An Introduction To Quantum Gravity

An introduction to quantum gravity
arXiv:1108.3269v1 [hep-th] 16 Aug 2011
Giampiero Esposito,
INFN, Sezione di Napoli,
Complesso Universitario di Monte S. Angelo,
Via Cintia, Edificio 6, 80126 Napoli, Italy
August 17, 2011
Abstract
Quantum gravity was born as that branch of modern theoretical
physics that tries to unify its guiding principles, i.e., quantum me-
chanics and general relativity. Nowadays it is providing new insight
into the unification of all fundamental interactions, while giving rise to
new developments in mathematics. The various competing theories,
e.g. string theory and loop quantum gravity, have still to be checked
against observations. We review the classical and quantum founda-
tions necessary to study field-theory approaches to quantum gravity,
the passage from old to new unification in quantum field theory, canon-
ical quantum gravity, the use of functional integrals, the properties
of gravitational instantons, the use of spectral zeta-functions in the
quantum theory of the universe, Hawking radiation, some theoretical
achievements and some key experimental issues.
1 Introduction
The aim of theoretical physics is to provide a clear conceptual framework for
the wide variety of natural phenomena, so that not only are we able to make
accurate predictions to be checked against observations, but the underlying
mathematical structures of the world we live in can also become sufficiently
well understood by the scientific community. What are therefore the key
elements of a mathematical description of the physical world? Can we derive
1
all basic equations of theoretical physics from a set of symmetry principles?
What do they tell us about the origin and evolution of the universe? Why is
gravitation so peculiar with respect to all other fundamental interactions?
The above questions have received careful consideration over the last
decades, and have led, in particular, to several approaches to a theory aimed
at achieving a synthesis of quantum physics on the one hand, and general
relativity on the other hand. This remains, possibly, the most important
task of theoretical physics. In early work in the thirties, Rosenfeld [131, 132]
computed the gravitational self-energy of a photon in the lowest order of
perturbation theory, and obtained a quadratically divergent result. With
hindsight, one can say that Rosenfeld’s result implies merely a renormaliza-
tion of charge rather than a non-vanishing photon mass [40]. A few years after
Rosenfeld’s papers [131, 132], Bronstein realized that the limitation posed by
general relativity on the mass density radically distinguishes the theory from
quantum electrodynamics and would ultimately lead to the need to reject
Riemannian geometry and perhaps also to reject our ordinary concepts of
space and time [20, 135].
Indeed, since the merging of quantum theory and special relativity has
given rise to quantum field theory in Minkowski spacetime, while quantum
field theory and classical general relativity, taken without modifications, have
given rise to an incomplete scheme such as quantum field theory in curved
spacetime [65], which however predicts substantially novel features like Hawk-
ing radiation [87, 88], here outlined in section 7, one is led to ask what would
result from the “unification” of quantum field theory and gravitation, despite
the lack of a quantum gravity phenomenology in earth-based laboratories.
The resulting theory is expected to suffer from ultraviolet divergences [157],
and the one-loop [94] and two-loop [74] calculations for pure gravity are out-
standing pieces of work. As is well described in Ref. [157], if the coupling
constant of a field theory has dimension massd in h̄ = c = 1 units, then the
integral for a Feynman diagram of order N behaves at large momenta like
R A−N d
p dp, where A depends on the physical process considered but not on
the order N. Thus, the “harmful” interactions are those having negative val-
ues of d, which is precisely the case for Newton’s constant G, where d = −2,
since G = 6.67 × 10−39 GeV−2 in h̄ = c = 1 units. More precisely, since the
scalar curvature contains second derivatives of the metric, the corresponding
momentum-space vertex functions behave like p2 , and the propagator like
p−2 . In d dimensions each loop integral contributes pd , so that with L loops,
V vertices and P internal lines, the superficial degree D of divergence of a
2
Feynman diagram is given by [53]
D = dL + 2V − 2P. (1)
Moreover, a topological relation holds:
L = 1 − V + P, (2)
which leads to [53]

D = (d − 2)L + 2. (3)
In other words, D increases with increasing loop order for d > 2, so that it
clearly leads to a non-renormalizable theory.
A quantum theory of gravity is expected, for example, to shed new light on
singularities in classical cosmology. More precisely, the singularity theorems
prove that the Einstein theory of general relativity leads to the occurrence
of spacetime singularities in a generic way [86]. At first sight one might be
tempted to conclude that a breakdown of all physical laws occurred in the
past, or that general relativity is severely incomplete, being unable to predict
what came out of a singularity. It has been therefore pointed out that all these
pathological features result from the attempt of using the Einstein theory well
beyond its limit of validity, i.e. at energy scales where the fundamental theory
is definitely more involved. General relativity might be therefore viewed as a
low-energy limit of a richer theory, which achieves the synthesis of both the
basic principles of modern physics and the fundamental interactions in
the form currently known.
So far, no less than 16 major approaches to quantum gravity have been
proposed in the literature. Some of them make a direct or indirect use of the
action functional to develop a Lagrangian or Hamiltonian framework. They
are as follows.
1. Canonical quantum gravity [16, 17, 43, 44, 32, 99, 100, 6, 54, 144].
2. Manifestly covariant quantization [116, 33, 94, 74, 7, 152, 21, 103].
3. Euclidean quantum gravity [68, 90].
4. R-squared gravity [142].
5. Supergravity [64, 148].
6. String and brane theory [162, 98, 10].
3
7. Renormalization group and Weinberg’s asymptotic safety [129, 106].
8. Non-commutative geometry [26, 75].
Among these 8 approaches, string theory is peculiar because it is not field-
theoretic, spacetime points being replaced by extended structures such as
strings.
A second set of approaches relies instead upon different mathematical
structures with a more substantial (but not complete) departure from con-
ventional pictures, i.e.
9. Twistor theory [122, 123].
10. Asymptotic quantization [67, 5].
11. Lattice formulation [114, 22].
12. Loop space representation [133, 134, 136, 145, 154].
13. Quantum topology [101], motivated by Wheeler’s quantum geometrody-
namics [159].
14. Simplicial quantum gravity [72, 1, 109, 2] and null-strut calculus [102].
15. Condensed-matter view: the universe in a helium droplet [155].
16. Affine quantum gravity [105].
After such a concise list of a broad range of ideas, we hereafter focus
on the presentation of some very basic properties which underlie whatever
treatment of classical and quantum gravity, and are therefore of interest for
the general reader rather than (just) the specialist. He or she should revert
to the above list only after having gone through the material in sections 2–7.
2 Classical and quantum foundations

Before any attempt to quantize gravity we should spell out how classical
gravity can be described in modern language. This is done in the subsection
below.
4
2.1 Lorentzian spacetime and gravity
In modern physics, thanks to the work of Einstein [51], space and time are
unified into the spacetime manifold (M, g), where the metric g is a real-valued
symmetric bilinear map
g : Tp (M) × Tp (M) → R
of Lorentzian signature. The latter feature gives rise to the light-cone struc-
ture of spacetime, with vectors being divided into timelike, null or spacelike
depending on whether g(X, X) is negative, vanishing or positive, respectively.
The classical laws of nature are written in tensor language, and gravity is the
curvature of spacetime. In the theory of general relativity, gravity couples to
the energy-momentum tensor of matter through the Einstein equations
1 8πG
Rµν − gµν R = 4 Tµν . (4)
2 c
The Einstein–Hilbert action functional for gravity, giving rise to Eq. (4),
is diffeomorphism-invariant, and hence general relativity belongs actually to
the general set of theories ruled by an infinite-dimensional [31] invariance
group (or pseudo-group). With hindsight, following DeWitt [39], one can
say that general relativity was actually the first example of a non-Abelian
gauge theory (about 38 years before Yang–Mills theory [164]).
Note that the spacetime manifold is actually an equivalence class of pairs
(M, g), where two metrics are viewed as equivalent if one can be obtained
from the other through the action of the diffeomorphism group Diff(M). The
metric is an additional geometric structure that does not necessarily solve
any field equation.
2.2 From Schrödinger to Feynman

Quantum mechanics deals instead, mainly, with a probabilistic description
of the world on atomic or sub-atomic scale. It tells us that, on such scales,
the world can be described by a Hilbert space structure, or suitable gener-
alizations. Even in the relatively simple case of the hydrogen atom, the ap-
propriate Hilbert space is infinite-dimensional, but finite-dimensional Hilbert
spaces play a role as well. For example, the space of spin-states of a spin-s
particle is C2s+1 and is therefore finite-dimensional. Various pictures or for-
mulations of quantum mechanics have been developed over the years, and
their key elements can be summarized as follows:
5
(i) In the Schrödinger picture, one deals with wave functions evolving in
time according to a first-order equation. More precisely, in an abstract
Hilbert space H, one studies the Schrödinger equation
dψ
ih̄ = Ĥψ, (5)
dt
where the state vector ψ belongs to H, while Ĥ is the Hamiltonian

operator. In wave mechanics, the emphasis is more immediately put
on partial differential equations, with the wave function viewed as a
complex-valued map ψ : (x, t) → C obeying the equation
!
∂ψ h̄2
ih̄ = − △ +V ψ, (6)
∂t 2m
where −△ is the Laplacian in Cartesian coordinates on R3 (with this

sign convention, its symbol is positive-definite).
(ii) In the Heisenberg picture, what evolves in time are instead the opera-
tors, according to the first-order equation
dÂ
ih̄ = [Â, Ĥ]. (7)
dt
Heisenberg performed a quantum mechanical re-interpretation of kine-
matic and mechanical relations [93] because he wanted to formulate
quantum theory in terms of observables only.
(iii) In the Dirac quantization, from an assessment of the Heisenberg ap-
proach and of Poisson brackets [41], one discovers that quantum me-
chanics can be made to rely upon the basic commutation relations
involving position and momentum operators:
[q̂ j , q̂ k ] = [p̂j , p̂k ] = 0, (8)
[q̂ j , p̂k ] = ih̄δ jk . (9)

For generic operators depending on q̂, p̂ variables, their formal Taylor
series, jointly with application of (8) and (9), should yield their com-
mutator.
6
(iv) Weyl quantization. The operators satisfying the canonical commuta-
tion relations (9) cannot be both bounded [57], whereas it would be
nice to have quantization rules not involving unbounded operators and
domain problems. For this purpose, one can consider the strongly con-
tinuous one-parameter unitary groups having position and momentum
as their infinitesimal generators. These read as V (t) ≡ eitq̂ , U(s) ≡ eisp̂ ,
and satisfy the Weyl form of canonical commutation relations, which
is given by
U(s)V (t) = eisth̄ V (t)U(s). (10)
Here the emphasis was, for the first time, on group-theoretical meth-
ods, with a substantial departure from the historical development, that
relied instead heavily on quantum commutators and their relation with
classical Poisson brackets.
(v) Feynman quantization (i.e., Lagrangian approach). The Weyl approach
is very elegant and far-sighted, with several modern applications [57],
but still has to do with a more rigorous way of doing canonical quanti-
zation, which is not suitable for an inclusion of relativity. A spacetime
approach to ordinary quantum mechanics was instead devised by Feyn-
man [62] (and partly Dirac himself [42]), who proposed to express the
Green kernel of the Schrödinger equation in the form
Z
G[xf , tf ; xi , ti ] = eiS/h̄ dµ, (11)
all paths
where dµ is a suitable (putative) measure on the set of all spacetime

paths (including continuous, piecewise continuous, or even discontinu-
ous paths) matching the initial and final conditions. This point of view
has enormous potentialities in the quantization of field theories, since
it preserves manifest covariance and the full symmetry group, being
derived from a Lagrangian.
It should be stressed that quantum mechanics regards wave functions

only as a technical tool to study bound states (corresponding to the discrete
spectrum of the Hamiltonian operator Ĥ), scattering states (corresponding
instead to the continuous spectrum of Ĥ), and to evaluate probabilities (of
finding the values taken by the observables of the theory). Moreover, it is
meaningless to talk about an elementary phenomenon on atomic (or sub-
atomic) scale unless it is registered [160], and quantum mechanics in the
7
laboratory needs also an external observer and assumes the so-called reduc-
tion of the wave packet (see [57] and references therein). There exist indeed
different interpretations of quantum mechanics, e.g. Copenhagen [160], hid-
den variables [15], many worlds [60, 35].
2.3 Spacetime singularities

Now we revert to the geometric side. In Riemannian or pseudo-Riemannian
geometry, geodesics are curves whose tangent vector X moves by parallel
transport [85], so that eventually
dX λ
+ Γλµν X µ X ν = 0, (12)
ds
where s is the affine parameter and Γλµν are the connection coefficients. In
general relativity, timelike geodesics correspond to the trajectories of freely
moving observers, while null geodesics describe the trajectories of zero-rest-
mass particles (section 8.1 of Ref. [85]). Moreover, a spacetime (M, g) is said
to be singularity-free if all timelike and null geodesics can be extended to ar-
bitrary values of their affine parameter. At a spacetime singularity in general
relativity, all laws of classical physics would break down, because one would
witness very pathological events such as the sudden disappearance of freely
moving observers, and one would be completely unable to predict what came
out of the singularity. In the sixties, Penrose [121] proved first an important
theorem on the occurrence of singularities in gravitational collapse (e.g. for-
mation of black holes). Subsequent work by Hawking [79, 80, 81, 82, 83],
Geroch [66], Ellis and Hawking [84, 52], Hawking and Penrose [86] proved
that spacetime singularities are generic properties of general relativity, pro-
vided that physically realistic energy conditions hold. Very little analytic use
of the Einstein equations is made, whereas the key role emerges of topological
and global methods in general relativity.
On the side of singularity theory in classical cosmology, explicit mention
should be made of the work in Ref. [14], since it has led to significant progress
by Damour et al. [27], despite having failed to prove singularity avoidance
in classical cosmology. As pointed out in Ref. [27], the work by Belinsky et
al. is remarkable because it gives a description of the generic asymptotic be-
haviour of the gravitational field in four-dimensional spacetime in the vicinity
of a spacelike singularity. Interestingly, near the singularity the spatial points
essentially decouple, i.e. the evolution of the spatial metric at each spatial
8
point is asymptotically governed by a set of second-order, non-linear ordinary
differential equations in the time variable [14]. Moreover, the use of quali-
tative Hamiltonian methods leads naturally to a billiard description of the
asymptotic evolution, where the logarithms of spatial scale factors define a
geodesic motion in a region of the Lobachevskii plane, interrupted by geomet-
ric reflections against the walls bounding this region. Chaos follows because
the Bianchi IX billiard has finite volume [27]. A self-contained derivation
of the billiard picture for inhomogeneous solutions in D dimensions, with
dilaton and p-form gauge fields, has been obtained in Ref. [27].
2.4 Unification of all fundamental interactions

