
e iHt iHt
P t( MPhys ) MPhys e
ß 0

In a forthcoming article, we push forward the idea following which, that at 0 scale, the Lorentzian

dynamic is replaced by an intrinsic "Euclidean dynamic". A first path to follow would be to investigate

the zero limit of the The Euclidean dynamic engendered by the non-stellar automorphisms of the
0,1
algebra MTop implies, following the results of [6], a "spectral increase" in the diameters of the space

of states d( ) in Euclidean time (dual to the space of observables in Lorentzian time). This
0,1
Euclidean pseudo-dynamic, linked with semi-group automorphisms ß ( MTop ) is described in a natural

way by the flow of weights (in the Connes-Takesaki [9] sense) of algebra Mq ; we suggest equally

102
 (ii) that the Euclidean modular flow representing the evolution of a system in imaginary time can be

associated with an increase in the spectral distance separating the states of the system. Finally, it has

been proposed by one of us [7] that the Euclidean dynamic raised above results from the existence of
~
the topological amplitude yield by the topological charge Q = d 4 x Tr R R of the zero size

singular gravitational instanton connected to the (topological) origin of space-time.

Acknowledgements It is a pleasure to acknowledge the help and encouragement received during many discussions

with S.Majid , of the Mathematics Laboratory of the Queen Mary and Westfield College, C. Kounnas, of the Theoretical

Physics Department of the Ecole Normale Supérieure, F.Combes, of the Mathematics Departement of the University of

Orleans, C.M. Marle, of the Mathematics Department of the University of Paris VI and M. Enock, of the Mathematics
tel-00001503, version 1 - 24 Jul 2002

Department of the University of Paris VII.

REFERENCES

1. ALVAREZ-GAUMÈ L. "Supersymmetry and the Atiyah-Singer Theorem".Commun. Math. Phys. 90, (1983)

2. ALVAREZ E. ALVAREZ-GAUMÈ L. LOZANO Y. "An Introduction to T-Duality in String Theory"

Nucl. Phys. B 451 (1995)

3. ANTONIADIS I. DEREDINGER J.P. KOUNNAS, C. "Non-Perturbative Supersymmetry Breaking and Finite

Temperature Instabilities in N=4 Superstrings". hep-th 9908137 (1999)

4. ATICK J.J. WITTEN E. "The Hagedorn Transition and the Number of Degrees of Freedom in String Theory". Nucl.
Phys. B 310 (1988)

5. BACCHAS C.P. BAIN P. GREEN M.B. "Curvature Terms in D-Branes Actions and their M-theory Origin". Hep-
th 990 3210 (1999)

6. BOGDANOFF G. "Fluctuations Quantiques de la Signature de la Metrique à l'Echelle de Planck" Th. Doctorat Univ.
de Bourgogne (1999)

7. BOGDANOFF I. "Topological Theory of Inertia" CERN Preprint ext 2000-001

8. CAROW-WATUMARA U. SCHLIEKER M. SCHOLL M. WATUMARA S. "Tensor Representation of the

Quantum Group SLq(2, C) and Quantum Minkowski Space". Z Phys. C 48 (1990)

9. CONNES A. TAKESAKI M. "Flot des Poids sur les Facteurs de Type III". CR Acad. Sci. Paris 278, Série A,
(1974)

10. CONNES A. TAKESAKI M. "The flow of weights on factors of type III" Tohoku Math. Journ. 29 (1977)

103
 11. CONNES A "Non Commutative Geometry ". Academic Press (1994)

12. CONNES A. ROVELLI C. "von Neumann Algebra Automorphisms and Time Thermodynamics Relation in General
Covariant Quantum Theories". gr-qc 9406019 (1994)

13 DEREDINGER J.P. LUCCHESI C. "Realizations of Thermal Supersymmetry" Hep-Ph 9807403

14. DONALSON S.K. "Polynomial Invariants for Smooth Four Manifolds". Topology 29, 3 (1990)

15. DONALSON S.K. KRONHEIMER P.B. "The Geometry of Four Manifolds". Oxford Science Publ. Oxford
University Press (1990)

16. DRINFELD V.G. "Quantum Groups". Proc. of the Int. Congress of Maths. Berkeley (1986)

17. DRINFELD V.G. "Quasi-Hopf Algebras". Leningrad Math. J. 1 (1990)

18. FLOER A. "An Instanton Invariant for 3-Manifolds". Commun. Math. Phys. 118, (1995)

19. FRÉ P. SORIANI P. "The N=2 Wonderland. From Calabi-Yau Manifolds to Topological Field Theory". World
Scientific Publ. (1995)
tel-00001503, version 1 - 24 Jul 2002

20. GETZLER E. "A Short Proof of the Local Atiyah-Singer Index Theorem". Topology 25 1 (1986)

21. GILKEY P. "The Index Theorem and the Heat Equation" Math. Lecture Ser. 4 Publish or Perish, Boston (1974)

22. GILKEY P. "Invariance Theory, the Heat Equation and the Atiyah-Singer Index Theorem". Math. Lecture Ser. 4
Publish or Perish, Wilmington (1984)

23. GREGORI A. KOUNNAS C. PETROPOULOS P.M. Nuclear Phys. B 537 (1999)

24. HAAG. R. HUGENHOLZ N. WINNINK M. "On the Equilibrium States in Quantum Statistical Mechanics".
Commun. Math. Phys. 5, (1967)

25. KAKU M. "Strings, Conformal fields and M-theory". Springer (2000)

26. KLIMCIK C. SEVERA P. "Poisson Lie T-Duality and Loop groups of Drinfeld Doubles". Phys. Let. 372 (1996)

27. KOUNNAS C. "Universal Thermal Instabilities and the High-Temperature Phase of the N=4 Superstrings". hep-th
9902072 (1999)

28. LUCCHESI C. "Thermal Supersymmetry in Thermal Superspace" hep-ph 9808435

29. MAJID S. "Foundations of Quantum Groups" Cambridge University Press (1995)

30. MAJID S. "q-Euclidean Space and Quantum Wick Rotation by Twisting". J. Math. Phys. 35, N° 9 (1994)

31. NASH C. "Differential Topology and Quantum field Theory" . Academic Press (1996)

32. SEIBERG N. WITTEN E. "Electric-Magnetic Duality, Monopole Condensation and Confinement in N=2
Supersymmetric Yang-Mills Theory". Nucl. Phys. B 426 (1994)

33. SEIBERG N. WITTEN E. "Gauge Dynamics and Compactification to Three Dimensions". hep-th 9607163 (1996)

34. WEINBERG S. "Gauge and Global Symmetries at High Temperature". Phys. Rev. D 9 12 (1974)

104
 35. WESS J. ZUMINO B. "Covariant Differential Calculus on the Quantum Hyperplane". Pro. SuppL Nucl. Phys. B,
18B, (1990)

36. WITTEN E. "Constraints on Supersymmetric Breaking". Nucl. Phys. B 202 (1982)

37. WITTEN E. "Topological Quantum Field Theory". Com. Math. Phys. 117 (1988)

38. WITTEN E. "Quantum Field Theory and the Jones Polynomial" Com. Math. Phys. 121 (1989)

39. WITTEN E. "Lectures on Quantum Field Theory". Quantum Fields and Strings Am. Math. Soc. 1-2 (1991)

40. WITTEN E. "Supersymmetric Yang-Mills Theory on a four Manifold". J. Math. Phys. Vol 35 N°10 (1994)

41. WITTEN E. "Small Instantons in String Theory". hep-th - 9511030 (1995)

42. WITTEN E. "Anti de Sitter Space and Holography" hep-th 9802150 (1998)
tel-00001503, version 1 - 24 Jul 2002

105
 Annals of Physics 296 90-97 (2002)

Spacetime Metric and the KMS Condition at the Planck Scale

Grichka Bogdanoff and Igor Bogdanoff

Mathématical Physics Laboratory CNRS UMR 5029 Bourgogne University

U.F.R. Sciences et Techniques
B.P. 47870
F-21078 Dijon Cedex FRANCE

Abstract

Considering the expected thermal equilibrium chcaracterising the physices at the Planck scale, it is here stated, for the
tel-00001503, version 1 - 24 Jul 2002

first time, that, as a global system, the space-time at the Planck scale must be considered as subject to the Kubo-

Martin-Schwinger (KMS) condition. Consequently, in the interior of the KMS strip, i.e. from the scale ß = 0 to the

scale ß = Planck, the fourth coordinate g44 must be considered as complex, the two real poles being ß = 0 and ß =

Planck. This means that within the limits of the KMS strip, the Lorentzian and the Euclidean metric are in a

"quantum superposition state" (or coupled), this entailing a "unification" (or coupling) between the topological

(Euclidean) and the physical (Lorentzian) states of space-time.

1. INTRODUCTION

Many descriptions have been recently proposed in the literature regarding the physical state of the

universe at the vicinity of the Planck scale. Non commutative geometry, string theory, supergravity

or quantum gravity have contributed, independently of each others, to establish on solid basis the data

of a "transition phase" in the physical content and the geometric structures of the (pre)spacetime at

such a scale. But what is the nature of this dramatic change? In the present paper, we propose a novel

approach, based on one of the most natural and realistic physical condition predicted by the Standard

Model for the (pre)universe. In agreement with some well-established results [1-2-3] (recalled in [4] )

and, more recently, the approach of C. Kounnas and al [5-6], we argue that at the Planck scale, the

106
 "spacetime system" is in a thermodynamical equilibrium state [2]. This notion of equilibrium state at

the Planck scale has been recently stated with some new interesting arguments in [7-8-9]. As a

consequence, according to [10], we suggest hereafter that the (pre)universe should be considered as

subject to the Kubo-Martin-Schwinger (KMS) condition [11] at such a scale. Surprisingly, the well

known KMS and modular theories [12] have never been applied to the study of the metric properties

in the context of quantum cosmology ; however, the KMS condition might have dramatic

consequences onto Planck scale physics. For which reasons? Because, when applied to quantum

spacetime, the KMS statistics are such that , within the limits of the "KMS strip" (i.e. between the

scale zero and the Planck scale), the time like direction of the system should be considered as
complex : t tc tr iti . G.Bogdanoff has showed in [10] that at the scale zero, the theory is
tel-00001503, version 1 - 24 Jul 2002

projected onto the pure imaginary boundary t tc iti of the KMS strip. Namely, there exists,

around ß 0, a non-trivial topological limit of quantum field theory, dual to the usual topological

limit associated with ß in the partition function (2). Such a topological state of the (pre)spacetime

can be described by the topological invariant

Z = Tr (-1)s (1)
ß 0

given by the ß 0 limit of the partition function (2). S is here the instanton number of the theory. It

has been demonstrated [10] that this topological index is isomorphic to the first Donaldson Invariant
[13]. This suggests that at zero scale, the observables O i must be replaced by the homology cycles
( k)
Hi mod in the moduli space of gravitational instantons [14]. We get then a deep correspondence

-a symmetry of duality- [15-16], between physical theory and topological field theory. Conversely,
on the (classical) infrared limit ß Planck , the imaginary component cancels and the time like

direction becomes pure real t tc t r So, within the limits of the KMS strip (i.e. for 0 < ß <

Planck ), the Lorentzian and the Euclidean metric are in a "quantum superposition state" (or coupled).

Such a superposition state, described in detail in [10], is entailing a "unification" (or coupling)

between the topological (Euclidean) and the physical (Lorentzian) states of spacetime. The
(Lorentzian / Euclidean) states of the metric g are given by the partition function

107
 Z = Tr (-1)s e ßH (2)

1
with ß = as usual and n being the "metric number" of the theory. To avoid any difficulty of
kT
interpretation, let's remark that the scale parameter ß admits two possible intepretations : (i) either
ß kT can be seen as a real time parameter of a Lorentzian (3+1)-dimensional theory [17] or (ii)
ß = it can be interpreted as the fourth space-like direction of a Riemannian 4-dimensional Euclidean
theory (e.g. ß Planck ) [18]. In the second case, ß is a periodic imaginary time interval [5-6].

Considering the hypothesis of holomorphicity formulated in section 4, we use here the two

interpretations. In such a context, the KMS state of the (pre)spacetime may be considered as a

transition phase from the Euclidean topological phase (ß = 0) to the Lorentzian physical phase,
tel-00001503, version 1 - 24 Jul 2002

beyond the Planck scale [10-19].

The present article is organized as follows. In section 2 we recall that at the Planck scale, the

"spacetime system" should most likely be considered as being in a thermodynamical equilibrium state.

In section 3, we show that, as a natural consequence of this equilibrium state, the spacetime must be

considered as subject to the KMS condition. In section 4, we suggest that, considering the KMS
properties, the time-like direction g44 of the metric should be seen as complex tc = tr + i ti within the

limits of the KMS strip. In section 5, we discuss briefly the transition from imaginary time ti to real

time tr in terms of KMS breaking beyond the Planck scale.

2. THERMODYNAMICAL EQUILIBRIUM OF THE SPACETIME AT THE PLANCK

SCALE

It is well known that at the Planck scale, one must expect a thermodynamical "phase transition",

closely related to (i) the existence of an upper limit in the temperature growth -the Hagedorn

temperature [1]- and (ii) the equilibrium state characterizing, most likely, the (pre)spacetime at such a

scale. In such a context the seminal investigations of K. Huang and S. Weinberg [2], Dolan and R.