The fully established unifications of modern physics are as follows.
(i) Maxwell: electricity and magnetism are unified into electromagnetism.
All related phenomena can be described by an antisymmetric rank-two
tensor field, and derived from a one-form, called the potential.
(ii) Einstein: space and time are unified into the spacetime manifold. More-
over, inertial and gravitational mass, conceptually different, are actu-
ally unified as well.
(iii) Standard model of particle physics: electromagnetic, weak and strong
forces are unified by a non-Abelian gauge theory, normally considered
in Minkowski spacetime (this being the base space in fibre-bundle lan-
guage).
The physics community is now familiar with a picture relying upon four
fundamental interactions: electromagnetic, weak, strong and gravitational.
The large-scale structure of the universe, however, is ruled by gravity only.
All unifications beyond Maxwell involve non-Abelian gauge groups (either
Yang–Mills or Diffeomorphism group). At least three extreme views have
been developed along the years, i.e.,
(i) Gravity arose first, temporally, in the very early Universe, then all other
fundamental interactions.
(ii) Gravity might result from Quantum Field Theory (this was the Sakharov
idea [139]).
(iii) The vacuum of particle physics is regarded as a cold quantum liquid
in equilibrium. Protons, gravitons and gluons are viewed as collective
excitations of this liquid [155].
9
3 Canonical quantum gravity
Although Hamiltonian methods differ substantially from the Lagrangian ap-
proach used in the construction of the functional integral (see the follow-
ing sections), they remain nevertheless of great importance both in cos-
mology and in light of modern developments in canonical quantum gravity
[6, 136, 144], which is here presented within the original framework of quan-
tum geometrodynamics. For this purpose, it may be useful to describe the
main ideas of the Arnowitt-Deser-Misner (hereafter referred to as ADM) for-
malism. This is a canonical formalism for general relativity that enables one
to re-write Einstein’s field equations in first-order form and explicitly solved
with respect to a time variable. For this purpose, one assumes that four-
dimensional spacetime (M, g) can be foliated by a family of t = constant
spacelike surfaces St , giving rise to a 3 + 1 decomposition of the original
4-geometry. The basic geometric data of this decomposition are as follows
[53].
(1) The induced 3-metric h of the three-dimensional spacelike surfaces St .
This yields the intrinsic geometry of the three-space. h is also called the first
fundamental form of St , and is positive-definite with our conventions.
(2) The way each St is imbedded in (M, g). This is known once we are
able to compute the spatial part of the covariant derivative of the normal n
to St . On denoting by ∇ the four-connection of (M, g), one is thus led to
define the tensor
Kij ≡ −∇j ni . (13)
Note that Kij is symmetric if and only if ∇ is symmetric. In general relativity,
an equivalent definition of Kij is Kij ≡ − 21 (Ln h)ij , where Ln denotes the Lie
derivative along the normal to St . The tensor K is called extrinsic-curvature
tensor, or second fundamental form of St .
(3) How the coordinates are propagated off the surface St . For this pur-
pose one defines the vector (N, N 1 , N 2 , N 3 )dt connecting the point (t, xi )
with the point (t + dt, xi ). Thus, given the surface x0 = t and the surface
x0 = t + dt, Ndt ≡ dτ specifies a displacement normal to the surface x0 = t.
Moreover, N i dt yields the displacement from the point (t, xi ) to the foot of
the normal to x0 = t through (t + dt, xi ). In other words, the N i arise since
the xi = constant lines do not coincide in general with the normals to the
t = constant surfaces. According to a well-established terminology, N is the
lapse function, and the N i are the shift functions. They are the tool needed
10
to achieve the desired space-time foliation.
In light of points (1)–(3) as above, the 4-metric g can be locally cast in
the form
g = hij dxi + N i dt ⊗ dxj + N j dt − N 2 dt ⊗ dt. (14)
This implies that
g00 = − N 2 − Ni N i , (15)
gi0 = g0i = Ni , (16)
gij = hij , (17)
whereas, using the property g λν gνµ = δ λµ , one finds
1
g 00 = − , (18)
N2
Ni
g i0 = g 0i =, (19)
N2
N iN j
g ij = hij − . (20)
N2
Interestingly, the covariant gij and hij coincide, whereas the contravariant
g ij and hij differ as shown in (20). In terms of N, N i and h, the extrinsic-
curvature tensor defined in (13) takes the form

1 ∂hij
Kij ≡ − + Ni|j + Nj|i , (21)
2N ∂t
where the stroke | denotes covariant differentiation on the spacelike 3-surface
St , and indices of Kij are raised using hil . Equation (21) can be also written
as
∂hij
= Ni|j + Nj|i − 2NKij . (22)
∂t
Equation (22) should be supplemented by another first-order equation ex-
pressing the time evolution of Kij (recall that π ij is related to K ij ), i.e.
∂Kij h i
(3)
= −N|ij + N Rij + Kij (trK) − 2Kim Kjm
∂t h i
+ N m Kij|m + N|im Kjm + N|jm Kim . (23)
11
On using the ADM variables described so far, the form of the action
integral I for pure gravity that is stationary under variations of the metric
vanishing on the boundary is (in c = 1 units)
Z Z √
1 (4) √ 4 1
I ≡ R −g d x + Kii h d3 x
16πG M 8πG ∂M
1 Z (3) 2 √
= R + Kij K − Kii N h d3 x dt.
ij
(24)
16πG M
The boundary term appearing in (24) is necessary since (4) R contains second
derivatives of the metric, and integration by parts in the Einstein–Hilbert
part Z
1 (4) √
IH ≡ R −g d4 x
16πG M
1 R i
√ 3
of the action also leads to a boundary term equal to − 8πG ∂M Ki h d x. On
(4) 1 (4)
denoting by Gµν the Einstein tensor Gµν ≡ Rµν − 2 gµν R, and defining
1 h i
δΓρµν ≡ g ρλ ∇µ δgλν + ∇ν δgλµ − ∇λ δgµν , (25)
2
one then finds [165]
Z Z
√ µν 4 √
(16πG)δIH = − −g G δgµν d x+ −g g µν δρσ −g µσ δρν δΓρµν d3 x ,
M ∂M σ
(26)
which clearly shows that IH is stationary if the Einstein equations hold, and
the normal derivatives of the variations of the metric vanish on the boundary
∂M. In other words, IH is not stationary under arbitrary variations of the
metric, and stationarity is only achieved after adding to IH the boundary
term appearing in (24), if δgµν is set to zero on ∂M. Other useful forms
of the boundary term can be found in [68, 165]. Note also that, strictly, in
writing down (24) one should also take into account a term arising from IH
[32]:
1 Z Z h√ i
It ≡ dt d3 x ∂i h Kll N i − hij N|j . (27)
8πG ∂M
However, we have not explicitly included It since it does not modify the
results derived or described hereafter.
We are now ready to apply Dirac’s technique to the Hamiltonian quanti-
zation of general relativity. This requires that all classical constraints which
are first-class are turned into operators that annihilate the wave functional
12
[53]. Hereafter, we assume that this step has already been performed. As
we know, consistency of the quantum constraints is proved if one can show
that their commutators lead to no new constraints [53]. For this purpose, it
may be useful to recall the equal-time commutation relations of the canonical
variables, i.e. h i
N(x), π(x′ ) = iδ(x, x′ ), (28)
h i ′
Nj (x), π k (x′ ) = iδjk , (29)
h ′ ′
i ′ ′
hjk , π l m = iδjkl m . (30)
Note that, following [32], primes have been used, either on indices or on the
variables themselves, to distinguish different points of three-space. In other
words, one defines ′
δij ≡ δij δ(x, x′ ), (31)
′ ′
δij k l ≡ δij kl δ(x, x′ ), (32)
1 k l
δij kl ≡
δi δj + δil δjk . (33)
2
The reader can check that, since [32]
√
H ≡ h Kij K ij − K 2 − (3) R , (34)

Hi ≡ −2π,jij − hil 2hjl,k − hjk,l π jk , (35)
one has
h i h i h i h i
π(x), π i (x′ ) = π(x), Hi (x′ ) = π(x), H(x′ ) = π i (x), Hj (x′ )
h i
= π i (x), H(x′ ) = 0. (36)
h i h i h i
It now remains to compute the three commutators Hi , Hj ′ , Hi , H′ , H, H′ .
The first two commutators are obtained by using Eq. (35) and defining
Hi ≡ hij Hj . Interestingly, Hi is homogeneous bilinear in the hij and π ij ,
with the momenta always to the right. As we said before, following Dirac,
the operator version of constraints should annihilate the wave function since
the classical constraints are first-class (i.e. their Poisson brackets are linear
combinations of the constraints themselves). This condition reads as
Z
Hξ d3 x ψ = 0 ∀ξ, (37)
St
13
Z
Hi ξ i d3 x ψ = 0 ∀ξ i . (38)
St
In the applications to cosmology, Eq. (37) is known as the Wheeler–DeWitt
equation, and the functional ψ is then called the wave function of the universe
[78].
We begin by computing [32]
Z
′
hjk , i Hk′ δξ k d3 x′ = −hjk,l δξ l − hlk δξ l ,j − hjl δξ l ,k , (39)
St
Z
k′
jk
π ,i H δξ d x = − π jk δξ l
k′
3 ′
+ π lk δξ j ,l + π jl δξ k,l . (40)
St ,l
This calculation shows that the Hi are generators of three-dimensional co-

ordinate transformations xi = xi + δξ i . Thus, by using the definition of
structure constants of the general coordinate-transformation group [32], i.e.
′′ ′′ ′′ ′′ ′′
ck ij ′ ≡ δ k i,l′′ δjl ′ − δ k j ′,l′′ δil , (41)
the results (39)–(40) may be used to show that
h i Z
′ ′′
Hj (x), Hk (x ) = −i Hl′′ cl jk ′ d3 x′′ , (42)
St
h i
Hj (x), H(x′ ) = iH δ,j (x, x′ ). (43)
Note that the only term of H which might lead to difficulties is the one
quadratic in the momenta. However, all factors appearing in this term
have homogeneous linear transformation laws under the three-dimensional
coordinate-transformation group. They thus remain undisturbed in position
when commuted with Hj [32]. h i
Last, we have to study the commutator H(x), H(x′ ) . The following
remarks are in order:
(i) Terms quadratic in momenta contain no derivatives of hij or π ij with
respect to three-space
q coordinates.
Hence
q they commute;

(ii) The terms h(x) R(x) and h(x′ ) (3) R(x′ ) contain no momenta,
(3)
so that they also commute;

(iii) The only commutators we are left with are the cross-commutators,
and they can be evaluated by using the variational formula [32]
√ √ ij kl
δ h (3) R = h h h δhik,jl − δhij,kl
√ (3) ij 1 ij (3)
− h R − h R δhij , (44)
2
14
which leads to
Z Z Z
3 3
H ξ1 d x, H ξ2 d x = i Hl ξ1 ξ2,l − ξ1,l ξ2 d3 x. (45)
St St St
The commutators (42)–(43) and (45) clearly show that the constraint equa-
tions of canonical quantum gravity are first-class. The Wheeler-DeWitt equa-
tion (37) is an equation on the superspace (here Σ is a Riemannian 3-manifold
diffeomorphic to St )
S(Σ) ≡ Riem(Σ)/Dif f (Σ).
In this quotient space, two Riemannian metrics on Σ are identified if they

are related through the action of the diffeomorphism group Diff(Σ).
Two very useful classical formulae frequently used in Lorentzian canonical
gravity are √
ij kl h (3)
H ≡ (16πG)Gijklp p − R , (46)
16πG
h √ i
H ≡ (16πG)−1 Gijml Kij Kml − h (3) R , (47)
where the rank-4 tensor density is the DeWitt supermetric on superspace,
with covariant and contravariant forms
1
Gijkl ≡ √ hik hjl + hil hjk − hij hkl , (48)
2 h
√
h ik jl
Gijkl ≡ h h + hil hjk − 2hij hkl , (49)
2
√
h
and pij is here defined as − 16πG K ij − hij K . Note that the factor −2
multiplying hij hkl in (49) is needed so as to obtain the identity
1 k l
Gijmn Gmnkl = δi δj + δil δjk . (50)
2
Equation (46) clearly shows that H contains a part quadratic in the momenta
and a part proportional to (3) R (cf. (34)). On quantization, it is then hard
2
to give a well-defined meaning to the second functional derivative δhijδδhkl ,
whereas the occurrence of (3) R makes it even more difficult to solve exactly
the Wheeler-DeWitt equation.
15
It should be stressed that wave functions built from the functional integral
(see the following sections) which generalizes the path integral of ordinary
quantum mechanics (see (11)) do not solve the Wheeler–DeWitt equation
(37) unless some suitable assumptions are made [77], and counterexamples
have been built, i.e. a functional integral for the wave function of the universe
which does not solve the Wheeler–DeWitt equation [37].
4 From old to new unification

Here we outline how the space-of-histories formulation provides a common
ground for describing the ‘old’ and ‘new’ unifications of fundamental theories.
4.1 Old unification

Quantum field theory begins once an action functional S is given, since the
first and most fundamental assumption of quantum theory is that every iso-
lated dynamical system can be described by a characteristic action functional
[31]. The Feynman approach makes it necessary to consider an infinite-
dimensional manifold such as the space Φ of all field histories ϕi . On this
space there exist (in the case of gauge theories) vector fields
δ
Qα = Qiα (51)
δϕi
that leave the action invariant, i.e. [39]
Qα S = 0. (52)
The Lie brackets of these vector fields lead to a classification of all gauge
theories known so far.
4.2 Type-I gauge theories

The peculiar property of type-I gauge theories is that these Lie brackets are
equal to linear combinations of the vector fields themselves, with structure
constants, i.e. [38]
[Qα , Qβ ] = C γαβ Qγ , (53)
16
δC γ
where δϕαβ i = 0. The Maxwell, Yang–Mills, Einstein theories are all ex-
ample of type-I theories (this is the ‘unifying feature’). All of them, being
gauge theories, need supplementary conditions, since the second functional
derivative of S is not an invertible operator. After imposing such conditions,
the theories are ruled by an invertible differential operator of D’Alembert
type (or Laplace type, if one deals instead with Euclidean field theory), or a
non-minimal operator at the very worst (for arbitrary choices of gauge pa-
rameters). For example, when Maxwell theory is quantized via functional
integrals in the Lorenz [110] gauge, one deals with a gauge-fixing functional
Φ(A) = ∇µ Aµ , (54)
and the second-order differential operator acting on the potential in the

gauge-fixed action functional reads as

1
Pµν = −δµν + Rµν + 1− ∇µ ∇ν , (55)
α
where α is an arbitrary gauge parameter. The Feynman choice α = 1 leads
to the minimal operator
Peµν = −δµν + Rµν ,
which is the standard wave operator on vectors in curved spacetime. Such

operators play a leading role in the one-loop expansion of the Euclidean
effective action, i.e. the quadratic order in h̄ in the asymptotic expansion of
the functional ruling the quantum theory with positive-definite metrics.
The closure property expressed by Eq. (53) implies that the gauge group
decomposes the space of histories Φ into sub-spaces to which the Qα are
tangent. These sub-spaces are known as orbits, and Φ may be viewed as a
principal fibre bundle of which the orbits are the fibres. The space of orbits
is, strictly, the quotient space Φ/G, where G is the proper gauge group, i.e.
the set of transformations of Φ into itself obtained by exponentiating the
infinitesimal gauge transformation
δϕi = Qiα δξ α , (56)
and taking products of the resulting exponential maps. Suppose one performs
the transformation [39]
ϕi → I A , K α (57)
17
from the field variables ϕi to a set of fibre-adapted coordinates I A and K α .
With this notation, the I’s label the fibres, i.e. the points in Φ/G, and are
gauge invariant because
Qα I A = 0. (58)
The K’s label the points within each fibre, and one often makes specific
choices for the K’s, corresponding to the choice of supplementary condition
[31], more frequently called gauge condition. One normally picks out a base
point ϕ∗ in Φ and chooses the K ′ s to be local functionals of the ϕ’s in such
a way that the formula
Fbβα ≡ Qβ K α = K,iα Qiβ (59)
defines a non-singular differential operator, called the ghost operator, at and
in a neighbourhood of ϕ∗ . Thus, what is often called choosing a gauge
amounts to choosing a hypersurface K α = constant in a fibre-adapted coor-
dinate patch. The fields acted upon by the ghost operator are called ghost
fields, and have opposite statistics with respect to the fields occurring in the
gauge-invariant action functional (see Refs. [63, 61, 33] for the first time that
ghost fields were considered in quantized gauge theories). The gauge-fixed
action in the functional integral reads as [39]
1 ′
Sg.f. = S + K α ωαβ ′ K β , (60)
2
where ωαβ ′ is a non-singular matrix of gauge parameters (strictly, it is written
with matrix notation, but it contains Dirac’s delta, i.e. ωαβ ′ ≡ ωαβ δ(x, x′ )).
4.3 Type-II gauge theories

For type-II gauge theories, Lie brackets of vector fields Qα are as in Eq. (53)
for type-I theories, but the structure constants are promoted to structure
functions. An example is given by simple supergravity (a supersymmetric
[73, 158] gauge theory of gravity, with a symmetry relating bosonic and
fermionic fields) in four spacetime dimensions, with auxiliary fields [148].
4.4 Type-III gauge theories

In this case, the Lie bracket (53) is generalized by
[Qα , Qβ ] = C γαβ Qγ + U iαβ S,i , (61)
18
and it therefore reduces to (53) only on the mass-shell, i.e. for those field
configurations satisfying the Euler–Lagrange equations. An example is given
by theories with gravitons and gravitinos such as Bose–Fermi supermulti-
plets of both simple and extended supergravity in any number of spacetime
dimensions, without auxiliary fields [148].
4.5 From general relativity to supergravity

It should be stressed that general relativity is naturally related to supersym-
metry, since the requirement of gauge-invariant Rarita–Schwinger equations
[128] in curved spacetime implies Ricci-flatness in four dimensions [29], which
is then equivalent to vacuum Einstein equations. Of course, despite such a
relation does exist, general relativity can be (and is) formulated without any
use of supersymmetry.
The Dirac operator [56] is more fundamental in this framework, since
the m-dimensional spacetime metric is entirely re-constructed from the γ-
matrices, in that
g µν = 2−[m/2]−1 tr(γ µ γ ν + γ ν γ µ ). (62)
In four-dimensional spacetime, one can use the tetrad formalism, with Latin
indices (a, b) corresponding to tensors in flat space (the tangent frames, the
freely falling lifts) while Greek indices (µ, ν) correspond to coordinates in
curved space. The contravariant form of the spacetime metric g is then given
by
g µν = η ab eµa eνb , (63)
where η ab is the Minkowski metric and eµa are the tetrad vectors. The curved-
space γ-matrices γ µ are then obtained from the flat-space γ-matrices γ a and
from the tetrad according to
γ µ = γ a eµa . (64)
In Ref. [64], the authors assumed that the action functional describing the
interaction of tetrad fields and Rarita–Schwinger fields in curved spacetime,
subject to the Majorana constraint ψρ (x) = Cψ ρ (x)T , reads as
Z
1 −2 √ 1
I= d4 xκ −gR − ǫλρµν ψ λ (x)γ5 γµ Dν ψρ (x) , (65)
4 2
where the covariant derivative of Rarita–Schwinger fields is defined by
1
Dν ψρ (x) ≡ ∂ν ψρ (x) − Γσνρ ψσ + ωνab σ ab ψρ . (66)
2
19
With a standard notation, Γσνρ are the Christoffel symbols built from the
curved spacetime metric g µν , ωνab is the spin-connection (the gauge field
associated to the generators of the Lorentz algebra)
1h µ i
ωνab = ea (∂ν ebµ − ∂µ ebν ) + eρa eσb (∂σ ecρ )ecν − (a ↔ b), (67)
2
while σab is proportional to the commutator of γ-matrices in Minkowski
spacetime, i.e.
1
σab ≡ [γa , γb ]. (68)
4
To investigate the possible supersymmetry possessed by the above action
functional, the authors of Ref. [64] considered the transformation laws
δψµ (x) = κ−1 Dµ ǫ(x), (69)
δeaµ (x) = iκǫ(x)γ a ψµ (x), (70)

h i
δgµν (x) = iκǫ(x) γµ ψν (x) + γν ψµ (x) , (71)
where the supersymmetry parameter is taken to be an arbitrary Majorana
spinor field ǫ(x) of dimension square root of length. The assumption of local
supersymmetry was non-trivial, and was made necessary by the coordinate-
invariant Lagrangian (i.e. at that stage one had to avoid, for consistency, the
coordinate-dependent notion of constant, space-time-independent spinor).
After a lengthy calculation the authors of Ref. [64] managed to prove full
gauge invariance of the supergravity action. With geometrical hindsight, one
can prove it in a quicker and more elegant way by looking at a formulation
of Supergravity as a Yang–Mills Theory [149].
4.6 New unification

In modern high energy physics, the emphasis is no longer on fields (sections of
vector bundles in classical field theory [156], operator-valued distributions in
quantum field theory [161]), but rather on extended objects such as strings
[28]. In string theory, particles are not described as points, but instead
as strings, i.e., one-dimensional extended objects. While a point particle
sweeps out a one-dimensional worldline, the string sweeps out a worldsheet,
i.e., a two-dimensional real surface. For a free string, the topology of the
worldsheet is a cylinder in the case of a closed string, or a sheet for an
20
open string. It is assumed that different elementary particles correspond to
different vibration modes of the string, in much the same way as different
minimal notes correspond to different vibrational modes of musical string
instruments [28]. The five different string theories [4] are different aspects of
a more fundamental unified theory, called M-theory [13].
In the latest developments, one deals with ‘branes’, which are classical
solutions of the equations of motion of the low-energy string effective action,
that correspond to new non-perturbative states of string theory, break half of
the supersymmetry, and are required by duality arguments in theories with
open strings. They have the peculiar property that open strings have their
end-points attached to them [45, 46]. Branes have made it possible not only
to arrive at the formulation of M theory, but also to study perturbative and
non-perturbative properties of the gauge theories living on the world-volume
[47]. The so-called Dirichlet branes [124], or Dp branes, admit indeed two
distinct descriptions. On the one hand, they are classical solutions of the
low-energy string effective action (as we said before) and may be therefore
described in terms of closed strings. On the other hand, their dynamics is
determined by the degrees of freedom of the open strings with endpoints at-
tached to their world-volume, satisfying Dirichlet boundary conditions along
the directions transverse to the branes. They may be thus described in terms
of open strings as well. Such a twofold description of Dp branes laid the foun-
dations of the Maldacena conjecture [112] providing the equivalence between
a closed string theory, as the IIB theory on five-dimensional anti-de Sitter
space times the 5-sphere, and N = 4 super Yang–Mills with degrees of free-
dom corresponding to the massless excitations of the open strings having
their endpoints attached to a D3 brane.
For the impact of braneworld picture on phenomenology and unification,
we refer the reader to the seminal work in Refs. [126, 127], while for the
role of extra dimensions in cosmology we should mention also the work in
Refs. [137, 138]. With the language of pseudo-Riemannian geometry, branes
are timelike surfaces embedded into bulk spacetime [11, 10]. According to
this picture, gravity lives on the bulk, while standard-model gauge fields are
confined on the brane [12]. For branes, the normal vector N is spacelike with
respect to the bulk metric GAB , i.e.,
GAB N A N B = NC N C > 0. (72)
For a wide class of brane models, the action functional S pertaining to the
combined effect of bulk and brane geometry can be taken to split into the
21
sum [10] (gαβ (x) being the brane metric)
S = S4 [gαβ (x)] + S5 [GAB (X)], (73)
while the effective action [38] Γ is formally given by

Z
iΓ
e = DGAB (X) eiS × gauge − fixing term. (74)
In the functional integral, the gauge-fixed action reads as (here there is sum-
mation as well as integration over repeated indices [31, 38, 10])
1 1
Sg.f. = S4 + S5 + F A ΩAB F B + χµ ωµν χν , (75)
2 2
where F A and χµ are bulk and brane gauge-fixing functionals, respectively,
while ΩAB and ωµν are non-singular matrices of gauge parameters, similarly
to the end of section 4.2. The gauge-invariance properties of bulk and brane
action functionals can be expressed by saying that there exist vector fields
on the space of histories such that (cf. Eq. (52))
RB S5 = 0, Rν S4 = 0, (76)
whose Lie brackets obey a relation formally analogous to Eq. (53) for ordi-
nary type-I theories, i.e.
[RB , RD ] = C ABD RA , (77)
[Rµ , Rν ] = C λµν Rλ . (78)

Equations (77) and (78) refer to the sharply different Lie algebras of dif-
feomorphisms on the bulk and the brane, respectively. The bulk and brane
ghost operators are therefore
QAB ≡ RB F A = F A,a RaB , (79)
J µν ≡ Rν χµ = χµ,i Riν , (80)

respectively, where the commas denote functional differentiation with respect
to the field variables. The full bulk integration means integrating first with
respect to all bulk metrics GAB inducing on the boundary ∂M the given
brane metric gαβ (x), and then integrating with respect to all brane metrics.
22
Thus, one first evaluates the cosmological wave function [10] of the bulk
spacetime (which generalizes the wave function of the universe encountered
in canonical quantum gravity), i.e.
Z
e
ψBulk = µ(GAB , SC , T D )eiS5 , (81)
GAB [∂M ]=gαβ
where µ is taken to be a suitable measure, the SC , T D are ghost fields, and

1
Se5 ≡ S5 [GAB ] + F A ΩAB F B + SA QAB T B . (82)
2
Eventually, the effective action results from
Z
iΓ e
e = e αβ , ργ , σ δ )eiS4 ψBulk ,
µ(g (83)
where µe is another putative measure, ργ and σ δ are brane ghost fields, and
1
Se4 ≡ S4 + χµ ωµν χν + ρµ J µν σ ν . (84)
2
We would like to stress here that infinite-dimensional manifolds are the
natural arena for studying the quantization of the gravitational field, even
prior to considering a space-of-histories formulation. There are, indeed, at
least three sources of infinite-dimensionality in quantum gravity:
(i) The infinite-dimensional Lie group (or pseudo-group) of spacetime dif-

feomorphisms, which is the invariance group of general relativity in the
first place [31], [143].
(ii) The infinite-dimensional space of histories in a functional-integral quan-
tization [38, 39].
(iii) The infinite-dimensional Geroch space of asymptotically simple space-
times [67].
5 Functional integrals and background fields

We now study in greater detail some aspects of the use of functional integrals
in quantum gravity, after the previous (formal) applications to a space of
histories formulation.
23
5.1 The one-loop approximation
In the one-loop approximation (also called stationary phase or JWKB meth-
od) one first expands both the metric g and the fields φ coupled to it about
a metric g0 and a field φ0 which are solutions of the classical field equations:
g = g0 + g, (85)
φ = φ0 + φ. (86)
One then assumes that the fluctuations g and φ are so small that the dom-
inant contribution to the functional integral for the in-out amplitude comes
from the quadratic order in the Taylor-series expansion of the action about
the background fields g0 and φ0 [90]:
IE [g, φ] = IE [g0 , φ0 ] + I2 [g, φ] + higher − order terms, (87)
so that the logarithm of the quantum-gravity amplitude Ae can be expressed

as Z
e
log A ∼ −IE [g0 , φ0] + log D[g, φ]e−I2 [g,φ] . (88)
It should be stressed that background fields need not be a solution of any field
equation [36], but this possibility will not be exploited in our presentation.
For our purposes we are interested in the second term appearing on the right-
hand side of (88). An useful factorization is obtained if φ0 can be set to zero.
One then finds that I2 [g, φ] = I2 [g] + I2 [φ], which implies [91]
Z Z
log Ae ∼ −IE [g0 ] + log D[φ]e−I2 [φ] + log D[g]e−I2 [g] . (89)
The one-loop term for matter fields with various spins (and boundary con-
ditions) is extensively studied in the literature. We here recall some basic
results, following again Ref. [91].
A familiar form of I2 [φ] is
Z
1 √
I2 [φ] = φBφ g0 d4 x, (90)
2
where the elliptic differential operator B depends on the background metric
g0 . Note that B is a second-order operator for bosonic fields, whereas it
is first-order for fermionic fields.
n o In light of (90) it is clear that we are
interested in the eigenvalues λn of B, with corresponding eigenfunctions
24
n o
φn . If boundaries are absent, it is sometimes possible to know explicitly
the eigenvalues with their degeneracies. This is what happens for example
in de Sitter space. If boundaries are present, however, very little is known
about the detailed form of the eigenvalues, once boundary conditions have
been imposed.
We here assume for simplicity to deal with bosonic fields subject to (ho-
mogeneous) Dirichlet conditions on the boundary surface: φ = 0 on ∂M , and
φn = 0 on ∂M , ∀n. It is in fact well-known that the Laplace operator subject
to Dirichlet conditions has a positive-definite spectrum [23]. The field φ can
then be expanded in terms of the eigenfunctions φn of B as
∞
X
φ= y n φn , (91)
n=n0
where the eigenfunctions φn are normalized so that

Z
√
φn φm g0 d4 x = δnm . (92)
Another formula we need is the one expressing the measure on the space of
all fields φ as
∞
Y
D[φ] = µ dyn , (93)
n=n0
where the normalization parameter µ has dimensions of mass or (length)−1 .