Jackiw [3] , then later of several others, have renewed the initial idea of Hagedorn concerning the

108
 existence, at very high temperature, of a limit restricting the growth of states excitation. Already some

time ago, J.J. Atik and E.Witten have shown the existence of a Hagedorn limit around the Planck

scale in string theory [1]. The reason is that, as recently recalled by C. Kounnas [6] in the context of

N = 4 superstrings, at finite temperature, the partition function Z(ß) and the mean energy U(ß)

develop some power pole singularities in ß T since the density of states of a system grows

exponentially with the energy E :

ßE 1
Z( ß) dE (E)e ~
( ß b)(k 1)

1
U( ß) ln Z ~ (k 1) (3)
ß b
tel-00001503, version 1 - 24 Jul 2002

Clearly, equ.(3) exhibits the existence, around the Planck scale, of a critical temperature TH b 1,
where the (pre)spacetime system must be viewed as in a thermodynamical equilibrium state. Indeed,
a(t) being the cosmological scale factor, the global temperature T follows the well-known law :

a(t p )
T (t ) Tp
a(t )
1
Ep C5 2
and around the Planck time, T is reaching the critical limit T p kB 1 1,4.1032 K . In
kB G

fact, it is currently admitted in string theory that, before the inflationary phase, the ratio between the
interaction rate ( ) of the initial fields and the (pre)spacetime expansion (H) is << 1, so that the
H
system can reasonably be considered in equilibrium state. This has been established a long time ago

within some precursor works already quoted in ref. [1,2,3]. Recently the same approach has been

successfully considered by C. Kounnas et al in the superstrings context [5]. However, this natural

notion of equilibrium, when viewed as a global gauge condition, has dramatic consequences regarding

physics at the Planck scale. Among those consequences, unquestionably the most important is that the

(pre)spacetime at the Planck scale must be considered as subject to the famous "KMS condition", a

very special and interesting physical state that we are now going to describe.

109
 3. THE (PRE)SPACETIME IN KMS STATE AT THE PLANCK SCALE

Let's first recall on mathematical basis what an equilibrium state is.

Definition H being an autoadjoint operator and the Hilbert space of a finite system, the
Tr (e ßH A)
equilibrium state of this system is described by the Gibbs condition ( A) and
Tr (e ßH )
satisfy the KMS condition.

This well-known definition has been proposed for the first time in [11]. Now, it is usual (and

natural) to oppose the notion of equilibrium to the one of evolution of a system. In fact, the famous

Tomita-Takesaki modular theory has established that the "intrinsic" dynamic of a quantum system
tel-00001503, version 1 - 24 Jul 2002

corresponds, in a unique manner, to the strongly continuous one parameter * - automorphism group

t of some von Neumann C* - algebra A [12] :

t (A) = ei Ht A e i Ht (4)

This one parameter group describes the time evolution of the observables of the system and

corresponds to the well-known Heisenberg algebra. Nothing mysterious at this stage. However, we

are here brought to find the remarkable discovery of Takesaki and Winnink, connecting the evolution
group t (A) of a system (more precisely the modular group M = it M it) with the equilibrium
ßH
Tr(Ae )
state (A) of this system [11-20]. With this more or less unexpected relation between
Tr(e ßH )

evolution t (A) and equilibrium (A), we now meet the famous "KMS condition". More exactly, in

the frame of quantum statistical mechanic, the KMS condition provides a rigorous mathematical

formulation about the coexistence of different possible equilibrium states at the same given

temperature T.

Let's recall now how such a relation between equilibrium state and evolution of a system is realized
by the KMS condition. It has been clearly established [11] that a state on the C* -algebra A and the

continuous one parameter automorphism group of A at the temperature ß = 1 / k T verify the KMS

110
 condition if, for any pair A, B of the * - sub-algebra of A, t - invariant and of dense norm, it exists

a ƒ(tc) function holomorphic in the strip {tc = t + i ß , Im t c [ 0 , ß ] } such that :

(i) ƒ (t) = (A ( t B)) ,

(ii) ƒ (t + i ß ) = ( t (B)A) , t . (5)

Moreover, a state on the C* -algebra A is separator if the given algebraic representation is a von

Neumann algebra W* endowed with a separator and cyclic vector. The sets

Il = {A A (A * A ) = 0}

and
tel-00001503, version 1 - 24 Jul 2002

Ir = {A A : (AA*) =0}

are forming a left and right ideal in A. For any KMS state, we have I l = I r.

The above definition expresses the bijective relation between equilibrium state, holomorphic state of

the measure parameters and KMS state.

Now, considering the general properties raised by the KMS condition, if we admit that around the

Planck scale, the (pre)spacetime system is in a thermal equilibrium state, then we are also bound to

admit that this system is in a KMS state. Indeed, it has been shown a long time ago [11] that if a state

of a system satisfies the equilibrium condition ([h, t (A)])dt 0, A U , then, satisfies

the KMS condition. So, there is a biunivoque relation between equilibrium state and KMS state. So,
if we admit that around Planck , the (pre)spacetime system is in a thermal equilibrium state, then

according to [11], we are also bound to admit that this system is in a KMS state.

Next, let's push forwards the consequences raised by the holomorphicity of the KMS strip.

111
 4. HOLOMORPHIC TIME FLOW AT THE PLANCK SCALE

As a critical consequence of the KMS condition, we are induced to consider that the time-like
coordinate g00 becomes holomorphic within the limits of the KMS strip. Indeed, as demonstrated in

details in [10-19], within the KMS strip, we necessarily should have :

t tc = tr + i ti (6)

In the same way, the physical (real) temperature should also be considered as complex at the Planck

scale :
T Tc = Tr + iTi (7)

as proposed by Atick and Witten in another context [1]. This unexpected effect is simply due to the

fact that, given a von Neumann algebra W* and two elements A, B of W* , then there exists a
tel-00001503, version 1 - 24 Jul 2002

function ƒ(tc) holomorphic in the strip tc , im tc 0, ß such that :

ƒ(t) = (A ( t B)) and ƒ(tr +i ß ) = ( t (B)A), t (8)

Here, t is the usual time parameter of the 3D theory, like ß kT . So in our case, within the limits
of the KMS strip, i.e. from the scale zero (ß = 0) to the Planck scale (ß = Planck ), the "time-like"

direction of the system must be extended to the complex variable

tc = tr + i ti , Im t c [i ti , tr] (9)

Of course, the holomorphicity of the time like direction of the spacetime is induced in a natural manner

by the fact that in our approach, the thermodynamical system is the spacetime itself. Such a situation

has been investigated in details in [10] in the context of "quantum groups" and non-commutative

geometry.

Indeed, according to Tomita's modular theory [20], the KMS condition, when applied to the

spacetime as a system, allows, within the KMS strip, the existence of an "extended" (holomorphic)

automorphism "group of evolution", which depends, in the classification of factors [12], on a "type
III factor" Mq (a factor is a special type of von Neumann algebra, whose the center is reduced to the

scalars a ). The extended automorphism group has the following form :

Hßc Hßc
Mq ßc (Mq ) e M q e

112
 with the ßc = ßr +i ßi parameter being formally complex and able to be interpreted as a complex time

t and / or temperature T Tc = Tr + iTi . So, the KMS condition suggests the existence at the Planck

scale, of an effective one loop potential coupled, in N = 2 supergravity, to the complex dilaton + axion
1 i
field = 2 i and yielding the dynamical form diag(1 , 1 , 1 , e ) for the metric. The
g
signature of is Lorentziann (i.e. physical) for =± and can become Euclidean (topological)

for = 0. Consequently, the "KMS signature" of the metric is (+++±). This is as it should be since,
considering the quantum fluctuations of g , there is no more invariant measure on the non

commutative metric Therefore, according to von Neumann algebra theory, the "good factor"

addressing those constraints is uniquely a non commutative traceless algebra, i.e. the type III factor
tel-00001503, version 1 - 24 Jul 2002

Mq , of the general form constructed by Connes [12] :

*
0,1 0,1
Mq = MTop ß MTop ß S1 (11)

*
0,1 0,1
MTop being a type II factor and ß the group acting periodically on MTop . The relation between
2
the periods and ß is such that , so that when ß , we get 0 (the periodicity is
ß

suppressed).

At the Lie group level, this "superposition state" can simply be given by the symmetric homogeneous

space constructed in [10] :

SO(3 , 1) SO (4)
h= (12)
SO (3)

to which corresponds, at the level of the underlying metric spaces involved, the topological quotient

space :
3, 1 4
top = (13)
SO(3)

In the non commutative context, G. Bogdanoff has constructed, again in [10], the "cocyle

bicrossproduct" :

Uq(so(4)op Uq(so(3, 1)) (14)

113
 where Uq(so(4)op and Uq(so(3, 1)) are Hopf algebras (or "quantum groups"[21]) and a 2-cocycle of

q-deformation. The bicrossproduct (7) suggests an unexpected kind of "unification" between the

Lorentzian and the Euclidean Hopf algebras at the Planck scale and yields the possibility of a

"q-deformation" of the signature from the Lorentzian (physical) mode to the Euclidean (topological)

mode [10-19].

Now, starting from equ. (10), it appears clearly that the Tomita-Takesaki modular automorphisms
group ßc (M q ) corresponds to the "unification" given by equ.(14) and induces, within the KMS

field, the existence of two dual flows.

(i) On the boundary ßi Planck , the first possible flow, of the form
tel-00001503, version 1 - 24 Jul 2002

eiH Mq e iH
t (Mq ) (15)

corresponds to the algebra of observables of the system and to the Lorentzian flow in real time. In this
perspective, this flow is a "physical flow", that we call P . This scale represents the physical part of
ß 0

the light cone and, consequently, the notion of (Lebesgue) measure is fully defined. Therefore, the

(commutative) algebra involved at such a scale is endowed with a hyperfinite trace and is given on the
infinite Hilbert space ( ), with = L2( ). Then, (L2( )) is a type I factor, indexed by the real
group , which we call MPhys . Of course, at this scale, the theory is Lorentzian, controlled by SO(3,

1).

(ii) On the "zero scale" ß i = 0 limit, the second flow takes necessarily the non unitary form :

H H
r r
iß (Mq ) e Mq e (16)

giving on Mq the semi-group of unbounded and non-stellar operators. This initial «topological» scale

corresponds to the imaginary vertex of the light cone, i.e. a zero-size gravitational instanton [19]. All
the measures performed on the Euclidean metric being -equivalent up to infinity, the system is

ergodic. As shown by A. Connes, any ergodic flow for an invariant measure in the Lebesgue measure
class gives a unique type II hyperfinite factor [12]. This strongly suggests that the singular 0-scale

114
 should be described by a type II factor, endowed with a hyperfinite trace noted Tr . We have called
0,1
MTop such a "topological" factor, which is an infinite tensor product of matrices algebra (ITPFI)
0,1
of the RO,1 Araki-Woods type [22]. Since Mq = MTop ß S1 , on the ß i = 0 limit, we get
0,1
Mq MTop .With respect to the analytic continuation between () and (), iß (Mq ) represents a "current

in imaginary time". We have stated in [19] that this current is another way to interpret the "flow of

weights" of the algebra Mq [23]. Clearly, according to [23], the flow of weights of Mq is an ergodic
flow, which represents an invariant of Mq . Then, iß (Mq ) yields a pure topological amplitude [24]

and, as such, "propagates" in imaginary time from zero to infinity. iß (Mq )is not defined on the

whole algebra Mq but on an ideal { of Mq. One can demonstrate that in this case, the theory is

Riemannian, the isometries of the metric being given by SO(4). As showed in [10, 19] this zero scale
tel-00001503, version 1 - 24 Jul 2002

corresponds to the first Donaldson invariant I = ( 1)n i and can be described by the topological
i

quantum field theory proposed by E. Witten in [24].

To finish, let's observe that the topological flow does not commute with the physical flow. Again,

this is a direct and natural consequence of the KMS condition.

5. DISCUSSION

It is interesting to remark that in the totally different context of superstrings, J.J. Atick and E. Witten

were the first to propose such an extension of the real temperature towards a complex domain [1].

Recently, in N=4 supersymmetric string theory, I. Antoniadis, J.P. Deredinger and C. Kounnas [5]

have also suggested to shift the real temperature to imaginary one by identification with the inverse

radius of a compactified Euclidean time on S1, with R = 1 / 2 T. Consequently, one can introduce a
complex temperature in the thermal moduli space, the imaginary part coming from the B
S / T/ U S / T/ U
antisymmetric field under type IIA type IIB Heterotic string-string dualities.

More precisely, in Antoniadis and al approach, the field controlling the temperature comes from the

product of the real parts of three complex fields : s= Re S, t= Re T and u= Re U. Within our KMS

approach, the imaginary parts of the moduli S, T, U can be interpreted in term of Euclidean

115
 temperature. Indeed, from our point of view, a good reason to consider the temperature as complex at

the Planck scale is that a system in thermodynamical equilibrium state must be considered as subject

to the KMS condition [11].

(ii) On the other hand, according to most of the models, supergravity is considered as broken for

scales greater than the Planck scale. But supersymmetry breaking is also closely connected to the

cancellation of the thermodynamic equilibrium state. C. Kounnas has recently demonstrated that a

five-dimensional (N=4) supersymmetry can be described by a four-dimensional theory in which

supersymmetry is spontaneously broken by finite thermal effects [6]. This scenario may be applied to

our setting. As a matter of fact, the end of the thermal equilibrium phase at the Planck scale might
tel-00001503, version 1 - 24 Jul 2002

bring about the breaking of KMS state and of supersymmetry N = 4. This corresponds exactly, in our

case, to the decoupling between imaginary time and real time.

To sum up, the chain of events able to explain the transition from the topological phase to the

physical phase of the spacetime might be the following :

thermodynamical equilibrium breaking KMS state breaking imaginary time / real time decoupling
topological state / physical state decoupling Supersymmetry breaking

We have given a detailed description of such a transition in [19]. Likewise, the supersymmetry is

broken in [5-6] by the finite temperature, which corresponds in our view to the decoupling between

real and imaginary (topological) temperature (the topological temperature being identified, in Kounnas
model, with the inverse radius of a compactified Euclidean time on S1 : 2 T 1 R ). Applying this

representation, the partition function in our case is given by the (super)trace over the thermal spectrum

of the theory in 4 dimensions. According to this, supersymmetry breaking and transition from

topological state to physical state might be deeply connected.