Note that, if gauge fields appear in the calculation, the choice of gauge-
fixing and the form of the measure in the functional integral are not a trivial
problem.
On using well-known results about Gaussian integrals, the one-loop am-
(1)
plitudes Aeφ can be now obtained as
Z
(1)
Aeφ ≡ D[φ]e−I2 [φ]
∞ Z
Y λn 2
= µ dyn e− y
2 n
n=n0
∞
Y 1
= 2πµ2 λ−1
n
2
n=n0
1
= r . (94)
1 −1 −2
det 2 π µ B
25
When fermionic fields appear in the functional integral for the in-out
amplitude, one deals with a first-order elliptic operator, the Dirac opera-
e These are anticommuting
tor, acting on independent spinor fields ψ and ψ.
Grassmann variables obeying the Berezin integration rules
Z Z
dw = 0, w dw = 1. (95)
The formulae (95) are all what we need, since powers of w greater than or
equal to 2 vanish in light of the anticommuting property. The reader can
then check that the one-loop amplitude for fermionic fields is

(1) 1 −2
Aeψ = det µ B . (96)
2
The main difference with respect to bosonic fields is the direct proportionality
to the determinant. The following comments can be useful in understanding
the meaning of (96).
Let us denote again by γ µ the curved-space γ-matrices, and by λi the
eigenvalues of the Dirac operator γ µ Dµ , and suppose that no zero-modes ex-
ist. More precisely, the eigenvalues of γ µ Dµ occur in equal and opposite pairs:
±λ1 , ±λ2 , ..., whereas the eigenvalues of the Laplace operator on spinors oc-
cur as (λ1 )2 twice, (λ2 )2 twice, and so on. For Dirac fermions (D) one thus
finds ! !
∞
Y ∞
Y ∞
Y
detD γ µ Dµ = | λi | | λi | = | λi |2 , (97)
i=1 i=1 i=1
whereas in the case of Majorana spinors (M), for which the number of degrees
of freedom is halved, one finds
∞ r
Y
µ
detM γ Dµ = | λi |= detD γ µ Dµ . (98)
i=1
5.2 Zeta-function regularization of functional integrals

The formal expression (94) for the one-loop quantum amplitude clearly di-
verges since the eigenvalues λn increase without bound, and a regularization
is thus necessary. For this purpose, the following technique has been de-
scribed and applied by many authors [48, 89, 91].
Bearing in mind that Riemann’s zeta-function ζR (s) is defined as
∞
X
ζR (s) ≡ n−s , (99)
n=1
26
one first defines a generalized (also called spectral) zeta-function ζ(s) ob-
tained from the (positive) eigenvalues of the second-order, self-adjoint oper-
ator B. Such a ζ(s) can be defined as (cf. [141])
∞
X ∞
X
ζ(s) ≡ dm (n)λ−s
n,m . (100)
n=n0 m=m0
This means that all the eigenvalues are completely characterized by two in-
teger labels n and m, while their degeneracy dm only depends on n. Note
that formal differentiation of (100) at the origin yields
′
det B = e−ζ (0) . (101)
This result can be given a sensible meaning since, in four dimensions, ζ(s)
converges for Re(s) > 2, and one can perform its analytic extension to a
meromorphic

function

of s which only has poles at s = 12 , 1, 32 , 2. Since
det µB = µζ(0) det B , one finds the useful formula
1 1
log Aeφ = ζ ′(0) + log 2πµ2 ζ(0). (102)
2 2
As we said following (90), it may happen quite often that the eigenvalues
appearing in (100) are unknown, since the eigenvalue condition, i.e. the
equation leading to the eigenvalues by virtue of the boundary conditions, is
a complicated equation which cannot be solved exactly for the eigenvalues.
However, since the scaling properties of the one-loop amplitude are still given
by ζ(0) (and ζ ′ (0)) as shown in (102), efforts have been made to compute
ζ(0) also in this case. The various steps of this program are as follows [89].
(1) One first studies the heat equation for the operator B, i.e.
∂
F (x, y, τ ) + BF (x, y, τ ) = 0, (103)
∂τ
where the Green’s function F satisfies the initial condition F (x, y, 0) =
δ(x, y). n o
(2) Assuming completeness of the set φn of eigenfunctions of B, the
field φ can be expanded as
∞
X
φ= an φn .
n=ni
27
(3) The Green’s function F (x, y, τ ) is then given by
∞
X ∞
X
F (x, y, τ ) = e−λn,m τ φn,m(x) ⊗ φn,m (y). (104)
n=n0 m=m0
(4) The corresponding integrated heat kernel is then

Z ∞
X ∞
X
√ 4
G(τ ) = Tr F (x, x, τ ) g d x = e−λn,m τ . (105)
M n=n0 m=m0
(5) In light of (100) and (105), the generalized zeta-function can be also
obtained as an integral transform (also called inverse Mellin transform) of
the integrated heat kernel, i.e. [89, 71]
1 Z ∞ s−1
ζ(s) = τ G(τ ) dτ. (106)
Γ(s) 0
(6) The hard part of the analysis is now to prove that G(τ ) has an
asymptotic expansion as τ → 0+ [76]. This property has been proved for
all boundary conditions such that the Laplace operator is self-adjoint and
the boundary-value problem is strongly elliptic [71, 8]. The corresponding
asymptotic expansion of G(τ ) can be written as
3 1
√
G(τ ) ∼ A0 τ −2 + A 1 τ − 2 + A1 τ −1 + A 3 τ − 2 + A2 + O τ , (107)
2 2
which implies
ζ(0) = A2 . (108)
The result (108) is proved by splitting the integral in (106) into an integral
from 0 to 1 and an integral from 1 to ∞. The asymptotic expansion of
R 1 s−1
0 τ G(τ ) dτ is then obtained by using (107).
In other words, for a given second-order self-adjoint elliptic operator,
we study the corresponding heat equation, and the integrated heat kernel
G(τ ). The ζ(0) value is then given by the constant term appearing in the
asymptotic expansion of G(τ ) as τ → 0+ . The ζ(0) value also yields the
one-loop divergences of the theory for bosonic and fermionic fields [55].
5.3 Gravitational instantons

This section is devoted to the study of the background gravitational fields.
These gravitational instantons are complete four-geometries solving the Ein-
stein equations R(X, Y )−Λg(X, Y ) = 0 when the four-metric g has signature
28
+4 (i.e. it is positive-definite, and thus called Riemannian). They are of in-
terest because they occur in the tree-level approximation of the partition
function, and in light of their role in studying tunnelling phenomena. More-
over, they can be interpreted as the stationary phase metrics in the path
integrals for the partition functions, Z, of the thermal canonical ensemble
and the volume canonical ensemble. In these cases the action of the instan-
ton gives the dominant contribution to − log Z. Following [125], essentially
three cases can be studied.
5.3.1 Asymptotically locally Euclidean instantons

Even though it might seem natural to define first the asymptotically Eu-
clidean instantons, it turns out that there is not much choice in this case,
since the only asymptotically Euclidean instanton is flat space. It is in fact
well-known that the action of an asymptotically Euclidean metric with van-
ishing scalar curvature is ≥ 0, and it vanishes if and only if the metric is
flat. Suppose now that such a metric is a solution of the Einstein equations
R(X, Y ) = 0. Its action should be thus stationary also under constant con-
formal rescalings g → k 2 g of the metric. However, the whole action rescales
then as IE → k 2 IE , so that it can only be stationary and finite if IE = 0. By
virtue of the theorem previously mentioned, the metric g must then be flat
[70, 108].
In the asymptotically locally Euclidean case, however, the boundary at
infinity has topology S 3 /Γ rather than S 3 , where Γ is a discrete subgroup
of the group SO(4). Many examples can then be found. The simplest
was discovered by Eguchi and Hanson [49, 50], and corresponds to Γ = Z2
and ∂M = RP 3 . nThis o instanton is conveniently described using three left-
invariant 1-forms ωi on the 3-sphere, satisfying the SU(2) algebra dωi =
− 12 ǫi jk ωj ∧ ωk , and parametrized by Euler angles as follows:
ω1 = (cos ψ)dθ + (sin ψ)(sin θ)dφ, (109)
ω2 = −(sin ψ)dθ + (cos ψ)(sin θ)dφ, (110)
ω3 = dψ + (cos θ)dφ, (111)
where θ ∈ [0, π], φ ∈ [0, 2π]. The metric of the Eguchi–Hanson instanton
may be thus written in the Bianchi-IX form [125]
!−1 " ! #
a4 r2 a4
g1 = 1 − 4 dr ⊗ dr + (ω1 )2 + (ω2 )2 + 1 − 4 (ω3 )2 , (112)
r 4 r
29
where r ∈ [a, ∞[. The singularity of g1 at r = a is only a coordinate singu-
2 4
larity. We may get rid of it by defining 4 aρ2 ≡ 1 − ar4 , so that, as r → a, the
metric g1 is approximated by the metric
h i2 a2 h i
g2 = dρ ⊗ dρ + ρ2 dψ + (cos θ)dφ + dθ ⊗ dθ + (sin θ)2 dφ ⊗ dφ . (113)
4
Regularity of g2 at ρ = 0 is then guaranteed provided that one identifies ψ
with period 2π. This implies in turn that the local surfaces r = constant have
topology RP 3 rather than S 3 , as we claimed. Note that at r = a ⇒ ρ = 0
the metric becomes that of a 2-sphere of radius a2 . Following Ref. [69], we
∂
say that r = a is a bolt, where the action of the Killing vector ∂ψ has a
two-dimensional fixed-point set [125].
A whole family of multi-instanton solutions is obtained by taking the
group Γ = Zk . They all have a self-dual Riemann-curvature tensor, and
their metric takes the form
2
g = V −1 dτ + γ · dx + V dx · dx. (114)
Following [125], V = V (x) and γ = γ(x) on an auxiliary flat 3-space with

metric dx · dx. This metric g solves the Einstein vacuum equations provided
that grad V = curlγ, which implies △V = 0. If one takes
n
X 1
V = , (115)
i=1 | x − xi |
one obtains the desired asymptotically locally Euclidean multi-instantons. In

particular, if n = 1 in (115), g describes flat space, whereas n = 2 leads to the
Eguchi–Hanson instanton. If n > 2, there are (3n − 6) arbitrary parameters,
related to the freedom to choose the positions xi of the singularities in V .
These singularities correspond actually to coordinate singularities in (113),
and can be removed by using suitable coordinate transformations [125].
5.3.2 Asymptotically flat instantons

This name is chosen since the underlying idea is to deal with metrics in the
functional integral which tend to the flat metric in three directions but are
periodic in the Euclidean-time dimension. The basic example is provided
30
(1)
by the Riemannian version gR (also called Euclidean) of the Schwarzschild
solution, i.e.
−1
(1) M M
gR = 1−2 dτ ⊗ dτ + 1 − 2 dr ⊗ dr + r 2 Ω2 , (116)
r r
where Ω2 = dθ ⊗ dθ + (sin θ)2 dφ ⊗ dφ is the metric on a unit 2-sphere. It
is indeed well-known that, in the Lorentzian case, the metric gL is more
conveniently written by using Kruskal–Szekeres coordinates
r

gL = 32M 3 r −1 e− 2M − dz ⊗ dz + dy ⊗ dy + r 2 Ω2 , (117)
where z and y obey the relations

r r
− z2 + y2 = − 1 e 2M , (118)
2M
(y + z) t
= e 2M . (119)
(y − z)
In the Lorentzian case, the coordinate singularity at r = 2M can be thus
avoided, whereas the curvature singularity at r = 0 remains and is described
by the surface z 2 − y 2 = 1. However, if we set ζ = iz, the analytic continu-
ation to the section of the complexified space-time where ζ is real yields the
positive-definite (i.e. Riemannian) metric
r

(2)
gR = 32M 3 r −1 e− 2M dζ ⊗ dζ + dy ⊗ dy + r 2 Ω2 , (120)
where r

r
ζ 2 + y2 = − 1 e 2M . (121)
2M
It is now clear that also the curvature singularity at r = 0 has disappeared,
since the left-hand side of (121) is ≥ 0, whereas the right-hand side of (121)
would be equal to −1 at r = 0. Note also that, by setting z = −iζ and
t = −iτ in (119), and writing ζ 2 + y 2 as (y + iζ)(y − iζ) in (121), one finds
r
iτ r r
y + iζ = e 4M − 1 e 4M , (122)
2M
r
τ r r
y = cos − 1 e 4M , (123)
4M 2M
31
which imply that the Euclidean time τ is periodic with period 8πM. This
periodicity on the Euclidean section leads to the interpretation of the Rie-
mannian Schwarzschild solution as describing a black hole in thermal equi-
librium with gravitons at a temperature (8πM)−1 [125]. Moreover, the fact
that any matter-field Green’s function on this Schwarzschild background is
also periodic in imaginary time leads to some of the thermal-emission prop-
erties of black holes. This is one of the greatest conceptual revolutions in
modern gravitational physics.
Interestingly, a new asymptotically flat gravitational instanton has been
found by Chen and Teo in Ref. [24]. It has an U(1) × U(1) isometry group
and some novel global features with respect to the other two asymptotically
flat instantons, i.e. Euclidean Schwarzschild and Euclidean Kerr.
There is also a local version of the asymptotically flat boundary condition
in which ∂M has the topology of a non-trivial S 1 -bundle over S 2 , i.e. S 3 /Γ,
where Γ is a discrete subgroup of SO(4). However, unlike the asymptot-
ically Euclidean boundary condition, the S 3 is distorted and expands with
increasing radius in only two directions rather than three [125]. The simplest
example of an asymptotically locally flat instanton is the self-dual (i.e. with
self-dual curvature 2-form) Taub-NUT solution, which can be regarded as a
special case of the two-parameter Taub-NUT metrics
(r + M) (r − M) h i
g= dr ⊗ dr + 4M 2 (ω3 )2 + r 2 − M 2 (ω1 )2 + (ω2 )2 , (124)
(r − M) (r + M)
n o
where the ωi have been defined in (109)–(111). The main properties of
the metric (124) are
(I) r ∈ [M, ∞[, and r = M is a removable coordinate singularity provided
that ψ is identified modulo 4π;
(II) the r = constant surfaces have S 3 topology;
(III) r = M is a point at which the isometry generated by the Killing
∂
vector ∂ψ has a zero-dimensional fixed-point set.
In other words, r = M is a nut, using the terminology in Ref. [69].