116
 Acknowledgements. This work has benefited of the encouragement from François Combes, of the Mathematics

Department of the University of Orleans, Daniel Sternheimer, from the University of Bourgogne and Jac Verbaarschot, of

the University of Stony Broock.

REFERENCES
1. ATICK J.J. WITTEN E. 1988 "The Hagedorn Transition and the Number of Degrees of Freedom in String Theory". Nucl.
Phys. B 310 291-334

2. HUANG K. WEINBERG S. 1970 "Ultimate Temperature and the Early Universe" Phys. Rev. Letters 25 895-897
tel-00001503, version 1 - 24 Jul 2002

3. DOLAN L. JACKIW R. 1974 “Symmetry Behavior at Finite Temperature” Phys. Rev. D9 3320-3341.

4. WITTEN E. 1991 "Lectures on Quantum Field Theory". Quantum Fields and Strings Am. Math. Soc. Vol, 1, 2 419-509,
1119-424

5. ANTONIADIS I. DEREDINGER J.P. KOUNNAS, C. 1999 "Non-Perturbative Supersymmetry Breaking and Finite Temperature
Instabilities in N=4 Superstrings". Preprint hep-th 9908137

6. KOUNNAS C. 1999 "Universal Thermal Instabilities and the High-Temperature Phase of the N=4 Superstrings". Preprint hep-
th 9902072

7. TRUCKS M. 1998 “A KMS-like state of Hadamard type on Rob ertson-Walker spacetimes and its time evolution “ Commun.
Math. Phys. 197 387-404

8. LONGO R. 2000 CERN “Notes for a Quantum Index Theorem “ Preprint Math. OA - 0003082

9. BUCHHOLZ D.A., LONGO R. 1999 “Graded KMS Functionals and the Breackdown of Supersymmetry” Adv. Theor. Math.
Phys. 3 615-626

10. BOGDANOV G. 1999 "Fluctuations Quantiques de la Signature de la Métrique à l'Echelle de Planck" Th. Doctorat Univ. de
Bourgogne

11. HAAG. R. HUGENHOLZ N. WINNINK M. 1967 "On the Equilibrium States in Quantum Statistical Mechanics". Commun.
Math. Phys. 5, 215-236

12. CONNES A 1994 "Non Commutative Geometry ". Academic Press

13. DONALSON S.K. 1990 "Polynomial Invariants for Smooth Four Manifolds". Topology 29, 257-315

14. FRÉ P. SORIANI P. 1995 "The N=2 Wonderland. From Calabi-Yau Manifolds to Topological Field Theory". World Scientific
Publ.

15. ALVAREZ E. ALVAREZ-GAUMÈ L. LOZANO Y. 1995"An Introduction to T-Duality in String Theory" Nucl. Phys. B 451
1-20

16. VAFA C. 1997 "Lectures on Strings and Dualities.Preprint hep-th/9702201.

17. CONNES A. ROVELLI C. 1994 "Von Neumann Algebra Automorphisms and Time Thermodynamics Relation in General
Covariant Quantum Theories". Class. Quantum Grav. 11 2899-2917

117
 18. DEREDINGER J.P. LUCCHESI C. 1998 "Realizations of Thermal Supersymmetry" Hep-Ph 980740319.

19. BOGDANOV G. BOGDANOV I. 2001 "Topological Field Theory of the Initial Singularity of Spacetime". Class. Quantum
Grav. 18 4341-4372

20. TAKESAKI M. 1970 "Tomita's Theory of Modular Hilbert Algebras and its Applications". Lecture Notes 128 Springer-
Verlaag

21. MAJID S. 1995 "Foundations of Quantum Groups" Cambridge University Press

22. ARAKI H. WOODS E.J. 1968 "A Classification of Factors" Publ. Res. Inst. Math. Sci. Kyoto Univ. 4

23. CONNES A. TAKESAKI M. 1977 "The flow of weights on factors of type III" Tohoku Math. Journ. 29 473-575

25. WITTEN E. 1988 "Topological Quantum Field Theory". Com. Math. Phys. 117 353-386
tel-00001503, version 1 - 24 Jul 2002

118
 CHINESE JOURNAL OF PHYSICS VOL 40 NO 2 APRIL 2002

The KMS State of Spacetime at the Planck Scale

Igor Bogdanoff

Mathématical Physics Laboratory CNRS UMR 5029 Bourgogne University

U.F.R. Sciences et Techniques
B.P. 47870
F-21078 Dijon Cedex FRANCE
tel-00001503, version 1 - 24 Jul 2002

Considering the expected thermal equilibrium characterizing the physics at the Planck scale it is here stated, for the first

time, that as a system the spacetime at the Planck scale must be considered as subject to the Kubo-Martin-Schwinger

(KMS) condition. Consequently in the interior of the KMS strip, i.e. from the scale ß = 0 to the scale ß = Planck, the

fourth coordinate g44 must be considered as complex, the two real poles being ß = 0 and ß = Planck. This means that

within the limits of the KMS strip the Lorentzian and the Euclidean metric are in a "quantum superposition state" (or

coupled), this entailing a "unification" (or coupling) between the topological (Euclidean) and the physical (Lorentzian)

states of spacetime.

1. INTRODUCTION

Much have been recently said regarding the physical state of the universe at the vicinity of the Planck

scale. String theory, non commutative geometry, supergravity or quantum gravity have contributed,

independently of each others, to establish on solid basis the data of a "transition phase" in the physical

content and the geometric structures of the (pre)spacetime at such a scale. But what is the nature of

this dramatic change? In the present paper, we propose a novel approach, based on one of the most

natural and realistic physical condition predicted by the Standard Model for the (pre)universe. In

agreement with some well-established results [1-2-3] (recalled in [4] ) and, more recently, the

119
 approach of C. Kounnas and al [5-6], we argue that at the Planck scale, the "spacetime system" is in a

thermodynamical equilibrium state [2]. This notion of equilibrium state at the Planck scale has been

recently stated with some new interesting arguments in [7-8-9]. As a consequence, according to [10],

we suggest hereafter that the (pre)universe should be considered as subject to the Kubo-Martin-

Schwinger (KMS) condition [11] at such a scale. Surprisingly, the well known KMS and modular

theories [12] have never been applied to the study of the metric properties in the context of quantum

cosmology ; however, the KMS condition might have dramatic consequences onto Planck scale

physics. For which reasons? Because, when applied to quantum spacetime, the KMS statistics are

such that , within the limits of the "KMS strip" (i.e. between the scale zero and the Planck scale), the
time like direction of the system should be considered as complex : t tc tr iti . G.Bogdanoff has
tel-00001503, version 1 - 24 Jul 2002

showed in [10] that at the scale zero, the theory is projected onto the pure imaginary boundary
t tc iti of the KMS strip. Namely, there exists, around ß 0, a non-trivial topological limit of

quantum field theory, dual to the usual topological limit associated with ß in the partition

function (2). Such a topological state of the (pre)spacetime can be described by the topological

invariant

Z = Tr (-1)s (1)
ß 0

given by the ß 0 limit of the partition function (2). S is here the instanton number of the theory. It

has been demonstrated [10] that this topological index is isomorphic to the first Donaldson Invariant
[13]. This suggests that at zero scale, the observables O i must be replaced by the homology cycles
( k)
Hi mod in the moduli space of gravitational instantons [14]. We get then a deep correspondence

-a symmetry of duality- [15-16], between physical theory and topological field theory. Conversely,
on the (classical) infrared limit ß Planck , the imaginary component cancels and the time like

direction becomes pure real t tc t r So, within the limits of the KMS strip (i.e. for 0 < ß <

Planck ), the Lorentzian and the Euclidean metric are in a "quantum superposition state" (or coupled).

Such a superposition state, described in detail in [10], is entailing a "unification" (or coupling)

120
 between the topological (Euclidean) and the physical (Lorentzian) states of spacetime. The
(Lorentzian / Euclidean) states of the metric g are given by the partition function

Z = Tr (-1)s e ßH (2)

1
with ß = as usual and n being the "metric number" of the theory. To avoid any difficulty of
kT
interpretation, let's remark that the scale parameter ß admits two possible intepretations : (i) either
ß kT can be seen as a real time parameter of a Lorentzian (3+1)-dimensional theory [17] or (ii)
ß = it can be interpreted as the fourth space-like direction of a Riemannian 4-dimensional Euclidean
theory (e.g. ß Planck ) [18]. In the second case, ß is a periodic imaginary time interval [5-6].

Considering the hypothesis of holomorphicity formulated in section 4, we use here the two
tel-00001503, version 1 - 24 Jul 2002

interpretations. In such a context, the KMS state of the (pre)spacetime may be considered as a

transition phase from the Euclidean topological phase (ß = 0) to the Lorentzian physical phase,

beyond the Planck scale [10-19].

The present article is organized as follows. In section 2 we recall that at the Planck scale, the

"spacetime system" should most likely be considered as being in a thermodynamical equilibrium state.

In section 3, we show that, as a natural consequence of this equilibrium state, the spacetime must be

considered as subject to the KMS condition. In section 4, we suggest that, considering the KMS
properties, the time-like direction g44 of the metric should be seen as complex tc = tr + i ti within the

limits of the KMS strip. In section 5, we discuss briefly the transition from imaginary time ti to real

time tr in terms of KMS breaking beyond the Planck scale.

2. THERMODYNAMICAL EQUILIBRIUM OF THE SPACETIME AT THE PLANCK

SCALE

It is well known that at the Planck scale, one must expect a thermodynamical "phase transition",

closely related to (i) the existence of an upper limit in the temperature growth -the Hagedorn

121
 temperature [1]- and (ii) the equilibrium state characterizing, most likely, the (pre)spacetime at such a

scale. In such a context the seminal investigations of K. Huang and S. Weinberg [2], Dolan and R.

Jackiw [3] , then later of several others, have renewed the initial idea of Hagedorn concerning the

existence, at very high temperature, of a limit restricting the growth of states excitation. Already some

time ago, J.J. Atik and E.Witten have shown the existence of a Hagedorn limit around the Planck

scale in string theory [1]. The reason is that, as recently recalled by C. Kounnas [6] in the context of

N = 4 superstrings, at finite temperature, the partition function Z(ß) and the mean energy U(ß)

develop some power pole singularities in ß T since the density of states of a system grows

exponentially with the energy E :

tel-00001503, version 1 - 24 Jul 2002

ßE 1
Z( ß) dE (E)e ~
( ß b)(k 1)

1
U( ß) ln Z ~ (k 1) (3)
ß b

Clearly, equ.(3) exhibits the existence, around the Planck scale, of a critical temperature TH b 1,
where the (pre)spacetime system must be viewed as in a thermodynamical equilibrium state. Indeed,
a(t) being the cosmological scale factor, the global temperature T follows the well-known law :

thermodynamical equilibrium breaking KMS state breaking imaginary time / real time decoupling
topological state / physical state decoupling Supersymmetry breaking
1
Ep C5 2
and around the Planck time, T is reaching the critical limit T p kB 1 1,4.1032 K . In
kB G

fact, it is currently admitted in string theory that, before the inflationary phase, the ratio between the
interaction rate ( ) of the initial fields and the (pre)spacetime expansion (H) is << 1, so that the
H
system can reasonably be considered in equilibrium state. This has been established a long time ago

within some precursor works already quoted in ref. [1,2,3]. Recently the same approach has been

successfully considered by C. Kounnas et al in the superstrings context [5]. However, this natural

notion of equilibrium, when viewed as a global gauge condition, has dramatic consequences regarding

122
 physics at the Planck scale. Among those consequences, unquestionably the most important is that the

(pre)spacetime at the Planck scale must be considered as subject to the famous "KMS condition", a

very special and interesting physical state that we are now going to describe.

3. THE (PRE)SPACETIME IN KMS STATE AT THE PLANCK SCALE

Let's first recall on mathematical basis what an equilibrium state is.

Definition H being an autoadjoint operator and the Hilbert space of a finite system, the
Tr (e ßH A)
tel-00001503, version 1 - 24 Jul 2002

equilibrium state of this system is described by the Gibbs condition ( A) and

Tr (e ßH )
satisfy the KMS condition.

This well-known definition has been proposed for the first time in [11]. Now, it is usual (and

natural) to oppose the notion of equilibrium to the one of evolution of a system. In fact, the famous

Tomita-Takesaki modular theory has established that the "intrinsic" dynamic of a quantum system
corresponds, in a unique manner, to the strongly continuous one parameter * - automorphism group

t of some von Neumann C* - algebra A [12] :

t (A) = ei Ht A e i Ht (4)

This one parameter group describes the time evolution of the observables of the system and

corresponds to the well-known Heisenberg algebra. Nothing mysterious at this stage. However, we

are here brought to find the remarkable discovery of Takesaki and Winnink, connecting the evolution
group t (A) of a system (more precisely the modular group M = it M it) with the equilibrium
Tr(Ae ßH )
state (A) of this system [11-20]. With this more or less unexpected relation between
Tr(e ßH )

evolution t (A) and equilibrium (A), we now meet the famous "KMS condition". More exactly, in

the frame of quantum statistical mechanic, the KMS condition provides a rigorous mathematical

123
 formulation about the coexistence of different possible equilibrium states at the same given

temperature T.