There is also a family of asymptotically locally flat multi-Taub-NUT in-
stantons. Their metric takes the form (114), but one should bear in mind
that the formula (115) is replaced by
n
X 2M
V =1+ . (125)
i=1 | x − xi |
32
Again, the singularities at x = xi can be removed, and the instantons are all
self-dual.
5.3.3 Compact instantons

Compact gravitational instantons occur in the course of studying the topo-
logical structure of the gravitational vacuum. This can be done by first of all
normalizing all metrics in the functional integral to have a given 4-volume
V , and then evaluating the instanton contributions to the partition function
as a function of their topological complexity. One then sends the volume
V to infinity at the end of the calculation. If one wants to constrain the
metrics in the functional integral to have a volume V , this can be obtained
Λ
by adding a term 8π V to the action. The stationary points of the modified
action are solutions of the Einstein equations with cosmological constant Λ,
i.e. R(X, Y ) − Λg(X, Y ) = 0.
The few compact instantons that are known can be described as follows [125].
(1) The 4-sphere S 4 , i.e. the Riemannian version of de Sitter space ob-
tained by analytic continuation to positive-definite metrics. Setting to 3 for
convenience the cosmological constant, the metric on S 4 takes the form [125]
1 h i
gI = dβ ⊗ dβ + (sin β)2 (ω1 )2 + (ω2 )2 + (ω3 )2 , (126)
4
where β ∈ [0, π]. The apparent singularities at β = 0, π can be made into
regular nuts, provided that the Euler angle ψ is identified modulo 4π. The
β = constant surfaces are topologically S 3 , and the isometry group of the
metric (126) is SO(5).
(2) If in C 3 we identify (z1 , z2 , z3 ) and (λz1 , λz2 , λz3 ), ∀λ ∈ C − {0},
we obtain, by definition, the complex projective space CP 2 . For this two-
dimensional complex space one can find a real four-dimensional metric, which
solves the Einstein equations with cosmological constant Λ. If we set Λ to 6
for convenience, the metric of CP 2 takes the form [125]
1 h i
gII = dβ ⊗ dβ + (sin β)2 (ω1 )2 + (ω2 )2 + (cos β)2 (ω3 )2 , (127)
4
h i
where β ∈ 0, π2 . A bolt exists at β = π2 , where ∂ψ ∂
has a two-dimensional
fixed-point set. The isometry group of gII is locally SU(3), which has a U(2)
subgroup acting on the three-spheres β = constant.
33
(3) The Einstein metric on the product manifold S 2 × S 2 is obtained as
the direct sum of the metrics on two 2-spheres, i.e.
2
1X
g= dθi ⊗ dθi + (sin θi )2 dφi ⊗ dφi . (128)
Λ i=1
The metric (128) is invariant under the SO(3) × SO(3) isometry group of
S 2 × S 2 , but is not of Bianchi-IX type as (126)-(127). This can be achieved
by a coordinate transformation leading to [125]
gIII = dβ ⊗ dβ + (cos β)2 (ω1 )2 + (sin β)2 (ω2 )2 + (ω3 )2 , (129)

h i
where Λ = 2 and β ∈ 0, π2 . Regularity at β = 0, π2 is obtained provided
that ψ is identified modulo 2π (cf. (126)). Remarkably, this is a regular
Bianchi-IX Einstein metric in which the coefficients of ω1 , ω2 and ω3 are all
different.
(4) The nontrivial S 2 -bundle over S 2 has a metric which, by setting Λ = 3,
may be cast in the form [118, 125]
h i
gIV = (1 + ν 2 ) f1 (x)dx ⊗ dx + f2 (x) (ω1 )2 + (ω2 )2 + f3 (x)(ω3 )2 , (130)
where x ∈ [0, 1], ν is the positive root of
w 4 + 4w 3 − 6w 2 + 12w − 3 = 0, (131)
and the functions f1 , f2 , f3 are defined by
(1 − ν 2 x2 )
f1 (x) ≡ , (132)
(3 − ν 2 − ν 2 (1 + ν 2 )x2 )(1 − x2 )
(1 − ν 2 x2 )
f2 (x) ≡ , (133)
(3 + 6ν 2 − ν 4 )
(3 − ν 2 − ν 2 (1 + ν 2 )x2 )(1 − x2 )
f3 (x) ≡ . (134)
(3 − ν 2 )(1 − ν 2 x2 )
The isometry group corresponding to (130) may be shown to be U(2).
(5) Another compact instanton of fundamental importance is the K3
surface, whose explicit metric has not yet been found. K3 is defined as the
compact complex surface whose first Betti number and first Chern class are
34
vanishing. A physical picture of the K3 gravitational instanton has been
obtained by Page [119].
Two topological invariants exist which may be used to characterize the var-
ious gravitational instantons studied so far. These invariants are the Euler
number χ and the Hirzebruch signature τ . The Euler number can be defined
as an alternating sum of Betti numbers, i.e.
χ ≡ B0 − B1 + B2 − B3 + B4 . (135)
The Hirzebruch signature can be defined as
τ ≡ B2+ − B2− , (136)
where B2+ is the number of self-dual harmonic 2-forms, and B2− is the number
of anti-self-dual harmonic 2-forms [in terms of the Hodge-star operator ∗ Fab ≡
1
ǫ F cd , self-duality of a 2-form F is expressed as ∗ F = F , and anti-self-
2 abcd
duality as ∗ F = −F ]. In the case of compact four-dimensional manifolds
without boundary, χ and τ can be expressed as integrals of the curvature
[91] Z
1 √
χ= 2
Rλµνρ Rαβγδ ǫλµαβ ǫνργδ g d4 x, (137)
128π M
Z
1 √
τ= 2
Rλµνρ Rλµαβ ǫνραβ g d4 x. (138)
96π M
For the instantons previously listed one finds [125]
Eguchi–Hanson: χ = 2, τ = 1.
Asymptotically locally Euclidean multi-instantons: χ = n, τ = n − 1.
Schwarzschild: χ = 2, τ = 0.
Taub-NUT: χ = 1, τ = 0.
Asymptotically locally flat multi-Taub-NUT instantons: χ = n, τ = n − 1.
S 4 : χ = 2, τ = 0.
CP 2 : χ = 3, τ = 1.
S 2 × S 2 : χ = 4, τ = 0.
S 2 -bundle over S 2 : χ = 4, τ = 0.
K3: χ = 24, τ = 16.
35
6 Spectral zeta-functions in one-loop quan-
tum cosmology
In the late nineties a systematic investigation of boundary conditions in
quantum field theory and quantum gravity has been performed (see Refs.
[111, 150, 55, 117, 8] and the many references therein). It is now clear that
the set of fully gauge-invariant boundary conditions in quantum field theory,
providing a unified scheme for Maxwell, Yang–Mills and General Relativity
is as follows: h i
πA = 0, (139)
∂M
h i
Φ(A) = 0, (140)
∂M
[ϕ]∂M = 0, (141)
where π is a projection operator, A is the Maxwell potential, or the Yang–
Mills potential, or the metric (more precisely, their perturbation about a
background value which can be set to zero for Maxwell or Yang–Mills), Φ is
the gauge-fixing functional, ϕ denotes the set of ghost fields for these bosonic
theories. Equations (139) and (140) are both preserved under infinitesimal
gauge transformations provided that the ghost obeys homogeneous Dirichlet
conditions as in Eq. (141). For gravity, it may be convenient to choose Φ so
as to have an operator P of Laplace type in the Euclidean theory.
6.1 Eigenvalue condition for scalar modes

In a quantum theory of the early universe via functional integrals, the semi-
classical analysis remains a valuable tool, but the tree-level approximation
might be an oversimplification. Thus, it seems appropriate to consider at
least the one-loop approximation. On the portion of flat Euclidean 4-space
bounded by a 3-sphere, called Euclidean 4-ball and relevant for one-loop
quantum cosmology [78, 92, 55] when a portion of 4-sphere bounded by a
3-sphere is studied in the limit of small 3-geometry [140], the metric pertur-
bations hµν can be expanded in terms of scalar, transverse vector, transverse-
traceless tensor harmonics on the 3-sphere S 3 of radius a. For vector, tensor
and ghost modes, boundary conditions reduce to Dirichlet or Robin. For
scalar modes, one finds eventually the eigenvalues E = X 2 from the roots X
of [58, 59]
n
Jn′ (x) ± Jn (x) = 0, (142)
x
36

x n
Jn′ (x)
+ − ± Jn (x) = 0, (143)
2 x
where Jn are the Bessel functions of first kind. Note that both x and −x
solve the same equation.
6.2 Four spectral zeta-functions for scalar modes

By virtue of the Cauchy theorem and of suitable rotations of integration
contours in the complex plane [19], the eigenvalue conditions (142) and (143)
give rise to the following four spectral zeta-functions [58, 59]:
∞ Z ∂ ±
± sin(πs) X ∞ log FA,B (zn)
ζA,B (s) ≡ n−(2s−2) dz ∂z , (144)
π n=3 0 z 2s
where, denoting by In the modified Bessel functions of first kind (here β+ ≡

n, β− ≡ n + 2),

FA± (zn) ≡ z −β± znIn′ (zn) ± nIn (zn) , (145)
! !
z 2 n2
FB± (zn) ≡z −β±
znIn′ (zn) + ± n In (zn) . (146)
2
Regularity at the origin is easily proved in the elliptic sectors, corresponding
to ζA± (s) and ζB− (s) [58, 59].
6.3 Regularity at the origin of ζB+

1
With the notation in Refs. [58, 59], if one defines the variable τ ≡ (1 + z 2 )− 2 ,
one can write the uniform asymptotic expansion of FB+ in the form [58, 59]
 
∞
enη(τ ) (1 − τ 2 )  X rj,+ (τ ) 
FB+ ∼ √ 1+ . (147)
h(n) τ τ j=1 nj
R R R
On splitting the integral 01 dτ = 0µ dτ + µ1 dτ with µ small, one gets an
asymptotic expansion of the left-hand side of Eq. (144) by writing, in the
first interval on the right-hand side,
 
∞
X ∞
rj,+ (τ )  X Rj,+ (τ )
log 1 + ∼ , (148)
j=1 nj j=1 nj
37
and then computing [58, 59]
X4j
∂Rj,+ −j−1
Cj (τ ) ≡ = (1 − τ ) Ka(j) τ a . (149)
∂τ a=j−1
Remarkably, by virtue of the identity obeyed by the spectral coefficients Ka(j)

on the 4-ball, i.e.
4j
X Γ(a + 1)
g(j) ≡ Ka(j) = 0, (150)
a=j Γ(a − j + 1)
which holds ∀j = 1, ..., ∞, one finds [58, 59]

12
1X
lim sζB+ (s) = a(a − 1)(a − 2)Ka(3) = 0, (151)
s→0 6 a=3
and [58, 59]

12 ∞
5 1079 1 X X 296
ζB+ (0) = + − ω(a)Ka(3) + f (j)g(j) = , (152)
4 240 2 a=2 j=1 45
where, on denoting here by ψ the logarithmic derivative of the Γ-function

[58, 59],

1 Γ(a + 1) (6a2 − 9a + 1) Γ(a − 2)
ω(a) ≡ − log(2) −
6 Γ(a − 2) 4 Γ(a + 1)

+ 2ψ(a + 1) − ψ(a − 2) − ψ(4) , (153)
(−1)j h i
f (j) ≡ − 1 − 22−j + ζR (j − 2)(1 − δj,3 ) + γδj,3 . (154)
j!
Equation (150) achieves three goals:
(i) Vanishing of log(2) coefficient in (152);
P∞
(ii) Vanishing of j=1 f (j)g(j) in (152);
(iii) Regularity at the origin of ζB+ .
38
6.4 Interpretation of the result
Since all other ζ(0) values for pure gravity obtained in the literature are neg-
ative, the analysis here briefly outlined shows that only fully diffeomorphism-
invariant boundary conditions lead to a positive ζ(0) value for pure gravity on
the 4-ball, and hence only fully diffeomorphism-invariant boundary conditions
lead to a vanishing cosmological wave function for vanishing 3-geometries at
one-loop level, at least on the Euclidean 4-ball. If the probabilistic interpre-
tation is tenable for the whole universe, this means that the universe has
vanishing probability of reaching the initial singularity at a = 0, which is
therefore avoided by virtue of quantum effects [58, 59], since the one-loop
wave function is proportional to aζ(0) [140].
Interestingly, quantum cosmology can have observational consequences as
well. For example, the work in Ref. [104] has derived the primordial power
spectrum of density fluctuations in the framework of quantum cosmology,
by performing a Born–Oppenheimer approximation of the Wheeler–DeWitt
equation for an inflationary universe with a scalar field. In this way one
first recovers the scale-invariant power spectrum that is found as an ap-
proximation in the simplest inflationary models. One then obtains quantum
gravitational corrections to this spectrum, discussing whether they lead to
measurable signatures in the Cosmic Microwave Background anisotropy spec-
trum [104].
7 Hawking’s radiation
Hawking’s theoretical discovery of particle creation by black holes [87, 88] has
led, along the years, to many important developments in quantum field theory
in curved spacetime, quantum gravity and string theory. Thus, we devote
this section to a brief review of such an effect, relying upon the DAMTP
lecture notes by Townsend [147]. For this purpose, we consider a massless
scalar field Φ in a Schwarzschild black hole spacetime. The positive-frequency
outgoing modes of Φ are known to behave, near future null infinity F + , as
Φω ∼ e−iωu . (155)
According to a geometric optics approximation, a particle’s worldline is a

null ray γ of constant phase u, and we trace this ray backwards in time from
F + . The later it reaches F + , the closer it must approach the future event
39
horizon H+ in the exterior spacetime before entering the star. The ray γ
belongs to a family of rays whose limit as t → ∞ is a null geodesic generator,
denoted by γH , of H+ . One can specify γ by giving its affine distance from
γH along an ingoing null geodesic passing through H+ . The affine parameter
on this ingoing null geodesic is U, so U = −ǫ. One can thus write, on γ near
H+ (κ being the surface gravity),
1
u = − log ǫ, (156)
κ
so that positive-frequency outgoing modes have, near H+ , the approximate
form
iω
Φω ∼ e κ log ǫ . (157)
This describes increasingly rapid oscillations as ǫ → 0, and hence the geo-
metric optics approximation is indeed justified at late times.
The positive-frequency outgoing modes should be matched onto a solution
of the Klein–Gordon equation near past null infinity F − . When geometric
optics holds, one performs parallel transport of the vectors n parallel to F −
and l orthogonal to n back to F − along the continuation of γH . Such a
continuation can be taken to meet F − at v = 0. The continuation of the null
ray γ back to F − meets F − at an affine distance ǫ along an outgoing null
geodesic on F − . The affine parameter on outgoing null geodesics in F − is v,
because the line element takes on F − the form
ds2 = du dv + r 2 dΩ2 , (158)
dΩ2 being the line element on a unit 2-sphere, so that v = −ǫ and
iω
log(−v)
Φω ∼ e κ . (159)
This holds for negative values of v. When v is instead positive, an ingoing
null ray from F − passes through H+ and does not reach F + , hence the
positive-frequency outgoing modes depend on v on F − , where
iω
log(−v)
Φω (v) = 0 if v > 0, e κ if v < 0. (160)
Consider now the Fourier transform
Z ∞ ′
e
Φ eiω v Φω (v)dv
ω ≡
−∞
Z 0 ′ iω
= eiω v+ κ log(−v)
dv. (161)
−∞
40
In this integral, let us choose the branch cut in the complex v-plane to lie
along the real axis. For positive ω ′ let us rotate contour to the positive
imaginary axis and then set v = ix to get
Z
∞ −ω ′ x+ iω log xe−iπ/2
e (ω ′ ) = −i
Φ e κ
dx
ω
0
Z ∞
πω ′ iω
= −e 2κ e−ω x+ κ log(x)
dx. (162)
0
Since ω ′ is positive the integral converges. When ω ′ is negative one can rotate
the contour to the negative imaginary axis and then set v = −ix to get
Z
∞ ω ′ x+ iω log xeiπ/2
e
Φ ′
ω (ω ) = i e dx
κ
0
Z ∞
πω ′ iω
= e− 2κ eω x+ κ log(x)
dx. (163)
0
From the two previous formulae one gets

πω
e (−ω ′ ) = −e−
Φ κ e (ω ′ ) if ω ′ > 0.
Φ (164)
ω ω
Thus, a mode of positive frequency ω on F + matches, at late times, onto

positive and negative modes on F − . For positive ω ′ one can identify
e (ω ′ ),
Aωω′ = Φ (165)
ω
πω
e (−ω ′ ) = −e−
Bωω′ = Φ κ e (ω ′ ),
Φ (166)
ω ω
as the Bogoliubov coefficients. These formulae imply that

πω
Bij = −e− κ Aij . (167)
On the other hand, the matrices A and B should satisfy the Bogoliubov
relations, from which
δij = (AA† − BB † )ij

X
= (Ail A∗jl − Bil Bjl∗ )
l
h π(ωi +ωj ) iX
= e κ −1 Bil Bjl∗ , (168)
l
41
where we have inserted the formula relating Bij to Aij . Now one can take
i = j to get
1
(BB † )ii = 2πωi . (169)
e κ −1
Eventually, one needs the inverse Bogoliubov coefficients corresponding to
a positive-frequency mode on F − matching onto positive- and negative-
frequency modes on F + . Since the inverse B coefficient is found to be
B ′ = −B T , (170)
the late-time particle flux through F + , given a vacuum on F − , turns out to

be ∗
hNi iF + = (B ′ )† B ′ = B ∗ B T = BB T . (171)
ii ii ii
T
From reality of (BB )ii , the previous formulae lead to
1
hNi iF + = 2πωi . (172)
e κ −1
Remarkably, this is the Planck distribution for black body radiation from a
Schwarzschild black hole at the Hawking temperature
h̄κ
TH = . (173)
2π
8 Achievements and open problems

At this stage, the general reader might well be wondering what has been
gained by working on the quantum gravity problem over so many decades.
Indeed, at the theoretical level, at least the following achievements can be
brought to his (or her) attention:
(i) The ghost fields [63] necessary for the functional-integral quantization of
gravity and Yang–Mills theories [33, 61] have been discovered, jointly with a
deep perspective on the space of histories formulation.
(ii) The Vilkovisky–DeWitt gauge-invariant effective action [151, 38] has been
obtained and thoroughly studied.
(iii) We know that black holes emit a thermal spectrum and have temperature
and entropy by virtue of semiclassical quantum effects [87, 88, 40]. A full
42
theory of quantum gravity should account for this and should tell us whether
or not the black hole evaporation process comes to an end [153].
(iv) The manifestly covariant theory leads to the detailed calculation of phys-
ical quantities such as cross-sections for gravitational scattering of identical
scalar particles, scattering of gravitons by scalar particles, scattering of one
graviton by another and gravitational bremsstrahlung [34], but no laboratory
experiment is in sight for these effects.
(v) The ultimate laboratory for modern high energy physics is the whole uni-
verse. We have reasons to believe that either we need string and brane theory
with all their (extra) ingredients, or we have to resort to radically different
approaches such as, for example, loop space or twistors, the latter two living
however in isolation with respect to deep ideas such as supersymmetry and
supergravity (but we acknowledge that twistor string theory [163] is making
encouraging progress [113, 146]).
Although string theory may provide a finite theory of quantum grav-
ity that unifies all fundamental interactions at once, its impact on particle
physics phenomenology and laboratory experiments remains elusive. Some
key issues are therefore in sight:
(i) What is the impact (if any) of Planck-scale physics on cosmological

observations [166]?
(ii) Will general relativity retain its role of fundamental theory, or shall we
have to accept that it is only the low-energy limit of string or M-theory?
(iii) Are renormalization-group methods a viable way to do non-perturba-
tive quantum gravity [129, 18], after the recent discovery of a non-
Gaussian ultraviolet fixed point [106, 130, 107] of the renormalization-
group flow?
(iv) Is there truly a singularity avoidance in quantum cosmology [58, 59] or
string theory [97, 95, 96]?
8.1 Experimental side

As is well stressed, for example, in Ref. [75], gravity is so weak that it can
only produce measurable effects in the presence of big masses, and this makes
it virtually impossible to detect radiative corrections to it. Nevertheless, at
43
least four items can be brought to the attention of the reader within the
experimental framework.
(i) Colella et al. [25] have used a neutron interferometer to observe the
quantum-mechanical phase shift of neutrons caused by their interaction with
the Earth’s gravitational field.
(ii) Page and Geilker [120] have considered an experiment that gave results
inconsistent with the simplest alternative to quantum gravity, i.e. the semi-
classical Einstein equation. This evidence supports, but does not prove, the
hypothesis that a consistent theory of gravity coupled to quantized matter
should also have the gravitational field quantized [30].
(iii) Balbinot et al. have shown [9] that, in a black hole-like configuration
realized in a Bose–Einstein condensate, a particle creation of the Hawking
type does indeed take place and can be unambiguously identified via a char-
acteristic pattern in the density-density correlations. This has opened the
concrete possibility of the experimental verification of this effect.
(iv) Mercati et al. [115], for the study of the Planck-scale modifications [3]
of the energy-momentum dispersion relations, have considered the possible
role of experiments involving nonrelativistic particles and particularly atoms.
They have extended a recent result, establishing that measurements of atom-
recoil frequency can provide insight that is valuable for some theoretical
models.
We are already facing unprecedented challenges, where the achievements
of spacetime physics and quantum field theory are called into question. The
years to come will hopefully tell us whether the many new mathematical con-
cepts considered in theoretical physics lead really to a better understanding
of the physical universe and its underlying structures.
Acknowledgments The author is grateful to the Dipartimento di Sci-
enze Fisiche of Federico II University, Naples, for hospitality and support.
He dedicates to Maria Gabriella his work.
References
[1] Ambjorn, J., Carfora, M., Gabrielli, D. and Marzuoli, A. Crumpled
Triangulations and Critical Points in 4-D Simplicial Quantum Gravity,
Nuclear Physics B Vol. 542, pp. 349–394 (1999).
44
[2] Ambjorn, J., Jurkiewicz, J. and Loll, R. The Self-Organized de Sitter
Universe, International Journal of Modern Physics D Vol. 17, pp. 2515–
2520 (2009).
[3] Amelino–Camelia, G. Quantum Gravity Phenomenology, preprint

arXiv:0806.0339 [gr-qc].
[4] Angelantonj, C. and Sagnotti, A. Open Strings, Physics Reports Vol.

371, pp. 1–150 (2002).
[5] Ashtekar, A. Asymptotic Quantization of the Gravitational Field, Phys-

ical Review Letters Vol. 46, pp. 573–576 (1981).
[6] Ashtekar, A. New Variables for Classical and Quantum Gravity, Phys-
[7] Avramidi, I.G. A Covariant Technique for the Calculation of the One-
Loop Effective Action, Nuclear Physics B Vol. 355, pp. 712–754 (1991).
[8] Avramidi, I.G. and Esposito, G. Gauge Theories on Manifolds with

Boundary, Communications in Mathematical Physics Vol. 200, pp.
495–543 (1999).
[9] Balbinot, R., Carusotto, I., Fabbri, A. and Recati, A. Testing Hawking
Particle Creation by Black Holes Through Correlation Measurements,
International Journal of Modern Physics D Vol. 19, pp. 2371–2377
(2010).
[10] Barvinsky, A.O. Quantum Effective Action in Spacetimes with Branes

and Boundaries: Diffeomorphism Invariance, Physical Review D Vol.
74, 084033, 18 pages (2006).
[11] Barvinsky, A.O. and Nesterov, D.V. Quantum Effective Action in

Spacetimes with Branes and Boundaries, Physical Review D Vol. 73,
066012, 12 pages (2006).
[12] Barvinsky, A.O., Kamenshchik, A.Yu., Kiefer, C. and Nesterov, D.V.

Effective Action and Heat Kernel in a Toy Model of Brane-Induced
Gravity, Physical Review D Vol. 75, 044010, 14 pages (2007).
45
[13] Becker, K., Becker, M. and Schwarz, J.H. (2007). String Theory and
M-theory: a Modern Introduction, Cambridge University Press, Cam-
bridge.
[14] Belinsky, V.A., Khalatnikov, I.M. and Lifshitz, E.M. Oscillatory Ap-
proach to a Singular Point in the Relativistic Cosmology, Advances in
Physics Vol. 19, pp. 525–573 (1970).
[15] Bell, J.S. On the Problem of Hidden Variables in Quantum Mechanics,

Reviews of Modern Physics Vol. 38, pp. 447–452 (1966).
[16] Bergmann, P.G. Nonlinear Field Theories, Physical Review Vol. 75,
pp. 680–685 (1949).
[17] Bergmann, P.G. and Brunings, J.H.M. Nonlinear Field Theories II:
Canonical Equations and Quantization, Reviews of Modern Physics
Vol. 21, pp. 480–487 (1949).
[18] Bonanno, A. and Reuter, M. Cosmology of the Planck Era from a

Renormalization Group for Quantum Gravity, Physical Review D Vol.
65, 043508, 20 pages (2002).
[19] Bordag, M., Geyer, B., Kirsten, K. and Elizalde, E. Zeta-Function

Determinant of the Laplace Operator on the D-dimensional Ball, Com-
munications in Mathematical Physics Vol. 179, pp. 215–234 (1996).
[20] Bronstein, M.P. Quantentheories Schwacher Gravitationsfelder,

Physikalische Zeitschrift der Sowietunion Vol. 9, pp. 140–157 (1936).
[21] Canfora, F. A Large N Expansion for Gravity, Nuclear Physics B Vol.