Let's recall now how such a relation between equilibrium state and evolution of a system is realized
by the KMS condition. It has been clearly established [11] that a state on the C* -algebra A and the

continuous one parameter automorphism group of A at the temperature ß = 1 / k T verify the KMS
condition if, for any pair A, B of the * - sub-algebra of A, t - invariant and of dense norm, it exists

a ƒ(tc) function holomorphic in the strip {tc = t + i ß , Im t c [ 0 , ß ] } such that :

(i) ƒ (t) = (A ( t B)) ,

(ii) ƒ (t + i ß ) = ( t (B)A) , t . (5)
tel-00001503, version 1 - 24 Jul 2002

Moreover, a state on the C* -algebra A is separator if the given algebraic representation is a von

Neumann algebra W* endowed with a separator and cyclic vector. The sets

Il = {A A (A * A ) = 0}

and

Ir = {A A : (AA*) =0}

are forming a left and right ideal in A. For any KMS state, we have I l = I r.

The above definition expresses the bijective relation between equilibrium state, holomorphic state of

the measure parameters and KMS state.

Now, considering the general properties raised by the KMS condition, if we admit that around the

Planck scale, the (pre)spacetime system is in a thermal equilibrium state, then we are also bound to

admit that this system is in a KMS state. Indeed, it has been shown a long time ago [11] that if a state

of a system satisfies the equilibrium condition ([h, t (A)])dt 0, A U , then, satisfies

the KMS condition. So, there is a biunivoque relation between equilibrium state and KMS state. So,

124
 if we admit that around Planck , the (pre)spacetime system is in a thermal equilibrium state, then

according to [11], we are also bound to admit that this system is in a KMS state.

Next, let's push forwards the consequences raised by the holomorphicity of the KMS strip.

4. HOLOMORPHIC TIME FLOW AT THE PLANCK SCALE

As a critical consequence of the KMS condition, we are induced to consider that the time-like
coordinate g00 becomes holomorphic within the limits of the KMS strip. Indeed, as demonstrated in
tel-00001503, version 1 - 24 Jul 2002

details in [10-19], within the KMS strip, we necessarily should have :

t tc = tr + i ti (6)

In the same way, the physical (real) temperature should also be considered as complex at the Planck

scale :
T Tc = Tr + iTi (7)

as proposed by Atick and Witten in another context [1]. This unexpected effect is simply due to the

fact that, given a von Neumann algebra W* and two elements A, B of W* , then there exists a
function ƒ(tc) holomorphic in the strip tc , im tc 0, ß such that :

ƒ(t) = (A ( t B)) and ƒ(tr +i ß ) = ( t (B)A), t (8)

Here, t is the usual time parameter of the 3D theory, like ß kT . So in our case, within the limits
of the KMS strip, i.e. from the scale zero (ß = 0) to the Planck scale (ß = Planck ), the "time-like"

direction of the system must be extended to the complex variable

tc = tr + i ti , Im t c [i ti , tr] (9)

Of course, the holomorphicity of the time like direction of the spacetime is induced in a natural manner

by the fact that in our approach, the thermodynamical system is the spacetime itself. Such a situation

has been investigated in details in [10] in the context of "quantum groups" and non-commutative

geometry.

125
 Indeed, according to Tomita's modular theory [20], the KMS condition, when applied to the

spacetime as a system, allows, within the KMS strip, the existence of an "extended" (holomorphic)

automorphism "group of evolution", which depends, in the classification of factors [12], on a "type
III factor" Mq (a factor is a special type of von Neumann algebra, whose the center is reduced to the

scalars a ). The extended automorphism group has the following form :

Hßc Hßc
Mq ßc (Mq ) e M q e

with the ßc = ßr +i ßi parameter being formally complex and able to be interpreted as a complex time

t and / or temperature T Tc = Tr + iTi . So, the KMS condition suggests the existence at the Planck

scale, of an effective one loop potential coupled, in N = 2 supergravity, to the complex dilaton + axion
1 i
field = 2 i and yielding the dynamical form diag(1 , 1 , 1 , e ) for the metric. The
tel-00001503, version 1 - 24 Jul 2002

g
signature of is Lorentziann (i.e. physical) for =± and can become Euclidean (topological)

for = 0. Consequently, the "KMS signature" of the metric is (+++±). This is as it should be since,
considering the quantum fluctuations of g , there is no more invariant measure on the non

commutative metric Therefore, according to von Neumann algebra theory, the "good factor"

addressing those constraints is uniquely a non commutative traceless algebra, i.e. the type III factor

Mq , of the general form constructed by Connes [12] :

*
0,1 0,1
Mq = MTop ß MTop ß S1 (11)

*
0,1 0,1
MTop being a type II factor and ß the group acting periodically on MTop . The relation between
2
the periods and ß is such that , so that when ß , we get 0 (the periodicity is
ß

suppressed).

At the Lie group level, this "superposition state" can simply be given by the symmetric homogeneous

space constructed in [10] :

SO(3 , 1) SO (4)
h= (12)
SO (3)

126
 to which corresponds, at the level of the underlying metric spaces involved, the topological quotient

space :
3, 1 4
top = (13)
SO(3)

In the non commutative context, G. Bogdanoff has constructed, again in [10], the "cocyle

bicrossproduct" :

Uq(so(4)op Uq(so(3, 1)) (14)

where Uq(so(4)op and Uq(so(3, 1)) are Hopf algebras (or "quantum groups"[21]) and a 2-cocycle of

q-deformation. The bicrossproduct (7) suggests an unexpected kind of "unification" between the
tel-00001503, version 1 - 24 Jul 2002

Lorentzian and the Euclidean Hopf algebras at the Planck scale and yields the possibility of a

"q-deformation" of the signature from the Lorentzian (physical) mode to the Euclidean (topological)

mode [10-19].

Now, starting from equ. (10), it appears clearly that the Tomita-Takesaki modular automorphisms
group ßc (M q ) corresponds to the "unification" given by equ.(14) and induces, within the KMS

field, the existence of two dual flows.

(i) On the boundary ßi Planck , the first possible flow, of the form

eiH Mq e iH
t (Mq ) (15)

corresponds to the algebra of observables of the system and to the Lorentzian flow in real time. In this
perspective, this flow is a "physical flow", that we call P . This scale represents the physical part of
ß 0

the light cone and, consequently, the notion of (Lebesgue) measure is fully defined. Therefore, the

(commutative) algebra involved at such a scale is endowed with a hyperfinite trace and is given on the
infinite Hilbert space ( ), with = L2( ). Then, (L2( )) is a type I factor, indexed by the real
group , which we call MPhys . Of course, at this scale, the theory is Lorentzian, controlled by SO(3,

1).

127
 (ii) On the "zero scale" ß i = 0 limit, the second flow takes necessarily the non unitary form :

H H
r r
iß (Mq ) e Mq e (16)

giving on Mq the semi-group of unbounded and non-stellar operators. This initial «topological» scale

corresponds to the imaginary vertex of the light cone, i.e. a zero-size gravitational instanton [19]. All
the measures performed on the Euclidean metric being -equivalent up to infinity, the system is

ergodic. As shown by A. Connes, any ergodic flow for an invariant measure in the Lebesgue measure
class gives a unique type II hyperfinite factor [12]. This strongly suggests that the singular 0-scale

should be described by a type II factor, endowed with a hyperfinite trace noted Tr . We have called
0,1
MTop such a "topological" factor, which is an infinite tensor product of matrices algebra (ITPFI)
tel-00001503, version 1 - 24 Jul 2002

0,1
of the RO,1 Araki-Woods type [22]. Since Mq = MTop ß S1 , on the ß i = 0 limit, we get
0,1
Mq MTop .With respect to the analytic continuation between () and (), iß (Mq ) represents a "current

in imaginary time". We have stated in [19] that this current is another way to interpret the "flow of

weights" of the algebra Mq [23]. Clearly, according to [23], the flow of weights of Mq is an ergodic
flow, which represents an invariant of Mq . Then, iß (Mq ) yields a pure topological amplitude [24]

and, as such, "propagates" in imaginary time from zero to infinity. iß (Mq )is not defined on the

whole algebra Mq but on an ideal { of Mq. One can demonstrate that in this case, the theory is

Riemannian, the isometries of the metric being given by SO(4). As showed in [10, 19] this zero scale
corresponds to the first Donaldson invariant I = ( 1)n i and can be described by the topological
i

quantum field theory proposed by E. Witten in [24].

To finish, let's observe that the topological flow does not commute with the physical flow. Again,

this is a direct and natural consequence of the KMS condition.

5. DISCUSSION

It is interesting to remark that in the totally different context of superstrings, J.J. Atick and E. Witten

were the first to propose such an extension of the real temperature towards a complex domain [1].

128
 Recently, in N=4 supersymmetric string theory, I. Antoniadis, J.P. Deredinger and C. Kounnas [5]

have also suggested to shift the real temperature to imaginary one by identification with the inverse

radius of a compactified Euclidean time on S1, with R = 1 / 2 T. Consequently, one can introduce a
complex temperature in the thermal moduli space, the imaginary part coming from the B
S / T/ U S / T/ U
antisymmetric field under type IIA type IIB Heterotic string-string dualities.

More precisely, in Antoniadis and al approach, the field controlling the temperature comes from the

product of the real parts of three complex fields : s= Re S, t= Re T and u= Re U. Within our KMS

approach, the imaginary parts of the moduli S, T, U can be interpreted in term of Euclidean

temperature. Indeed, from our point of view, a good reason to consider the temperature as complex at

the Planck scale is that a system in thermodynamical equilibrium state must be considered as subject
tel-00001503, version 1 - 24 Jul 2002

to the KMS condition [11].

(ii) On the other hand, according to most of the models, supergravity is considered as broken for

scales greater than the Planck scale. But supersymmetry breaking is also closely connected to the

cancellation of the thermodynamic equilibrium state. C. Kounnas has recently demonstrated that a

five-dimensional (N=4) supersymmetry can be described by a four-dimensional theory in which

supersymmetry is spontaneously broken by finite thermal effects [6]. This scenario may be applied to

our setting. As a matter of fact, the end of the thermal equilibrium phase at the Planck scale might

bring about the breaking of KMS state and of supersymmetry N = 4. This corresponds exactly, in our

case, to the decoupling between imaginary time and real time.

To sum up, the chain of events able to explain the transition from the topological phase to the

physical phase of the spacetime might be the following :

thermodynamical equilibrium breaking KMS state breaking imaginary time / real time decoupling
topological state / physical state decoupling Supersymmetry breaking

129
 We have given a detailed description of such a transition in [19]. Likewise, the supersymmetry is

broken in [5-6] by the finite temperature, which corresponds in our view to the decoupling between

real and imaginary (topological) temperature (the topological temperature being identified, in Kounnas
model, with the inverse radius of a compactified Euclidean time on S1 : 2 T 1 R ). Applying this

representation, the partition function in our case is given by the (super)trace over the thermal spectrum

of the theory in 4 dimensions. According to this, supersymmetry breaking and transition from

topological state to physical state might be deeply connected.

tel-00001503, version 1 - 24 Jul 2002

Acknowledgements. This work has benefited of the encouragement from François Combes, of the Mathematics

Department of the University of Orleans, Daniel Sternheimer, from the University of Bourgogne and Jac Verbaarshot, of

the University of Stony Broock.

REFERENCES

1. ATICK J.J. WITTEN E. 1988 "The Hagedorn Transition and the Number of Degrees of
Freedom in String Theory". Nucl. Phys. B 310 291-334

2. HUANG K. WEINBERG S. 1970 "Ultimate Temperature and the Early Universe" Phys. Rev.
Letters 25 895-897

4. DOLAN L. JACKIW R. 1974 “Symmetry Behavior at Finite Temperature” Phys. Rev. D9

3320-3341.

4. WITTEN E. 1991 "Lectures on Quantum Field Theory". Quantum Fields and Strings Am. Math.
Soc. Vol, 1, 2 419-509, 1119-424

5. ANTONIADIS I. DEREDINGER J.P. KOUNNAS, C. 1999 "Non-Perturbative Supersymmetry

Breaking and Finite Temperature Instabilities in N=4 Superstrings". Preprint hep-th 9908137

130
 6. KOUNNAS C. 1999 "Universal Thermal Instabilities and the High-Temperature Phase of the N=4
Superstrings". Preprint hep-th 9902072

7. TRUCKS M. 1998 “A KMS-like state of Hadamard type on Robertson-Walker spacetimes and its
time evolution “ Commun. Math. Phys. 197 387-404

8. LONGO R. 2000 CERN “Notes for a Quantum Index Theorem “ Preprint Math. OA - 0003082

9. BUCHHOLZ D.A., LONGO R. 1999 “Graded KMS Functionals and the Breackdown of
Supersymmetry” Adv. Theor. Math. Phys. 3 615-626

10. BOGDANOV G. 1999 "Fluctuations Quantiques de la Signature de la Métrique à l'Echelle de

Planck" Th. Doctorat Univ. de Bourgogne

11. HAAG. R. HUGENHOLZ N. WINNINK M. 1967 "On the Equilibrium States in Quantum
tel-00001503, version 1 - 24 Jul 2002

Statistical Mechanics". Commun. Math. Phys. 5, 215-236

12. CONNES A 1994 "Non Commutative Geometry ". Academic Press

13. DONALSON S.K. 1990 "Polynomial Invariants for Smooth Four Manifolds". Topology 29,
257-315

14. FRÉ P. SORIANI P. 1995 "The N=2 Wonderland. From Calabi-Yau Manifolds to Topological
Field Theory". World Scientific Publ.

15. ALVAREZ E. ALVAREZ-GAUMÈ L. LOZANO Y. 1995"An Introduction to T-Duality in String

Theory" Nucl. Phys. B 451 1-20

16. VAFA C. 1997 "Lectures on Strings and Dualities.Preprint hep-th/9702201.

17. CONNES A. ROVELLI C. 1994 "Von Neumann Algebra Automorphisms and Time
Thermodynamics Relation in General Covariant Quantum Theories". Class. Quantum Grav. 11 2899-
2917

18. DEREDINGER J.P. LUCCHESI C. 1998 "Realizations of Thermal Supersymmetry" Hep-Ph

980740319.