731, pp. 389–405 (2005).
[22] Carfora, M. and Marzuoli, A. Bounded Geometries and Topologi-

cal Fluctuations in Lattice Quantum Gravity, Classical and Quantum
Gravity Vol. 9, pp. 595–627 (1992).
[23] Chavel, I. (1984) Eigenvalues in Riemannian Geometry, Academic

Press, New York.
[24] Chen, Y. and Teo, E. A New AF Gravitational Instanton,

arXiv:1107.0763 [gr-qc].
46
[25] Colella, R., Overhauser, A.W. and Werner, S.A. Observation of Gravi-
tationally Induced Quantum Interference, Physical Review Letters Vol.
34, pp. 1472–1474 (1975).
[26] Connes, A. and Marcolli, M. (2008). Non-Commutative Geometry,
Quantum Fields and Motives, American Mathematical Society, Provi-
dence.
[27] Damour, T., Henneaux, M. and Nicolai, H. Cosmological Billiards,
Classical and Quantum Gravity Vol. 20, pp. R145–R200 (2003).
[28] Deligne, P., Etingof, P., Freed, D.S., Jeffrey, L.C., Kazhdan, P., Mor-
gan, J.W., Morrison, D.R. and Witten, E. (1999). Quantum Fields and
Strings: a Course for Mathematicians, American Mathematical Soci-
ety, Providence.
[29] Deser, S. and Zumino, B. Consistent Supergravity, Physics Letters B
Vol. 62, pp. 335–337 (1976).
[30] DeWitt, B.S. (1962) The Quantization of Geometry, in Gravitation: An
Introduction to Current Research, pp. 266–381, ed. L. Witten, Wiley,
New York.
[31] DeWitt, B.S. (1965). Dynamical Theory of Groups and Fields, Gordon
& Breach, New York.
[32] DeWitt, B.S. Quantum Theory of Gravity. I. The Canonical Theory,
Physical Review Vol. 160, pp. 1113–1148 (1967).
[33] DeWitt, B.S. Quantum Theory of Gravity. II. The Manifestly Covari-
ant Theory, Physical Review Vol. 162, pp. 1195–1239 (1967).
[34] DeWitt, B.S. Quantum Theory of Gravity. III. Applications of the Co-
variant Theory, Physical Review Vol. 162, pp. 1239–1256 (1967).
[35] DeWitt, B.S. and Graham, R.N. Resource Letter IQM-1 on the Inter-
pretation of Quantum Mechanics, American Journal of Physics Vol. 39,
pp. 724–738 (1971).
[36] DeWitt, B.S. (1981) A Gauge Invariant Effective Action, in Quantum
Gravity, A Second Oxford Symposium, pp. 449–487, eds. C.J. Isham,
R. Penrose and D.W. Sciama, Clarendon Press, Oxford.
47
[37] DeWitt, B.S. (1999) The Quantum and Gravity: The Wheeler–DeWitt
Equation, in College Station 1998, Richard Arnowitt Fest: Relativity,
Particle Physics and Cosmology, pp. 70–92, ed. R.E. Allen, Singapore,
World Scientific.
[38] DeWitt, B.S. (2003). The Global Approach to Quantum Field Theory,
International Series of Monographs on Physics Vol. 114, Clarendon
Press, Oxford.
[39] DeWitt, B.S. (2005) The Space of Gauge Fields: Its Structure and
Geometry, in 50 Years of Yang–Mills Theory, pp. 15–32, ed. G. ’t Hooft,
World Scientific, Singapore.
[40] DeWitt, B.S. Quantum Gravity: Yesterday and Today, General Rela-
tivity and Gravitation Vol. 41, pp. 413–419 (2009).
[41] Dirac, P.A.M. The Fundamental Equations of Quantum Mechanics,

Proceedings Royal Society London A Vol. 109, pp. 642–653 (1925).
[42] Dirac, P.A.M. The Lagrangian in Quantum Mechanics, Physikalische

Zeitschrift USSR Vol. 3, pp. 64–72 (1933).
[43] Dirac, P.A.M. The Theory of Gravitation in Hamiltonian Form, Pro-

ceedings Royal Society of London A Vol. 246, pp. 333–343 (1958).
[44] Dirac, P.A.M. Fixation of Coordinates in the Hamiltonian Theory of

Gravitation, Physical Review Vol. 114, pp. 924–930 (1959).
[45] Di Vecchia, P. and Liccardo, A. D-Branes in String Theory. 1., NATO

Advanced Study Institute Series C Mathematical Physical Sciences Vol.
556, pp. 1–59 (2000).
[46] Di Vecchia, P. and Liccardo, A. D-Branes in String Theory. 2.,

hep-th/9912275.
[47] Di Vecchia, P. and Liccardo, A. (2007) Gauge Theories From D-Branes,

in Frontiers in Number Theory, Physics and Geometry II, Eds. P.
Cartier, B. Julia, P. Moussa and P. Vanhove, pp. 161–222, Springer–
Verlag, Berlin.
48
[48] Dowker, J.S. and Critchley, R. Effective Lagrangian and Energy-
Momentum Tensor in de Sitter Space, Physical Review D Vol. 13, pp.
3224–3232 (1976).
[49] Eguchi, T. and Hanson, A.J. Asymptotically Flat Self-Dual Solutions

to Euclidean Gravity, Physics Letters B Vol. 74, pp. 249–251 (1978).
[50] Eguchi, T. and Hanson, H.J. Self-Dual Solutions to Euclidean Gravity,

Annals of Physics (N.Y.), Vol. 120, pp. 82–106 (1979).
[51] Einstein, A. The Foundation of the General Theory of Relativity, An-

nalen der Physik Vol. 49, pp. 769–822 (1916).
[52] Ellis, G.F.R. and Hawking, S.W. The Cosmic Black Body Radiation
and the Existence of Singularities in our Universe, Astrophysical Jour-
nal Vol. 152, pp. 25–36 (1968).
[53] Esposito, G. (1994) Quantum Gravity, Quantum Cosmology and

Lorentzian Geometries, in Lecture Notes in Physics, New Series m:
Monographs, Vol. 12, Springer–Verlag, Berlin.
[54] Esposito, G., Gionti, G. and Stornaiolo, C. Spacetime Covariant Form

of Ashtekar’s Constraints, Nuovo Cimento B Vol. 110, pp. 1137–1152
(1995).
[55] Esposito, G., Kamenshchik, A.Yu. and Pollifrone, G. (1997) Euclidean

Quantum Gravity on Manifolds with Boundary, in Fundamental Theo-
ries of Physics, Vol. 85, Kluwer, Dordrecht.
[56] Esposito, G. (1998). Dirac Operators and Spectral Geometry, Cam-

bridge Lecture Notes in Physics Vol. 12, Cambridge University Press,
Cambridge.
[57] Esposito, G., Marmo, G. and Sudarshan, G. (2004). From Classical to

Quantum Mechanics, Cambridge University Press, Cambridge.
[58] Esposito, G., Fucci, G., Kamenshchik, A.Yu. and Kirsten, K. Spectral
Asymptotics of Euclidean Quantum Gravity with Diff-Invariant Bound-
ary Conditions, Classical and Quantum Gravity Vol. 22, pp. 957–974
(2005).
49
[59] Esposito, G., Fucci, G., Kamenshchik, A.Yu. and Kirsten, K. A
Non-Singular One-Loop Wave Function of the Universe from a New
Eigenvalue Asymptotics in Quantum Gravity, Journal of High Energy
Physics Vol. 0509, 063, 17 pages (2005).
[60] Everett, H. Relative State Formulation of Quantum Mechanics, Re-

views of Modern Physics Vol. 29, pp. 454–462 (1957).
[61] Faddeev, L.D. and Popov, V.N. Feynman Diagrams for the Yang–Mills
Field, Physics Letters B Vol. 25, pp. 29-30 (1967).
[62] Feynman, R.P. Space-Time Approach to Non-Relativistic Quantum Me-

chanics, Review of Modern Physics Vol. 20, pp. 367–387 (1948).
[63] Feynman R.P. Quantum Theory of Gravitation, Acta Physica Polonica

Vol. 24, pp. 697–722 (1963).
[64] Freedman, D.Z., van Nieuwenhuizen, P. and Ferrara, S. Progress To-

ward a Theory of Supergravity, Physical Review D Vol. 13, pp. 3214–
3218 (1976).
[65] Fulling, S.A. (1989) Aspects of Quantum Field Theory in Curved Space-
Time, Cambridge University Press, Cambridge.
[66] Geroch, R.P. Singularities in Closed Universes, Physical Review Let-

ters Vol. 17, pp. 445–447 (1966).
[67] Geroch, R.P. An Approach to Quantization of General Relativity, An-

nals of Physics (N.Y.) Vol. 62, pp. 582–589 (1971).
[68] Gibbons, G.W. and Hawking, S.W. Action Integrals and Partition
Functions in Quantum Gravity, Physical Review D Vol. 15, pp. 2752–
2756 (1977).
[69] Gibbons, G.W. and Hawking, S.W. Classification of Gravitational In-

stanton Symmetries, Communications in Mathematical Physics Vol.
66, pp. 291–310 (1979).
[70] Gibbons, G.W. and Pope, C.N. The Positive Action Conjecture and
Asymptotically Euclidean Metrics in Quantum Gravity, Communica-
tions in Mathematical Physics Vol. 66, pp. 267–290 (1979).
50
[71] Gilkey, P.B. (1995) Invariance Theory, The Heat Equation and the
Atiyah–Singer Index Theorem, CRC Press, Boca Raton.
[72] Gionti, G. Discrete Approaches Towards the Definition of Quantum

Gravity, SISSA Ph.D. Thesis, gr-qc/9812080 (1998).
[73] Golfand, A.Yu. and Likhtman, E.P. Extension of the Algebra of

Poincaré Group Generators and Violation of p Invariance, JETP Let-
ters Vol. 13, pp. 323–326 (1971).
[74] Goroff, M.H. and Sagnotti, A. The Ultraviolet Behavior of Einstein

Gravity, Nuclear Physics B Vol. 266, pp. 709–736 (1986).
[75] Gracia–Bondia, J. Notes on Quantum Gravity and Noncommutative

Geometry, in New Paths Towards Quantum Gravity, eds. B. Booss–
Bavnbek, G. Esposito and M. Lesch, Springer Lecture Notes in Physics
Vol. 807, pp. 3–58 (2010).
[76] Greiner, P. An Asymptotic Expansion for the Heat Equation, Archive

Rational Mechanics Analysis Vol. 41, pp. 163–218 (1971).
[77] Halliwell, J.J. and Hartle, J.B. Wave Functions Constructed From an
Invariant Sum Over Histories Satisfy Constraints, Physical Review D
Vol. 43, pp. 1170–1194 (1991).
[78] Hartle, J.B. and Hawking, S.W. Wave Function of the Universe, Phys-
ical Review D Vol. 28, pp. 2960–2975 (1983).
[79] Hawking, S.W. Occurrence of Singularities in Open Universes, Physical

Review Letters Vol. 15, pp. 689–690 (1965).
[80] Hawking, S.W. Singularities in the Universe, Physical Review Letters

Vol. 17, pp. 444-445 (1966).
[81] Hawking, S.W. The Occurrence of Singularities in Cosmology, Pro-

ceedings Royal Society London A Vol. 294, pp. 511–521 (1966).
[82] Hawking, S.W. The Occurrence of Singularities in Cosmology. II, Pro-

ceedings Royal Society London A Vol. 295, pp. 490–493 (1966).
51
[83] Hawking, S.W. The Occurrence of Singularities in Cosmology. III.
Causality and Singularities, Proceedings Royal Society London A Vol.
300, pp. 187–201 (1967).
[84] Hawking, S.W. and Ellis, G.F.R. Singularities in Homogeneous World
Models, Physics Letters Vol. 17, pp. 246–247 (1965).
[85] Hawking, S.W. and Ellis, G.F.R. (1973). The Large-Scale Structure of
Space-Time, Cambridge University Press, Cambridge.
[86] Hawking, S.W. and Penrose, R. The Singularities of Gravitational Col-
lapse and Cosmology, Proceedings Royal Society London A Vol. 314,
pp. 529–548 (1970).
[87] Hawking, S.W. Black Hole Explosions?, Nature Vol. 248, pp. 30–31
(1974).
[88] Hawking, S.W. Particle Creation by Black Holes, Communications in
Mathematical Physics Vol. 43, pp. 199–220 (1975).
[89] Hawking, S.W. Zeta Function Regularization of Path Integrals in
Curved Spacetime, Communications in Mathematical Physics Vol. 55,
pp. 133–148 (1977).
[90] Hawking, S.W. Quantum Gravity and Path Integrals, Physical Review
D Vol. 18, pp. 1747–1753 (1978).
[91] Hawking, S.W. (1979) The Path Integral Approach to Quantum Grav-
ity, in General Relativity, an Einstein Centenary Survey, pp. 746–789,
eds. S.W. Hawking and W. Israel, Cambridge University Press, Cam-
bridge.
[92] Hawking, S.W. The Quantum State of the Universe, Nuclear Physics
B Vol. 239, pp. 257–276 (1984).
[93] Heisenberg, W. Quantum-Mechanical Re-Interpretation of Kinematic
and Mechanical Relations, Zeitschrift für Physik Vol. 33, pp. 879–893
(1925).
[94] ’t Hooft, G. and Veltman, M. One-Loop Divergencies in the Theory
of Gravitation, Annales Institut Henri Poincaré A Vol. 20, pp. 69–94
(1974).
52
[95] Horowitz, G.T. and Marolf, D. Quantum Probes of Spacetime Singu-
larities, Physical Review D Vol. 52, pp. 5670–5675 (1995).
[96] Horowitz, G.T. and Myers, R. (1995) The Value of Singularities, Gen-
eral Relativity and Gravitation Vol. 27, pp. 915–919 (1995).
[97] Horowitz, G.T. and Steif, A.R. Space-Time Singularities in String The-
ory, Physical Review Letters Vol. 64, pp. 260–263 (1990).
[98] Horowitz, G.T. Spacetime in String Theory, New Journal of Physics

Vol. 7, 201, 13 pages (2005).
[99] Isham, C.J. and Kakas, A.C. A Group Theoretic Approach to the
Canonical Quantization of Gravity. 1. Construction of the Canonical
Group, Classical and Quantum Gravity Vol. 1, pp. 621–632 (1984).
[100] Isham, C.J. and Kakas, A.C. A Group Theoretical Approach to the
Canonical Quantization of Gravity. 2. Unitary Representations of the
Canonical Group, Classical and Quantum Gravity Vol. 1, pp. 633–650
(1984).
[101] Isham, C.J. Quantum Topology and Quantization on the Lattice of