19. BOGDANOV G. BOGDANOV I. 2001 "Topological Field Theory of the Initial Singularity of
Spacetime". Class. Quantum Grav. 18 4341-4372

131
 20. TAKESAKI M. 1970 "Tomita's Theory of Modular Hilbert Algebras and its Applications".
Lecture Notes 128 Springer-Verlaag

21. MAJID S. 1995 "Foundations of Quantum Groups" Cambridge University Press

22. ARAKI H. WOODS E.J. 1968 "A Classification of Factors" Publ. Res. Inst. Math. Sci.
Kyoto Univ. 4

23. CONNES A. TAKESAKI M. 1977 "The flow of weights on factors of type III" Tohoku Math.
Journ. 29 473-575

26. WITTEN E. 1988 "Topological Quantum Field Theory". Com. Math. Phys. 117 353-386
tel-00001503, version 1 - 24 Jul 2002

132
 CZECHOSLOVAK JOURNAL OF PHYSICS
Volume 51, 2001 Number 11 pp 1153-1236

Topological Origin of Inertia

Igor BOGDANOV
Mathematical physics Laboratory CNRS UPRES A 5029
BOURGOGNE University, FRANCE

The purpose of the present article, following "Mach's principle" (the main elements of which have contributed to
tel-00001503, version 1 - 24 Jul 2002

the foundations of general relativity) is to propose a new (non-local) interpretation of the inertial interaction. We
then suggest that the inertial interaction can be correctly described by the topological field theory proposed by
Witten in 1988. In such a context, the instantaneous propagation and the infinite range of the inertial interaction
might be explained in terms of the topological amplitude connected with the singular zero size gravitational
instanton corresponding to the Initial Singularity of space-time.

1. INTRODUCTION

The phenomenon of inertia - or "pseudo-force" according to E. Mach [1] - has recently been
presented by J.P. Vigier [2] as one of the "unsolved mysteries of modern physics". Indeed, our
point of view is that this important question, which is well formulated in the context of Mach's
principle, cannot be resolved or even understood in the framework of conventional field theory.

Here we suggest a novel approach, a direct outcome of the topological field theory proposed by
Edward Witten in 1988 [3]. According to this approach, beyond the interpretation proposed by
Mach, we consider inertia as a topological field, linked to the topological charge Q = 1 of the
"singular zero size gravitational instanton" [4] which, according to [5], can be identified with the
initial singularity of space-time in the standard model.

Evaluation of the total inertial (or inertial potential) contribution, resulting from the sum of the
masses in the universe may be given by:

GM
Uinertial total = 2 1 (1)
univers c r

133
 which turns out to be an invariant for each local mass. On the other hand, the topological charge
of the singular gravitational instanton, of the form

1
Q 2 d 4 x R R˜ =1 (2)
32

is also an invariant equal to unity, just as the total inertial contribution. Unexpectedly, the
equivalence between the inertial mass and the gravitational mass might be explained here in terms
of the quantization of the topological charge of the singular gravitational instanton [6].

We therefore claim that the topological charge of the singular gravitational instanton of zero size
represents the (non-local) source of the inertial interaction. Like the topological charge, the
tel-00001503, version 1 - 24 Jul 2002

inertial interaction is also invariant and propagates itself "instantaneously" from one point to
another in space-time. Such a property is not explicable within the framework of field theory but
may find a solution in topological field theory. In this new context, the initial singularity can be
viewed as the source of a "topological amplitude" [3] corresponding to the charge of the singular
zero size gravitational instanton, i.e. Q d 4 x R R˜ . This amplitude can be detected at the
boundary S3 of the singular gravitational instanton, which, according to [5], might have the
topology of the 4-dimensional Euclidean ball B4 bounded by the sphere S3. So, the pseudo-
observables of Riemannian space-time in the vicinity of the origin are here interpreted as co-
cycles on instanton moduli space and are associated with i cycles on the B4 four-dimensional
manifold (Donaldson mapping [7]). Considering a point X in B4, the topological amplitude
inducing the propagation of the instantonic charge can take the general form :

OS 3 . O X # (S 3 , X)

Starting from this, the pseudo-observables of the left member are giving the topological
Oi

amplitudes of the theory whereas the right member designates the number of intersections of i
B4. The function # (S 3 , X) is zero if the point X B4 and is equal to one if X B4 (the
case for which there exists topological amplitude).

The present article is organized as follows. In paragraph 2, we briefly recall the context in which
the problem of inertia is formulated (in classical mechanics) [2]. Likewise, we recall the
canonical formulation of Mach's principle, which suggests a non-local approach of the inertial

134
 interaction [1]. In paragraph 3, we consider the " topological Mach's principle", a new
formulation of Mach's principle in the context of topological field theory [3]. In paragraph 4, we
suggest that the propagation of the "causal signal" connected with the Initial Singularity of space-
time [5] can conveniently be described by the topological amplitude related to a singular
gravitational instanton of size zero [4]. We complete this approach by showing, from another
point of view, that if the "topological signal" at the origin has a Dirac mass (a distribution with a
support reduced to zero), then this signal is sent to infinity - i.e. onto the sphere S3 representing
the co-boundary of the space-time E4 and of the singular gravitational instanton B4 [5]. Last, we
discuss these diverse results and conjectures in paragraph 6.

2. REMINDER : INERTIA AND MACH'S PRINCIPLE

It is well known in classical mechanics that the inertial interaction is interpreted as an

tel-00001503, version 1 - 24 Jul 2002

instantaneous reaction to acceleration of any material object :

ƒi = ma

a being the acceleration of the system itself. According the Newtonian point of view, - and, to
some extent, in special relativity - , this reflects an inherent property of matter, which does not
conflict with the conception of absolute space and time.

On the contrary, according to general relativity, although independent of the observer, the
intrinsic properties of space-time and its geometry are described by a metric configuration
inseparable from the distribution of the matter. This relationship has been defined on axiomatic
basis in the "equivalence principle" between inertia and gravitational mass. However, how can
one explain the instantaneous propagation (i.e. non-physical in the strict sense) of the inertial
interaction ? How does this interaction propagate itself at "infinite speed" from one point to
another in space-time ?

The first attempt to provide a global response was qualitatively formulated by E. Mach [1] for
whom, on the basis of the relativity of all motion, the origin of inertia is the "interconnection" of
all matter in the universe. We take from this the generic expression of Mach's principle as
initially formulated :

Definition 2.1. (E. Mach) : the inertial reference point defined by local physics coincides with
the reference point in which distant objects are at rest, where it results that the most distant
masses distributed in the universe determine the inertial behavior of local masses.

135
 Numerous other variants of Mach's principle still exist, developed notably by Brans-Dicke [8],
Bondi-Samuel [9] or Brill [10] and some others. However, none of these approaches appears
able to explain, in a framework compatible with relativistic principles, the nature of the inertial
interaction as well as its mode of propagation.

We now propose hereafter to renew the approach of Mach's principle in the context of
topological field theory.

3. TOPOLOGICAL MACH'S PRINCIPLE

Our main suggestion within this paper is that the foundation on which Mach's principle lies (as
the notion of inertia) is not "physical" but falls within topological field theory, defined by Witten
tel-00001503, version 1 - 24 Jul 2002

in 1988 [3]. As an example, we consider Foucault's pendulum experiment , which cannot be

explained satisfactorily in either classical or relativistic mechanics. We recall that the problem is
the angular invariance of the plane of oscillation of the pendulum . Then, the "topological
Mach's principle" assumes that the interaction between and "global space-time " E is itself
of the topological type - which by its nature explains precisely the invariant and global
properties of the system formed by the oscillation plane of the pendulum and the rest of the
universe.

In [5] the existence of a zero scale limit in pre- space-time was suggested, such a limit
concentrating the energy density of the "whole universe", in the Mach sense. We begin by
showing that within such an approach, the zero scale configuration has no physical content
(there exists no stable physical length lower than the Planck length) but is a topological
configuration, corresponding to a singular gravitational instanton of zero size.

3.1 Zero scale of space-time and topological field theory

As Weinberg first stated [11], one can reasonably consider that space-time at the Planck scale

forms a system in global thermodynamical equilibrium. From an algebraic point of view, an

equilibrium state is a state on a C* - quasi-local algebra A, generated by a sub-algebra

corresponding to the kinematic observables of the sub-system. Such a state on A can be given,

at the temperature ß 1, by the Gibb's state :

136
 ßH
Tr (e A)
(A) ßH
Tr (e )

On the other hand, in the context of the Tomita-Takesaki modular theory, the famous Kubo-Martin-
Schwinger "(KMS) condition" [12] has established that each state (A) of the system can be
connected, in a unique manner, to the strongly continuous one parameter * - automorphisms group t

giving the time evolution of the observables and corresponding to the well known Heisenberg algebra

t (A) = ei Ht A e i Ht
tel-00001503, version 1 - 24 Jul 2002

Now, starting from the expected equilibrium state of the universe around the Planck scale, it has

been shown in [5] that (pre)space-time at the Planck scale can be seen as subject to the KMS

condition [12]. Therefore, within the holomorphic KMS strip (i.e. between the scale zero (ß = 0)
and the Planck scale (ß = Planck ), it is therefore natural to consider the timelike direction of the

metric as holomorphic, so that we should have :

t = tr + i ti

as showed in [5]. In the same way, the physical (real) temperature should also be viewed as

complex at the Planck scale :

T Tc = Tr + iTi

as proposed by Atick and Witten in another context [13]. Recently, in N=4 supersymmetric

string theory, I. Antoniadis, J.P. Deredinger and C. Kounnas [14] have also introduced at the

Planck scale a complex temperature in the thermal moduli space, the imaginary part coming from
S/T/U S/T/U
the B antisymmetric field under type IIA type IIB Heterotic string-

string dualities. So, the KMS condition suggests the existence at the Planck scale of an effective
one loop potential coupled, in N = 2 supergravity, to the complex dilaton + axion field =
1
i and yielding the following dynamical form of the metric
g2

137
 i i
diag(1 , 1 , 1 , e ) diag(1 , 1 , 1 , e ) (3)

The signature of the metric (3), endowed with a supplementary degree of freedom in the g44
component, is Lorentzian for = ± and can become Euclidean for = 0. The modular theory
of Tomita [15] suggests then the "dualisation" of the signature given by the generalized
automorphisms of the "quantum von Neumann algebra" Mq, which can be written :
Hßc Hßc
Mq ß (Mq ) e Mq e (4)
c

The temporal flow associated with (4) is formally holomorphic in the variable ßc = ßr+ ißi .
The group of modular automorphisms ß generates two dual flows. The first one, of the form
c
:

e iHt Mq e iHt
t (M q )`= , t . (5)
tel-00001503, version 1 - 24 Jul 2002

corresponds to the algebra of observables and to the Lorentzian flow in real time. On the other
hand, the dual current, is :
ßH
(Mq ) = e Mq e ßH (6)

giving on the topological von Neumann algebra Mq a semi-group of non-bounded and non-
stellar operators. The flow of Mq is defined not throughout the whole of Mq but on an
ideal { of Mq and coupled to the topological flow in pure ß = i t imaginary time. In the
model constructed in [5], the algebra of observables described by (5) is replaced at the scale zero
of space-time by an "algebra of states", dual to the algebra of Heisenberg in the form (6). Of
course, at the singular ß = 0 scale, it is no longer possible to conserve the notion of physical
observables; instead, we consider homology cycles in moduli space of (zero size) gravitational
instantons. This latter conclusion remains true, in pure imaginary time, for all real ß > 0. Such an
approach allows us to distinguish three different domains in the "cosmological light cone", each
of these domains being described by a specific von Neumann algebra called a "factor" [16]. If
0,1
we call MTop = R F the factor R0,1 of type II corresponding to the singular zero scale, as all
0,1
ergodic transformations starting from MTop (flows associated with zero scale) are weakly
0,1
equivalent, therefore MTop is a hyperfinite factor of the Araki-Woods type [17]. The factor
0,1
MTop is then canonical. More generally, there exist thus three scales (corresponding to the three
regions of the cosmological light cone) and described by three different "factors" :

(i) the classical scale (ß > Planck), described by the tracial factor MPhys of type I ;

138
 (ii) the quantum superposition scale (0 < ß < Planck) described by the traceless type
0,1
* + Then, we write Mq = MTop *+ ;
III factor such that III = II

(iii) the topological scale (zero scale associated with ß = 0) described by the tracial type II
0,1
factor MTop .

0,1
To finish, we note that the algebraic "flow of weights" [18] associated with the factor MTop of
0,1
type II at zero scale of space-time is an invariant of MTop . Thus, it has also been shown, again
in [5], that the initial space-time singularity, can be approached not in terms of divergences of
physical fields but in terms of topological field symmetries and associated invariants (such as the
first Donaldson invariant [7-19]) :

I= ( 1)n i
i
tel-00001503, version 1 - 24 Jul 2002

ni being defined by Donaldson as the instanton number of the theory [7]. i is indicating the i th
instanton and ni = 0 or 1 determines the sign of its contribution to Z. The contribution of the ith
instanton to Z is ( 1) ni . The index ( 1)n i is a non-polynomial topological invariant,
i
( k)
reduced to an integer when the dimension of the instanton moduli space dim mod = 0 [19].
Then, a possible resolution of the Initial Singularity consists in considering that the singular zero
scale, which cannot be described by the (perturbative) physical theory, should be described by
the (non-perturbative) dual theory, of the topological type.

3.2 Topological invariant of singularity

Starting from Witten [3], one normally defines topological field theory as the quantization of
zero, the Lagrangian of the theory being either (i) a zero mode, or (ii) a characteristic class c n (V )
of

a vectorial bundle V M built on space-time. A new topological limit of the theory has

therefore been defined [5], which is both non-trivial and no longer based on H = 0 but on ß = 0

and hence independent of H. ß plays the role of a coupling constant, such that it exists an infinite

number of states not interacting with each other and independent of H. The point is that for ß =

0, the action S is projected onto a non trivial minimum, corresponding to the self-duality

condition R = ± R * . Then, the minimum of the action is not zero (as it is in the trivial case) but

139
 has a non-trivial (invariant) value. But in this case, the field configuration is necessarily

Euclidean and defines a gravitational instanton, i.e. a topological configuration.

Let's recall that the ordinary topological limit of quantum field theory is described by the

Witten invariant

Zn = Tr(-1)n (7)

n being the zero energy states of the theory, for example the fermion number in supersymmetric
theories [20]. Equ. 7 is given by the limit of the partition function Z = Tr(-1)n e -ßH for zero (or
invariant) values of H. Here, Zn Tr( 1) n e ßH is simply the partition function at temperature
1
connected with a set described by the density matrix Q = (-1) n e ßH .
tel-00001503, version 1 - 24 Jul 2002

Now, we propose to extend Witten's results into the context of N=2 supergravity [21]. In such a
context, we analyze the zero mode of the space-time scale (i.e. the ß 0 limit). Hence, as
observed hereabove and according to the results of [5], at the Planck scale, the space-time, in
thermodynamical equilibrium, can be considered as submitted to the "KMS condition" [12].
Therefore, within the limits of the KMS strip - i.e. between the scale zero (ß = 0) and the Planck
scale (ß = Planck )- the time like direction of the metric might be viewed as holomorphic. The
corresponding quantum fluctuation of the signature of the metric from the Lorentzian form to the
Euclidean one and conversely has been exhibited in [5] and can be described by the symmetric
homogeneous space

SO(3,1) SO(4)
h =
SO(3)

SO(3) being diagonally embedded in SO(3, 1) SO(4). In terms of metric field, to such a
superposition state might correspond a "monopoles+instantons" configuration, the action of the
superposed theory being of the general form :

1 ~ g 2( ) 2
d4 x R d 4 x Tr R 2
m+i 2 R R R D
g ( ) 2

140
 According to the results of E. Kiritsis et C. Kounnas on a class of heterotic ground states in 4D [5],

equ.(11) can be true in certain classes of N = 2 supergravities (resulting of spontaneous breakdown of

N = 4). Such Lagrangians contain a scalar potential representing the kinetic terms and a Kähler

potential [21]. This type of action results typically from the Instanton-Monopole symmetry of duality

found in [5] and recently verified by C.P. Bacchas, P. Bain and M.B. Green in the context of string

theory [22]. Namely, 4-instantons (Euclidean metric) and (3+1)-monopôles (Lorentzian metric) are
1
connected by a complex dilaton field (or S-field) [21] of the form i playing the role of
g2

coupling constant of the superposed theory. So, on the infrared limit (ß Planck ) corresponding to g
1
R2
g 2( )
, the instanton contribution to the functional integral, given by e is suppressed. This
tel-00001503, version 1 - 24 Jul 2002

is a well-known result, demonstrated a long time ago [7]. On the contrary, it has been pointed in [5]

that the monopole configuration is well defined on this limit. So, the "superposed action" is reduced

to :
g 2( ) 2 2
i+m m = R D
2
corresponding to a 4D Lorentzian configuration of the metric on the infrared limit. Conversely, on the
ultraviolet limit ß 0, it has been showed, again in [5], that monopole contribution is suppressed

from the functional integral, just like the perturbative part of the instantonic action (tunneling effect).
~
However, one can easily observe that the topological charge Q = d 4 x Tr R R is conserved in

m+i , so that on the ß 0 limit, the theory contains only a Riemannian topological contribution

coming from zero size instantons [5]. Of course, for 0 < ß < Planck , the action takes the general

"superposed" form m+i mixing Lorentzian and Euclidean states of the metric field.

Now, we shall consider the partition function Zs describing this generalized symmetry of all
possible states (Lorentzian or Euclidean) of the metric beyond Planck :

Zs = Tr(-1)s e ßc H (8)

141
 with ßc and s representing the number of instantons of the theory. According to N=2
supergravity properties, H can be adequately described by the square of the Dirac operator D 2 .
Now, let's consider the scale ß = 0. Equ.(8) becomes :

Zs = Tr(-1)S

This new index, isomorphic to the Witten invariant Z = Tr(-1)n, can be explicitly associated
with the initial singularity of the pre-space-time, reached for the value ß = 0 (zero mode of the
scale) of the metric states partition function. One can therefore extend the iso-dimensional
monopole / instanton duality by suggesting that such a duality symmetry links the BRST
cohomology ring (physical sector of the theory) and the cohomology ring of the instanton moduli
space (topological sector). The BRST cohomology groups [21] having the generic form
tel-00001503, version 1 - 24 Jul 2002

(g)
(g) ker QBRST
H BRST = (g 1) (9)
imQBRST

we then consider that the topological field theory allows the injection of rings :

Uk dk
HgBRST (k) (i) (k)
HBRST g 0 H mod i 0H mod (10)

which provides an injective path of the physical mode into the topological mode. In terms of
(k ) ( k)
observables and homology cycles Hi
Oi mod in moduli space mod of configurations of the
gravitational instanton type [ (x)] on the gravitational fields of the theory, we bring out the
equivalence :

O1O 2 ... O n # (H1 H2 ... Hn ) (11)

where the physical sector of the theory is described by the observables and the dual sector, of
Oi

(k )
the topological type, by the homology cyclesHi mod . The oscillation of the signature of the
metric between physical and topological sectors is then induced by the
divergence Uk j d 4 x of the ghost flow [16][15] . When U = 0, as there exists no
j

embedding space for moduli space, we suggest that the theory is then projected in the Coulomb
( k)
branch, at the origin of mod , on a singular instanton of zero size [4] which we identify to
space-time at zero scale. The theory is ramified on the purely topological sector , the Hi

corresponding signature at this sector being Euclidean (++++).

142
 From this viewpoint, the image of Euclidean zero symmetry, described by the non-broken gauge
group of type SU(2) SU(2), is given by the first Donaldson invariant [7] associated with the
( k)
existence of a "topological amplitude" characterizing the theory. When dim mod 0, the
Donaldson invariants are given by the correlation function of the theory :

r r
S ( k)
Z( 1 ... r) DXe Wki Wki (Dim mod 0)
i 1 i 1
(12)

where Wki is a k form, for 0 k 4. In particular, as observed by E. Witten [3], if is a k

dimensional homology cycle on the manifold M, then the integral I( ) Wk is BRST

invariant, since Q,I Wk 0.

tel-00001503, version 1 - 24 Jul 2002

( k)
Now, what happens when dim mod = 0 ? The solution is in the correspondence between the

Donaldson invariants on 4D manifolds and the Floer homology groups on 3D manifolds [23]. In

fact, Donaldson invariants amount to the calculus of the partition function Z, expressed as an

algebraic sum over the instantons :

Z Z = ( 1)n i
(k)
mod 0 i

( k)
Donaldson has shown on topological grounds [7] that when dim mod = 0, then ( 1)n i is a
i
non-polynomial topological invariant, reduced to an integer. We find the same result starting
from T Q, .

Now, let's go further into the understanding of the zero scale limit of space-time. We claim that,
even if this limit is singular within the context of (perturbative) quantum field theory, it can be
resolved in the (non-perturbative) dual theory, i.e., the topological field theory. As noticed
previously, when ß 0, the partition function

Zs = Tr(-1)S e ßc H (13)

143
 describing the symmetry of all the (Lorentzian + Euclidean) states of the metric is reduced to

Z = Tr(-1)S (14)
ß= 0

What is here the meaning of Tr (-1)S ? In fact , at least in the context of supersymmetric theories

[20], it can be seen as the index of an operator acting on the Hilbert space . Dividing in
monopole and instanton sub-spaces = + i and Q being a generator of supersymmetry,
it follows from (21) that Q = 0 and Q* = 0. Then Tr (-1)S = Ker Q - Ker Q* so that

as topological index, Tr (-1)S is invariant under continuous deformations of parameters which do

not modify the asymptotic behavior of the Hamiltonian H at high energy. At the scale ß = 0, H is

given by H = dd* +d*d, the space of zero energy states corresponding to the set of even
tel-00001503, version 1 - 24 Jul 2002

harmonic forms on Mn:

n
Tr (-1)S = (M) = ( 1)k bk (15)
k 0

where bi is the ith Betti number and (M) the Euler-Poincaré characteristic of M. Obviously,
= Tr (-1)S is independent of ß, the sole contributions to coming from the topological sector of
zero energy. Consequently, is a topological invariant, i.e. the first Donaldson invariant. The
coupling constant g being dimensional, the limit ß = 0 implies = 0 and corresponds to the
sector of zero size instantons. Therefore, (15) is a topological index, corresponding to the
accumulation point of the zero size instantons of the theory. Let's note that all instantonic (i.e.
Euclidean) states of g , given by the topological charge of the singular gravitational instanton
are equivalent at zero scale. We call "symmetry zero" the generalized symmetry characterizing the
singular zero scale.

To conclude this part, we suggest as a geometric model of the instanton the ball B4 bounded by
the sphere S3. The so-called "propagation" of the solution depends then on the support of the
gravitational instanton : in the region of zero limit, there exists an accumulation of topological
charge above the singular point S0 such that the topological charge density RR* ; in the
dual situation, corresponding to the fundamental state, the support of the instanton is extended to
infinity and RR* 0. The "transition" from zero to infinity is then described by the
conformal transformations of the sphere (see in § 4 the Möbius group action (4.5)).

144
 From this point of view, the Machian interaction evidenced between the pendulum and the
global universe could be exactly described by the interaction between and the zero size
instanton - almost like a renormalisation transformation. We therefore propose a new formulation
of the Mach's principle in the context of topological field theory :

3.3 Topological Mach's principle. The topological amplitudes associated with the
propagation of the topological charge of the singular zero size gravitational instanton
corresponding to the Initial Singularity of space-time determines the inertial behavior of local
masses.

In the following paragraph, we put forward a number of natural arguments to establish

topological Mach's principle. In particular, we consider that the propagation of information
characterizing the Initial Singularity of space-time can conveniently be described by the
tel-00001503, version 1 - 24 Jul 2002

topological amplitude associated with the S0 singular zero size gravitational instanton.

4. SINGULAR ZERO SIZE INSTANTON AND TOPOLOGICAL

INTERACTION

The principal argument demonstrating the existence of an interaction between and the charge Q
of S0 lies in the fact that the topological charge of the instanton is entirely determined by the
asymptotic behavior of the gauge field (A in the Yang et Mills case and g in N = 2
supergravity). The field F, non-zero in E4 = 4 instanton space, cancels at the boundary E4 =
S3, the gauge potential becoming a pure gauge on the boundary. Hence, we suggest that E4 =
S3 represents the curved and compact three-dimensional physical space 3 within which the
rotation plane is inscribed. We draw from this a relation of topological type between the
topological charge defined by E4 and the fixed orientation of the oscillation plane 2 of ,
also defined by E4 = S3.

Proposition 4.1 The topological charge Q of the Yang-Mills instanton is completely

determined by the asymptotic behavior of the gauge field A , at the boundary E4 represented
by the three-dimensional sphere S 3 where the gauge potential becomes a pure gauge. E4 = S 3
represents the curved and compact three-dimensional physical space 3 in which the plane of
rotation 2 of is embedded.

Demonstration The Lagrangian of the Yang and Mills theory in Euclidean space-time has the
form :

145
 1
L= 2 d 4 xF F (16)
4g

abc
with being the structure constants of the gauge group SU(2) :

Fa abc b
A A A Ac (17)

The classical Yang and Mills action can then be written :

1 8 2
F˜ a
4 a 2
S 2 d x (F ) Q (18)
8g g2

1
with F˜ a Fa ,
tel-00001503, version 1 - 24 Jul 2002

Q being the topological charge (or Pontryagin index) of the

2
instanton:

1
Q= 2 d 4 x F a F˜ a (19)
32

As the action of the configuration must be finite, on spheres at infinity of radius r , the field
Fa must fall to 0 more rapidly than r -3 :

Fa ( x r) 0 (20)

such that A a must be a pure gauge at infinity (r ):

A aT a iU( x) U 1 (x) (21)

T a being the generators of the gauge group, i.e. in the case of the fundamental representation
1 a
with Pauli matrices a , T a = for SU(2). The element U SU(2). Considering the
2
charge Q, it is possible to represent it by an integral over the total derivative :

1
Q= 2 d 4x K (22)
32

that is to say, according to the Gauss theorem :

146
 1
Q= 2 K dS (23)
32 x r

where the vector K represents the Chern-Simons current :

1
(A a Aa abc
K =2 A a Ab Ac ) (24)
3

From equ. (24) we conclude that the topological charge Q is thus entirely determined by the
asymptotic behavior of the field A a . As established in [21], Q depends solely on the global
properties of the function A a ( x r) . Indeed, at infinity we have :

x
tel-00001503, version 1 - 24 Jul 2002

F 0

but this is not (necessarily) the case for the gauge potential A a which becomes a pure gauge :

x
A(x) U (x) U 1 (x) (25)

The gauge elements U(x) SU(2), x S 3 are such that :

U = A i B, A 2 B2 1 (26)

and U( x) represents :