Topologies, Classical and Quantum Gravity Vol. 6, pp. 1509–1534
(1989).
[102] Kheyfets, A., La Fave, N.J. and Miller, W.A. A Few Insights Into
the Nature of Classical and Quantum Gravity via Null Strut Calculus,
Classical and Quantum Gravity Vol. 6, pp. 659–682 (1989).
[103] Kiefer, C. (2007) Quantum Gravity, International Series of Monographs

on Physics Vol. 136, Clarendon Press, Oxford.
[104] Kiefer, C. and Krämer, M. Quantum Gravitational Contributions to

the CMB Anisotropy Spectrum, arXiv:1103.4967 (2011).
[105] Klauder, J.R. Fundamentals of Quantum Gravity, Journal of Physics

Conference Series Vol. 87, 012012, 9 pages (2007).
[106] Lauscher, O. and Reuter, M. Ultraviolet Fixed Point and Generalized

Flow Equation of Quantum Gravity, Physical Review D Vol. 65, 025013,
44 pages (2002).
53
[107] Lauscher, O. and Reuter, M. Asymptotic Safety in Quantum Ein-
stein Gravity: Nonperturbative Renormalizability and Fractal Space-
time Structure, hep-th/0511260 (2005).
[108] LeBrun, C.R. Counter-Examples to the Generalized Positive Action

Conjecture, Communications in Mathematical Physics Vol. 118, pp.
591–596 (1988).
[109] Loll, R. The Emergence of Spacetime or Quantum Gravity on Your

Desktop, Classical and Quantum Gravity Vol. 25, 114006, 17 pages
(2008).
[110] Lorenz, L. On the Identity of the Vibrations of Light with Electrical

Currents, Philosophical Magazine Vol. 34, pp. 287–301 (1867).
[111] Luckock, H.C. Mixed Boundary Conditions in Quantum Field Theory,

Journal of Mathematical Physics Vol. 32, pp. 1755–1766 (1991).
[112] Maldacena, J. The Large-N Limit of Superconformal Field Theories

and Supergravity, International Journal of Theoretical Physics Vol. 38,
pp. 1113–1133 (1999).
[113] Mason, L.J. Twistor Actions for Non-Self-Dual Fields; A New Foun-
dation for Twistor-String Theory, Journal of High Energy Physics Vol.
0510, 009, 22 pages (2005).
[114] Menotti, P. and Pelissetto, A. Poincaré, de Sitter, and Conformal

Gravity on the Lattice, Physical Review D Vol. 35, pp. 1194–1204
(1987).
[115] Mercati, F., Mazón, D., Amelino–Camelia, G., Carmona, J.M., Cortés,
J.L., Induráin, J., Lammerzahl, C. and Tino, G.M. Probing the
Quantum-Gravity Realm with Slow Atoms, Classical and Quantum
Gravity Vol. 27, 18 pages, 215003 (2010).
[116] Misner, C.W. Feynman’s Quantization of General Relativity, Reviews

of Modern Physics Vol. 57, pp. 497–509 (1957).
[117] Moss, I.G. and Silva, P.J. BRST Invariant Boundary Conditions for
Gauge Theories, Physical Review D Vol. 55, pp. 1072–1078 (1997).
54
[118] Page, D.N. A Compact Rotating Gravitational Instanton, Physics Let-
ters B Vol. 79, pp. 235–238 (1978).
[119] Page, D.N. A Physical Picture of the K3 Gravitational Instanton,

Physics Letters B Vol. 80, pp. 55–57 (1978).
[120] Page, D.N. and Geilker, C.D. Indirect Evidence of Quantum Gravity,
Physical Review Letters Vol. 47, pp. 979–982 (1981).
[121] Penrose, R. Gravitational Collapse and Space-Time Singularities, Phys-

[122] Penrose, R. and MacCallum, M.A.H. Twistor Theory: an Approach to

the Quantisation of Fields and Space-Time, Physics Reports Vol. 6, pp.
241–316 (1972).
[123] Penrose, R. Twistor Theory and the Einstein Vacuum, Classical and
Quantum Gravity Vol. 16, pp. A113–A130 (1999).
[124] Polchinski, J. Dirichlet Branes and Ramond–Ramond Charges, Physi-

cal Review Letters Vol. 75, pp. 4724–4727 (1995).
[125] Pope, C.N. (1981) The Role of Instantons in Quantum Gravity, in

Quantum Gravity, a Second Oxford Symposium, pp. 377–392, eds. C.J.
Isham, R. Penrose and D.W. Sciama, Clarendon Press, Oxford.
[126] Randall, L. and Sundrum, R. A Large Mass Hierarchy From a Small

Extra Dimension, Physical Review Letters Vol. 83, pp. 3370–3373
(1999).
[127] Randall, L. and Sundrum, R. An Alternative to Compactification, Phys-

[128] Rarita, W. and Schwinger, J. On a Theory of Particles with Half-

Integral Spin, Physical Review Vol. 60, p. 61 (1941).
[129] Reuter, M. Nonperturbative Evolution Equation for Quantum Gravity,

Physical Review D Vol. 57, pp. 971–985 (1998).
[130] Reuter, M. and Saueressig, F. Renormalization Group Flow of Quan-

tum Gravity in the Einstein–Hilbert Truncation, Physical Review D
Vol. 65, 065016, 26 pages (2002).
55
[131] Rosenfeld, L. Zur Quantelung der Wellenfelder, Annalen Physik Vol.
5, pp. 113–152 (1930).
[132] Rosenfeld, L. Uber die Gravitationswirkungen des Lichtes, Zeitschrift

Fur Physik Vol. 65, pp. 589–599 (1930).
[133] Rovelli, C. and Smolin, L. Knot Theory and Quantum Gravity, Physical
Review Letters Vol. 61, pp. 1155–1158 (1988).
[134] Rovelli, C. and Smolin, L. Loop Space Representation of Quantum Gen-

eral Relativity, Nuclear Physics B Vol. 331, pp. 80–152 (1990).
[135] Rovelli, C. Notes For a Brief History of Quantum Gravity,

arXiv:gr-qc/0006061 (2001).
[136] Rovelli, C. (2004) Quantum Gravity, Cambridge University Press,

Cambridge.
[137] Rubakov, V.A. and Shaposhnikov, M.E. Do We Live Inside a Domain

Wall?, Physics Letters B Vol. 125, pp. 136–138 (1983).
[138] Rubakov, V.A. and Shaposhnikov, M.E. Extra Space-Time Dimen-

sions: Towards a Solution to the Cosmological Constant Problem,
Physics Letters B Vol. 125, pp. 139–143 (1983).
[139] Sakharov, A.D. Vacuum Quantum Fluctuations in Curved Space and

the Theory of Gravitation, Soviet Physics Doklady Vol. 12, pp. 1040–
1041 (1968).
[140] Schleich, K. Semiclassical Wave Function of the Universe at Small

Three-Geometry, Physical Review D Vol. 32, pp. 1889–1898 (1985).
[141] Seeley, R. Complex Powers of an Elliptic Operator, American Math-

ematical Society Proceedings Symposium Pure Mathematics Vol. 10,
pp. 288–307 (1967).
[142] Stelle, K. Renormalization of Higher-Derivative Quantum Gravity,

Physical Review D Vol. 16, pp. 953–969 (1977).
[143] Streater, R.F. (2007). Lost Causes in and Beyond Physics, Springer,
New York.
56
[144] Thiemann, T. (2007) Modern Canonical Quantum General Relativity,
Cambridge University Press, Cambridge.
[145] Thiemann, T. Loop Quantum Gravity, International Journal of Modern

Physics A Vol. 23, pp. 1113–1129 (2008).
[146] Tomasiello, A. New String Vacua from Twistor Spaces, Physical Review
D Vol. 78, 046007, 9 pages (2008).
[147] Townsend, P.K. (1997) Black Holes, DAMTP lecture notes

(arXiv:gr-qc/9707012).
[148] van Nieuwenhuizen, P. Supergravity, Physics Reports Vol. 68, pp. 189–
398 (1981).
[149] van Nieuwenhuizen, P. (2005) Supergravity as a Yang–Mills Theory, in

50 Years of Yang–Mills Theory, pp. 433–456, ed. G. ’t Hooft, World
Scientific, Singapore.
[150] Vassilevich, D.V. Vector Fields on a Disk with Mixed Boundary Condi-
tions, Journal of Mathematical Physics Vol. 36, pp. 3174–3182 (1995).
[151] Vilkovisky, G.A. The Unique Effective Action in Quantum Field The-
ory, Nuclear Physics B Vol. 234, pp. 125–137 (1984).
[152] Vilkovisky, G.A. Effective Action in Quantum Gravity, Classical and

Quantum Gravity Vol. 9, pp. 895–903 (1992).
[153] Vilkovisky, G.A. Kinematics of Evaporating Black Holes, Annals of

Physics (N.Y.) Vol. 321, pp. 2717–2756 (2006).
[154] Vitale, P. A Field-Theoretic Approach to Spin Foam Models in Quan-

tum Gravity, arXiv:1103.4172 [gr-qc] (2011).
[155] Volovik, G.E. (2003). The Universe in a Helium Droplet, International

Series of Monographs on Physics Vol. 117, Clarendon Press, Oxford.
[156] Ward, R.S. and Wells, R.O. (1990) Twistor Geometry and Field The-
ory, Cambridge University Press, Cambridge.
57
[157] Weinberg, S. (1979) Ultraviolet Divergences in Quantum Theories of
Gravitation, in General Relativity: an Einstein Centenary Survey, pp.
790–831, eds. S.W. Hawking and W. Israel, Cambridge University
Press, Cambridge.
[158] Wess, J. and Zumino, B. Supergauge Transformations in Four Dimen-

sions, Nuclear Physics B Vol. 70, pp. 39–50 (1974).
[159] Wheeler, J.A. On the Nature of Quantum Geometrodynamics, Annals

of Physics (N.Y.) Vol. 2, pp. 604–614 (1957).
[160] Wheeler, J.A. and Zurek, W.H. Eds. (1983) Quantum Theory and Mea-
surement, Princeton University Press, Princeton.
[161] Wightman, A.S. How It Was Learned that Quantum Fields Are
Operator-Valued Distributions, Fortschritte der Physik Vol. 44, pp.
143–178 (1996).
[162] Witten, E. Comments on String Theory, hep-th/0212247 (2002).
[163] Witten, E. Perturbative Gauge Theory as a String Theory in Twistor

Space, Communications in Mathematical Physics Vol. 252, pp. 189–258
(2004).
[164] Yang, C.N. and Mills, R.L. Conservation of Isotopic Spin and Isotopic
Gauge Invariance, Physical Review Vol. 96, pp. 191–195 (1954).
[165] York, J.W. Boundary Terms in the Action Principles of General Rela-
tivity, Foundations of Physics Vol. 16, pp. 249–257 (1986).
[166] http://aether.lbl.gov/www/projects/cobe; http://map.gsfc.nasa.gov
58

An Introduction To Quantum Gravity

Uploaded by

Copyright:

Available Formats

An Introduction To Quantum Gravity

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

An Introduction To Quantum Gravity

Uploaded by

Copyright:

Available Formats

An introduction to quantum gravity

arXiv:1108.3269v1 [hep-th] 16 Aug 2011

Moreover, a topological relation holds:

which leads to [53]

2 Classical and quantum foundations

2.2 From Schrödinger to Feynman

where the state vector ψ belongs to H, while Ĥ is the Hamiltonian

where −△ is the Laplacian in Cartesian coordinates on R3 (with this

[q̂ j , q̂ k ] = [p̂j , p̂k ] = 0, (8)

[q̂ j , p̂k ] = ih̄δ jk . (9)

where dµ is a suitable (putative) measure on the set of all spacetime

It should be stressed that quantum mechanics regards wave functions

2.3 Spacetime singularities

2.4 Unification of all fundamental interactions

This calculation shows that the Hi are generators of three-dimensional co-

so that they also commute;

S(Σ) ≡ Riem(Σ)/Dif f (Σ).

In this quotient space, two Riemannian metrics on Σ are identified if they

4 From old to new unification

4.1 Old unification

that leave the action invariant, i.e. [39]

4.2 Type-I gauge theories

and the second-order differential operator acting on the potential in the

Peµν = −δµν + Rµν ,

which is the standard wave operator on vectors in curved spacetime. Such

δϕi = Qiα δξ α , (56)

4.3 Type-II gauge theories

4.4 Type-III gauge theories

4.5 From general relativity to supergravity

δψµ (x) = κ−1 Dµ ǫ(x), (69)

δeaµ (x) = iκǫ(x)γ a ψµ (x), (70)

4.6 New unification

S = S4 [gαβ (x)] + S5 [GAB (X)], (73)

while the effective action [38] Γ is formally given by

[RB , RD ] = C ABD RA , (77)

[Rµ , Rν ] = C λµν Rλ . (78)

QAB ≡ RB F A = F A,a RaB , (79)

J µν ≡ Rν χµ = χµ,i Riν , (80)

where µ is taken to be a suitable measure, the SC , T D are ghost fields, and

(i) The infinite-dimensional Lie group (or pseudo-group) of spacetime dif-

5 Functional integrals and background fields

IE [g, φ] = IE [g0 , φ0 ] + I2 [g, φ] + higher − order terms, (87)

so that the logarithm of the quantum-gravity amplitude Ae can be expressed

where the eigenfunctions φn are normalized so that

where the normalization parameter µ has dimensions of mass or (length)−1 .

5.2 Zeta-function regularization of functional integrals

(4) The corresponding integrated heat kernel is then

5.3 Gravitational instantons

5.3.1 Asymptotically locally Euclidean instantons

Following [125], V = V (x) and γ = γ(x) on an auxiliary flat 3-space with

one obtains the desired asymptotically locally Euclidean multi-instantons. In

5.3.2 Asymptotically flat instantons

where z and y obey the relations

In other words, r = M is a nut, using the terminology in Ref. [69].

5.3.3 Compact instantons

gIII = dβ ⊗ dβ + (cos β)2 (ω1 )2 + (sin β)2 (ω2 )2 + (ω3 )2 , (129)

where x ∈ [0, 1], ν is the positive root of

and the functions f1 , f2 , f3 are defined by

6.1 Eigenvalue condition for scalar modes