U: S3 SU(2) S3 (27)

where we find the applications of the sphere S 3 representing the compact physical space E3,
boundary of the space E4, on the isotopic space of SU(2), equally isomorphic to S 3 . We draw
the identification of S3, boundary of the 4-dimensional instanton solution to physical space, of
the double embedding of SU(2) in SL(2, C) - the universal convening of the Lorentz group -
and in SU(2) SU(2), the convening of SO(4). As SU(2) S3, we therefore propose to
interpret S3 both (i) as the 3-dimensional boundary of the 4-dimensional instanton
4
Euclidean solution B and (ii) as the 3-dimensional boundary of space-time . From this
3
S is designated by 3(S 3 ) and is such that :
3
identification, the application S

147
 3
3(S ) = (28)

such that the applications S 3 S3 are classified according to the homotopy classes characterized
by integers, in our case n = 1. Thus, the 2-dimensional plane of oscillation 2 of Foucault's
pendulum is embedded in the 3-dimensional physical space corresponding precisely to S 3 . As
the topological charge Q of the instanton is determined by the behavior of the gauge field on S 3 ,
it follows from this that 2 is determined by Q, as required. qed

We have established the relation between Q and 2 in the case of Yang and Mills instantons.
One can extend this result to supergravity, insofar as, in the context of non-linear curvature
theories, the field R is, like F, asymptotically free. The action of the gravitational instanton
becomes :
tel-00001503, version 1 - 24 Jul 2002

1 ~
S= 2 d4 x R R d 4 x Tr R R (29)
g ( )

the topological term being :

~
Q= d 4 x Tr R R (30)

In this case, the surface term K, associated with the Chern-Simons topological flow, becomes :

2
Q30 ( , R) ~ tr 3
d
3
2
Q30 ( , R) ~ tr 3
d (31)
3

representing the Christophel symbols, the spinorial connections and R the curvature of the
4-manifold. In putting R = d or R = d , we therefore find again the result of the
first part of (4.1.) in supergravity. Concerning the Chern-Simons form, we recall that Witten has
shown [24] that general relativity in dimension (2+1) is equivalent to Chern-Simons topological
theory in dimension 3. The Chern-Simons Lagrangian is :

2
h( A) Tr(A dA) A A A) (32)
3
Hence, the action of the theory of gravity in dimension (2+1) has the form :

148
 abc ijk a bc
L ei R jk ( )
(33)
Rbc bc be ec
jk j k j k (j k)

Analysis of the two Lagrangians (32) and (33) shows the equivalence between gravity (2+1) and
the Chern-Simon theory at the limit of zero scale associated with zero energy modes. In
dimension 4, we find then that at the limit of zero scale, Lorentzian (3+1) field theory must be
replaced by Donaldson topological field theory (4,0) [7].

One can therefore consider that the propagation of the initial singularity is induced by the
existence of a topological amplitude – the charge of the zero size gravitational instanton, i.e.
Q d 4 x R R˜ , - detectable at the boundary S3 of the singular gravitational instanton
provided with the topology of the B4 Euclidean ball of dimension D=4. As noticed in (1), the
pseudo-observables in 4D Euclidean space are connected with cycles i of the B4 4-manifold
tel-00001503, version 1 - 24 Jul 2002

(Donaldson application).

Now, we return to the inertial interaction. In the following conjecture, we suggest that the source
of inertia is not, strictly speaking, physical but has a topological content, linked to the
topological charge Q=1 of the singular gravitational instanton.

Conjecture 4.2. The inertial interaction may be interpreted as a topological interaction, of

which the source is the topological charge of the zero size singular gravitational instanton.

Elements of demonstration. We have shown that the total inertial force likely to contribute
in the case where the entire Universe would be submitted to an acceleration a in respect to a
'
given object, could be obtained by summing the forces F12 on masses other than m itself :

GM
Finertielle ma
c 2r (34)
If M represents the totality of the mass of the Universe, a good approximation of
GM
gives us :
c 2r
GM
Ftotale 2 1 (35)
univers c r
The total inertial interaction Ftotale is therefore of the order of unity and represents a topological
invariant -i.e. invariant of gravitational gauge and scale invariant -. In all points of space, the
contribution of Ftotale of is identical. Hence, we suggest that the total inertia Ftotale has its
origin in the topological charge Q :

149
 1 4 ˜ = 1 (36)
Q 2 d xR R
32
Indeed, we have seen that the topological charge Q of the instanton is entirely determined by the
asymptotic behavior of the gauge field at the boundary E 4 represented by the three-dimensional
sphere S3 where the gauge potential becomes a pure gauge. We show then that at the zero scale
limit, Ftotale is isomorphic to Q :
Ftotale Q
GM
At the zero scale limit, 2 is identifiable to the Schwartzshild radius lS of space-time
univers c r
quotiented by the radius r of the configuration in region of zero scale, i.e. :
GM lS
2 = (37)
univers c r univers r
GM
We therefore draw from (37) that represents the Gaussian curvature of the manifold, which
c2 r
tel-00001503, version 1 - 24 Jul 2002

gives :
GM
= Kr (38)
c2 r
Kr being the Gaussian curvature of the geodesic surface (r, ). Hence, at the Planck scale, we
must extend Einstein's gravitational term R towards an asymptotically free non-linear theory.
Our approach is that the presence of terms in R2 in the Lagrangian is the result of a superposition
of two scales of quantum gravity : (i) the gravitational scale - i.e. macroscopic scale -
characterized by the presence of an interaction in R and (ii) the quantum scale - i.e. microscopic
scale - characterized by the presence of dual gravity terms, in R* = iR. The curvature term Kr
takes then the new form in supergravity :
Kr (Planck) R R

lS
such that is identifiable to
univers r

= d 4x R R (39)

As the theory is auto-dual, one has therefore R = R* , and becomes :

= d 4 x R R˜ (40)

which, by construction, is isomorphic to the topological charge Q of the gravitational instanton,

as required. qed

150
 From the hereabove results, we can conclude that the inertial interaction might admit the
topological charge of the zero size gravitational instanton as source. The topological nature of
the inertial interaction could explains (i) its global invariance properties and (ii) its instantaneous
propagation between two points in space-time .

We complete the conjecture (4.2) in specifying why the plane of oscillation 2 should be static.

Corollary 4.3. The static behavior of the plane of oscillation 2 of the pendulum should
be induced by the structure of topological invariant of Q :

Elements of demonstration. Instanton theory has established that Q is a topological

invariant [6]. In our case, we have Q = 1. The resulting vacuum topology - and, as a
consequence, that of physical space S3, boundary of E4, in which 2 is embedded - is also
tel-00001503, version 1 - 24 Jul 2002

invariant. Indeed, the invariance of the topological charge implies the invariance of the
topological structure of S3, i.e. of the three-dimensional physical space in the course of
temporal evolution. It results from this that 2 S3 is itself invariant - i.e. static -.
Furthermore, the static character of 2 implies that the underlying symmetry is Euclidean and
can be described by the action of the group :

G S = SU(2) SU(2) (41)

G S being, precisely, the symmetry group describing the instanton configuration characterized
by the topological charge Q, as required. qed

We now show that a good model of the propagation of the topological interaction Int top can be
given by the conformal transformations Conf (S3) of the sphere S3.

4.4 Conformal group of S3

We propose to describe the transformations Conf (S3) by the Möbius group [25], defined from
the inversion of S3. Recalling that the pole inversion c, of strength a * is the application

i ic , :X c X where c is defined, in X vectorialised in c , by i(x) .x . From

x2
this :

151
 Definition 4.5 An inversion of S 3 is meaning all restriction to S 3 of an inversion of R 4
conserving S 3. The Möbius group Möb(3) of S 3 is the sub-group of the S 3 bijection group
induced by inversions of S 3.

At this point, we can establish a link between the application defining the topological charge of
the instanton and the similitudes of 4. We recall first that the group GL(n k) operates on M(n,
k) by similitude :

(P, M) P.M = P.M.P 1 (42)

The equivariant canonical bijection associated with the matrix M is then a diffeomorphism,
representing the similitude class of M :
tel-00001503, version 1 - 24 Jul 2002

(M) : GL(n, ) / Z(M) O (M) (43)

which leads us to the important definition, taken from the reduction of endomorphisms [22] :

Definition 4.6 The application : M (n, ) O (M) describing the similitude class of M is
given by :

X exp(X).M.exp(-X) (44)

From this definition, we establish in the two following propositions that the application defining
the topological charge of the instanton belongs to the conformal group of S3.

Proposition 4.7 For all similitude h Sim ( 3), the application defining the topological
charge of the instanton, i.e. ƒ : S 3 S 3 , defined by ƒ(n) = n and ƒ = g-1 o h o g on S 3 n
belongs to Möb (3) ).

Demonstration Let n be the north pole and s the south pole of the sphere S3. We can then
show that g-1 o h o g, conveniently extended to the whole of S3 , is in Möb(3) if h is an inversion
or a hyperplane symmetry. Yet, the h induce the existence of the group of similitudes Sim ( 3).
Indeed, considering , it has been shown [25] that if the kernel

Ker ( Id X ) 0 (45)

152
 then ƒ admits a unique fixed point corresponding to its center. Moreover, ƒ Sim (X) Is (X),
there exists a unique X such that ƒ( ) = . The point is the center of the similitude ƒ
and one can write :

h g = g h, h H , et g Is (X ) (46)

The above assumes convenient extensions, the most immediate consisting of attaching a point at
infinity on 3 and extending g to S3 3 by g(n) = . Thus, following [25], it exists in Möb
(3) the applications

1
g H0, g e H 0, e (47)

associated with vectorial homothetias of 3. If > 1, the application ƒ admits the north pole
tel-00001503, version 1 - 24 Jul 2002

n
as an attractor and the south pole as a repeller, i.e. the iterates (n ) making all points of
S3 s converge towards n. The only point of S3 out of the attraction of n is the south pole s.
qed

We show now that Möb(3) is the conformal group Conf(S3) of S3. Letting Conf (S3) describe
the scale invariance (i.e. the conformal invariance) of the sphere identified here (following the
inclusion S3 SL(2, C)) to physical space 3 compactification.

Proposition 4.8 Let Möb± (3) = Conf ± (S3). the radius r 0 of S 3 engendering S3r 0 ,
and ƒ Möb(3), then S3r 0 belongs to the bundle (S3) of spheres S 3. Reciprocally, a
bijection of S 3 verifying this property necessarily belongs to Möb(3) . The group Möb(3)
presents a natural isomorphism with PO( ) of the quadric of equation

4
q x2 x 25 .
i 1

Demonstration Let ic, be an inversion of X of dimension n. Let its derivative be i '( x )

composed of the vectorial hyperplane symmetry x and the homothetia of ratio / x 2 . We
n
show then that i '( x ) in all x X c is a direct similitude for < 0 and an indirect
n
similitude for > 0. In fact, i '( x ) conserves the right and right-oriented angles. Since the
product of two conformal applications is conformal, then Möb(3) Conf(S3). Reciprocally, as
Möb(3) Conf (S3) is transitive on S3, then ƒ Conf (S3) leaves the north pole n fixed.
According to the stereographic projection g of n, for ƒ(n) = n, we get :

153
 1
g g Conf ( 3) (48)

g and ƒ being conformal. Applying Liouville's similitudes theorem, we have

1
g g Sim ( 3) ƒ Möb(3) (49)

It follows from the inversion properties [25] that ƒ( ) conserves the structure of the sphere S3
when the radius r 0. Reciprocally, on putting ƒ(n) = n, g g 1 transforms the 3 (half)
lines into (half) lines, such that S3r 0 belongs to the bundle (S3) of spheres S3.

Finally, it has been established in [25] that Möb(3) = INCORPORER "Equation.3" \*

mergeformat , i.e. the Möbius group of S3 corresponds to the restriction of the group
tel-00001503, version 1 - 24 Jul 2002

PO(_) on im (_). qed

We now conjecture that the plane of oscillation of F conserves the initial singularity S for
inertial reference point, whatever the orientation of this plane in physical space R3

Conjecture 4.9 Whatever the orientation in physical space R3 of the plane of oscillation 2 of
the pendulum F, this 2-dimensional plane necessarily intersects the initial singularity S, i.e. 2
is always aligned on S .

Elements of demonstration We have established the identification between physical space R3

compactification and S3, boundary of space-time and equally boundary of the singular
gravitational instanton solution. Each orientation of the plane of oscillation 2 corresponds
therefore to an orientation in S3. We have also established that S3 can be identified to physical
space :

S3 3

such that the three-dimensional information coming from physical 3-geometry is concentrated on
the 3-surface S3 but is not detectable in the interior of the sphere. Thus, the conformal
invariance of S3 implies that the temporal direction x4 is necessarily orthogonal to the tangent
space in a point of S3. Putting this point as the south pole s of S3, we have shown above that
there exists in Möb (3) the applications

154
 1
g H0, g e H 0, e (50)

associated with the vectorial homothetias of 3. If > 1, the application ƒ admits the north
pole as attractor and the south pole as repeller, i.e. the iterations n (n ) make all points in
S3 s converge towards n. The only point in S3 escaping the attraction of n is the south pole
s. The north pole n of the 3-sphere is therefore the fixed point of the conformal transformation
Conf(S3) and the temporal direction x4, orthogonal to the tangential plane of the sphere in s,
intersects necessarily the center O of S3 as well as the north pole n. Since, by construction, the
plane of oscillation 2 contains x4 , then :

x4 2

it follows that the plane of oscillation 2 is orthogonal to the tangential plane of the sphere at the
tel-00001503, version 1 - 24 Jul 2002

south pole s and meets therefore necessarily the center O of S3 as well as the north pole n, the
singular attractor point of S3. qed

The above conjecture suggests therefore that the symmetry of rotation in 3 (exhibited by the
plane of oscillation of Foucault's pendulum) is explicitly linked to the symmetry of the zero
instanton configuration, SU(2) S3 being a sub-group of both SU(2) SU(2) and SL(2, C).
Once the identification Initial Singularity / instanton zero is admitted, the above approach
allows us reasonably to consider that, whatever the orientation of the plane of the pendulum in
physical space, this plane remains necessarily aligned with the singular origin of space-time,
identified here to the singular origin n of the sphere S3, n being the north pole of the 3-sphère,
the unique fixed point of the system. Indeed all possible orientations in physical space of the
plane of oscillation of the pendulum are given by all possible orthogonal directions to the tangent
plane at S3. We obtain the different orientations of 2 in 3 by making the south pole s "turn"
on the 3-surface S3, this rotation conserving the alignment between s, O and n in the same plane
2.

We draw from the above that whatever the orientation, the plane of oscillation of Foucault's
pendulum is necessarily aligned with the initial singularity marking the origin of physical space
S3, that of Euclidean space E4 (described by the family of instantons I of whatever radius ß)
and, finally, that of Lorentzian space-time M4.

The angular invariance of 2 comes therefore in fine from the fact that the direction x4
represents equally the fourth direction (in ß = i t imaginary time) of the Euclidean configuration

155
 of the type instanton E4 bounded by S3, such that at each point s of S3, 2 cuts O and the
origin of E4 represented by the north pole n. We suggest then that this interpretation of x4 as
imaginary time explains in a non-trivial way the nature of the inertial force as well as its
instantaneous propagation from one point to another in space-time .

5. DIRAC SIGNAL AND INITIAL SINGULARITY

We complete this paper by suggesting a subsidiary argument concerning the propagation of

topological type of a "causal information" from the singular point S, the origin of the system, to
the boundary of space-time. In the following, we consider that the topological signal at the
origin represents a Dirac shock and is sent to infinity - i.e. to S3, boundary of space-time .

In effect, the initial singularity can be interpreted as a causal signal giving rise to, at instant zero,
tel-00001503, version 1 - 24 Jul 2002

a shock at the origin corresponding to a Dirac signal [26]. The shock at the origin, or Imp(t),
distributed at the zero scale of space-time, must satisfy :

(i) t R, (t) 0
0 si t 0
(ii) Imp(t) =
si t 0 (51)

(iii ) Imp (t)dt = 1

R

The unity signal at the singular origin can then be considered as an ideal signal, of causal type.

Proposition 5.1 The initial singularity, distribution of zero support, can be interpreted as a
Dirac signal. It follows from this that the Fourier transform is a function that can be extended in
the complex plane under the form of a holomorphic entire function, or bilateral Laplace
transform.

Proof It has been shown in [5] that the initial singularity can be interpreted as a singular zero
size gravitational instanton, configuration built by E. Witten in [3]. The support of all associated
distributions is therefore reduced to the singular point . The function associated with the
curvature is therefore a Dirac distribution, such that, as established in [26], its Fourier transform
is holomorphic and given by :

2i 2i
ƒ[ ]= (x) , e [e ] 0 (52)

156
 or, when the scale ß (or the time t ) of the theory is zero :

ƒ[ ] = 1 (53)

becomes a real and even distribution which, insofar as ƒ[1] = must satisfy :

ƒ[ ] = [ ]=1 (54)

The holomorphic function resulting from the Fourier transform of the function of zero support
can equally be written in the form of a bilateral Laplace transform ƒ(ß) :
ßc H
ƒ(ßc) = (H)e dH (55)

where ß is a complex variable. By decomposing ß into real and imaginary parts, i.e. ßc = ßr +
ißi , we observe that, for ßr = 0 :
tel-00001503, version 1 - 24 Jul 2002

ißH
ƒ(ißi ) (H)e dH (56)

which, up to the change of variable, is the Fourier transform of ƒ(H). For a fixed ßr, we have:

ßH ißH
ƒ(ßr + ißi ) = e (H ) e dH (57)

ßH
which is the FT of ƒ(H) e . Deriving under the sum the expression for ƒ(ßc) :

ßc H
d/dßc ƒ(ßc) = ßc (H) e dH
(58)

or, in general :

(H )m ( H)e ßc H
dm/d(ßc)m ƒ(ßc) = dH (59)

and the summability abscissae of (H)m (H) are the same as those of ƒ(H), such that ƒ(ßc) is
indefinitely derivable for all values of ßc where ƒ(ßc) exists. ƒ(ßr + ißi ) is therefore holomorphic
in all the band of summability, i.e. for all values situated to the right of 0 in the complex plane
formed by ßr > 0 and ßi > 0. The function ßc is therefore analytic in this domain of the
complex plane. As the Dirac signal at the origin has for support the point , its FT describes
therefore an impulsional response non-decreasing at infinity. qed

To understand the necessarily non-compact character of the impulsional response giving the
evolution of the system, we complete the above proposition by the following corollary :

157
 Corollary 5.2 The Fourier transform of the singular distribution s of punctual support
describing the impulsional response of the space-time system cannot be of compact support.

Demonstration Let's consider the Dirac signal s E( ) and ƒ = ˆ s . The function ƒ is

analytic on , insofar as it is the trace of a holomorphic function on . We suppose that ƒ is of
compact support, hence ƒ cancels itself out on an non-empty open set of . Thus, if a general
analytical function on is zero on an non-empty open set, then it is identically zero. It follows
from this that ƒ cannot be of compact support. qed

The above results suggest in fact that the (topological) interaction here considered is ergodic. As
the Dirac signal at the origin has for support the point , its FT describes therefore an impulsion
response non-decreasing at infinity. Indeed, (i) the behavior of 2 is scale invariant and (ii) the
zero size singular gravitational instanton characterizing, according to [4], the initial singularity,
tel-00001503, version 1 - 24 Jul 2002

represents a critical point S0 in the system formed by the pre- space-time manifold and , such
that the correlation length of the system . From this viewpoint, the interaction Int top is
subject to the action of a renormalisation group Gn assuring the scale invariance of the system.

6. DISCUSSION

In this paper, taking the example of Foucault's experiment, we have suggested a new approach
concerning two open problems :

(i) the problem of the invariance of the inertial interaction onto all points in space-time ;

(ii) the problem of the "instantaneous propagation" of inertia from one point to another in space-
time - i.e. the Machian principle according to which the inertial reference frame defined by local
physics coincides with the reference frame in which distant objects are at rest. It follows that the
masses distributed most distantly in the universe determine the inertial behavior of local masses.

As a result of the hereabove research, our point of view is that the problem of inertia, well
formulated in the context of Mach's principle, cannot be resolved by ordinary field theory.
Indeed, as suggested in [5], beyond the Planck scale, quantum field theory must be analytically
extended towards topological field theory. In such a context, the initial singularity can be viewed
as a singular zero size gravitational instanton. This point like solution, endowed with an

158
 3, 1 4
Euclidean metric, corresponds to the origin of the topological quotient space top=
SO(3)
describing the q-deformation (quantum {Lorentzian Euclidean} superposition) of the metric
of the (pre)space-time between the scale zero and the Planck scale [5].

Our conclusion is then that onto zero scale, the topological charge Q = 1 of the singular zero
size instanton might represent the source of the global topological inertial interaction. This could
be tested by the angular invariance of the plane of oscillation of Foucault's pendulum or by the
Thirring- Lense effect.

In a subsequent paper, we will consider that this result can be reinforced by the hypothesis of a
correlation between the singular scale (t0, x0) and the macroscopic scale (t, x). We begin from
the observation of the cosmological radiation at 2.7 ° K and draw from this the existence of a
tel-00001503, version 1 - 24 Jul 2002

thermal Green function - or Euclidean Green function GE - describing the correlation between
the zero scale (ß = 0) and the macroscopic scale of space-time . Such an approach suggests
the topological nature of the interaction between zero and macroscopic scales. Indeed, the
correlation described by

GE (t0 , x 0 ; t, x) (t0 , x 0 ) (t, x) e d (60)

is such that all points P of space-time are correlated - by an Euclidean path - to the singular
point S0. Insofar as the path between S0 and P is Euclidean - which is the case since the space-
time system, considered at the non-zero temperature T = 2.7° K is the concern of statistical
mechanics - the interaction between S0 and P depends only on the boundary conditions and is
thus purely topological. Following this, we specify in this paper to come the notion of
"Euclidean propagation" of inertia, according to the flow of weights of the algebra of states
describing space-time in the region of the initial singularity. Starting from the von Neumann
algebra describing the zero size instantonic state, we conjecture, according to [16] that the
sole data from the algebra implies the existence of a "pseudo-dynamic" associated with and
characterized by the flow of weights of . Such a flow assures the propagation of the
topological charge Q of the zero instanton. In agreement with the results of Connes [16], the
homomorphism defining the canonical dynamic is such that : Out
Aut
= , this invariant having an intrinsic description in terms of flow of weights of . We
Int
suggest then that this "intrinsic dynamic" is based on the semi-group of automorphisms :

159
 2 2
(M) = e -ß D M e ß D (61)

corresponding to the evolution in imaginary time i t of the state M - i.e. to the expansion of the
space of states . This expansion of is indexed by increasing values of ß, the radius of . As
stated in [5], equ. (61) describes the flow of weights of the system and this flow being ergodic,
ß is necessarily increasing in the interval [0, ].

Then, we claim that a satisfactory test of the topological nature of the interaction existing between
the zero size singular gravitational instanton and local systems should be provided by the angular
invariance of the plane of oscillation of Foucault's pendulum. We have shown that Möb(3) is the
conformal group Conf (S3) of S3. Conf(S3) describes the scale invariance (i.e. conformal
invariance) of the sphere identified here, following the inclusion S3 SL(2, C), to physical
3
tel-00001503, version 1 - 24 Jul 2002

space compactification. We have then suggested that the flow of weights of the algebra M
giving the modular flow t (M) on S3 belongs to the class of similitude S3.

Finally, an interesting consequence of the above approach is that it allows us to establish an

explicit relation between the automorphism semi-group of the algebra of states A and the
renormalisation semi-group of the theory [27]. Introduced by Wilson then by Kadanoff [28], the
renormalisation program - in particular the renormalization group - allows us to encompass in a
unique formalism the different scales of the theory. Observation shows that the behavior of
Foucault's pendulum, notably the angular invariance of the plane of oscillation, is scale
invariant. Everything occurs therefore as if the dynamic of were subject to the action of a
renormalization group GR, the group whose structure we define below. The calculation of the
correlation length between two localized variables at different points takes then the form,
considering the variables n :

ß / ß0
n, m n m e (62)

where the distance ß depends on the number of points on the lattice between n et m. At zero
scale, the correlation length becomes, considering the coupling g0 :

ß0 (g0 ) (63)

160
 The correlation length is infinite, so that at zero scale, there exists an instantaneous interaction
between the point S0 representing the initial singularity of space-time and the boundary at
infinity of the 4-manifold, representing the 3-dimensional physical space.

Finally, all the hereabove results seem to confer a certain relevance (i) to the formulation of the
so-called "topological Mach's Principle" and, more generally (ii) to the "Singularity Principle".
Hopefully, this preliminary approach might open some new and interesting perspectives on the
origin of space-time and on some other open questions.
tel-00001503, version 1 - 24 Jul 2002

REFERENCES

[1] MACH E. La Mécanique Ed. Gabay , Paris (1989)

[2] VIGIER J.P. Foundations of Physics, 25 (1995) 1461

[3] WITTEN E. Commun. Math. Phys. 117 (1988) 353-386

[4] WITTEN E. hep-th / 9511030 (1995)

[5] BOGDANOV G. Ph.D. thesis Bourgogne University (1999)

[6] SHIFMAN M. Instantons in Gauge Theories World Scientific Publishing (1994)

[7] DONALDSON S.K. KRONHEIMER P.B. The Geometry of Four Manifolds Oxford Univ. Press (1990)

[8] BRANS C. DICKE R.H. Phys. Rev. 124 (1961) 925

[9] BONDI H. SAMUEL J. gr-qc 9607009 (1996)

[10] BRILL D.R. gr-qc 9402002 (1994)

161
 [11] WEINBERG S. Phys. Rev. D 9 (1974) 3367-3377

[12] HAAG. R. HUGENHOLZ N. WINNINK M. Commun. Math. Phys. 5 (1967) 215

[13] ATICK J.J. WITTEN E. Nucl. Phys. B 310 (1988) 291

[14] ANTONIADIS I. DEREDINGER J.P. KOUNNAS C. hep-th 9908137 (1999)

[15] TAKESAKI M. Lecture Notes in Maths 128 Springer Verlag, New-York (1970)

[16] CONNES A. Non-Commutative Geometry Academic Press , London (1994)

[17] ARAKI H. WOODS E.J. A classification of factors Publ. Res. Inst. Math. Sci. A 4 (1968) 51

[18] CONNES A. TAKESAKI M. Tohoku Math. Journ. 29 (1977) 473

[19] DONALDSON S.K. Topology 29 (1990) 257-315

tel-00001503, version 1 - 24 Jul 2002

[20] WITTEN E. Nucl. Phys. B 202 (1982) 253

[21] FRE P. SORIANI P. The N=2 Wonderland. From Calabi-Yau Manifolds to Topological Field Theory World
Scientific Publishing (1995)

[22] BACCHAS C.P. BAIN P. GREEN M.B. Hep-th 990 3210 (1999)

[23] FLOER A. Commun. Math. Phys. 118, (1995) 215

[24] WITTEN E. Phys. Let. B 206 (1988) 601

[25] BERGER M. Geometrie Ed. Nathan, Paris (1990)

[26] RODDIER F. Distributions et Transformation de Fourier Ediscience, Paris (1993)

[27] SEZGIN P. van NIEUWENHUIZEN E. Phys. Rev. D 22 (1980) 301-307

[28] KADANOFF L.P. Phase Transition and Critical Phenomena 6 Harcourt&Cie, Academic Press (1976)

162