4.1 Classical models

4.1.1 Classical sound

Sound in a non-relativistic moving fluid has already been extensively discussed in Section 2, and we will not repeat such discussion here. In contrast, sound in a solid exhibits its own distinct and interesting features, notably in the existence of a generalization of the normal notion of birefringence – longitudinal modes travel at a different speed (typically faster) than do transverse modes. This may be viewed as an example of an analogue model which breaks the “light cone” into two at the classical level; as such this model is not particularly useful if one is trying to simulate special relativistic kinematics with its universal speed of light, though it may be used to gain insight into yet another way of “breaking” Lorentz invariance (and the equivalence principle).

4.1.2 Sound in relativistic hydrodynamics

When dealing with relativistic sound, key historical papers are those of Moncrief [448Jump To The Next Citation Point] and Bilic [72Jump To The Next Citation Point], with astrophysical applications being more fully explored in [162, 161, 160], and with a more recent pedagogical follow-up in [639Jump To The Next Citation Point]. It is convenient to first quickly motivate the result by working in the limit of relativistic ray acoustics where we can safely ignore the wave properties of sound. In this limit we are interested only in the “sound cones”. Let us pick a curved manifold with physical spacetime metric g μν, and a point in spacetime where the background fluid 4-velocity is μ V while the speed of sound is cs. Now (in complete direct conformity with our discussion of the generalised optical Gordon metric) adopt Gaussian normal coordinates so that gμν → ημν, and go to the local rest frame of the fluid, so that Vμ → (1;βƒ—0) and

⌊ ⌋ ⌊ ⌋ − 1 0 0 0 0 0 0 0 || 0 1 0 0|| || 0 1 0 0|| gμν → ⌈ 0 0 1 0⌉ ; h μν = gμν + VμVν → ⌈ 0 0 1 0⌉ . (145 ) 0 0 0 1 0 0 0 1
In the rest frame of the fluid the sound cones are (locally) given by
− c2sdt2 + ||d βƒ—x||2 = 0, (146 )
implying in these special coordinates the existence of an acoustic metric
⌊ 2 ⌋ − cs 0 0 0 || 0 1 0 0|| 𝒒μν ∝ ⌈ 0 0 1 0⌉ . (147 ) 0 0 0 1
That is, transforming back to arbitrary coordinates:
2 { 2} 𝒒 μν ∝ − csV μVν + {gμν + VμV ν} ∝ gμν + 1 − cs V μVν. (148 )
Note again that in the ray acoustics limit, because one only has the sound cones to work with, one can neither derive (nor is it even meaningful to specify) the overall conformal factor. When going beyond the ray acoustics limit, seeking to obtain a relativistic wave equation suitable for describing physical acoustics, all the “fuss” is simply over how to determine the overall conformal factor (and to verify that one truly does obtain a d’Alembertian equation using the conformally-fixed acoustic metric).

One proceeds by combining the relativistic Euler equation, the relativistic energy equation, an assumed barotropic equation of state, and assuming a relativistic irrotational flow of the form [639Jump To The Next Citation Point]

μ gμν∇-νΘ- V = ||∇ Θ|| . (149 )
In this situation the relativistic Bernoulli equation can be shown to be
∫ p ---dp--- ln ||∇ Θ || = + 0 Ο±(p) + p, (150 )
where we emphasize that Ο± is now the energy density (not the mass density), and the total particle number density can be shown to be
[ ] ∫ Ο±(p) dΟ± n (p) = n(p=0) exp -------- . (151 ) Ο±(p=0)Ο± + p(Ο±)
After linearization around some suitable background [448, 72, 639Jump To The Next Citation Point], the perturbations in the scalar velocity potential Θ can be shown to satisfy a dimension-independent d’Alembertian equation
{ n2 [ 1 ] } ∇ μ ----0-- − -2V μ0 V0ν+ hμν ∇ νΘ1 = 0, (152 ) Ο±0 + p0 cs
which leads to the identification of the relativistic acoustic metric as
√ ---- n2 [ 1 ] − 𝒒 𝒒μν = ---0--- − -2V μ0 V ν0 + hμν . (153 ) Ο±0 + p0 cs
The dimension-dependence now comes from solving this equation for 𝒒 μν. Therefore, we finally have the (contravariant) acoustic metric
( ) −2βˆ•(d−1) μν -n20c−s1- 𝒒 = Ο±0 + p0 (154 ) { } − -1 VμV ν + hμν , (155 ) c2s 0 0
and (covariant) acoustic metric
( 2 −1 )2βˆ•(d−1) 𝒒 μν = -n0cs-- (156 ) Ο±0 + p0 {− c2[V ] [V ] + h }. (157 ) s 0 μ 0 ν μν
In the non-relativistic limit p0 β‰ͺ Ο±0 and Ο±0 ≈ ¯mn0, where ¯m is the average mass of the particles making up the fluid (which by the barotropic assumption is a time-independent and position-independent constant). So in the non-relativistic limit we recover the standard result for the conformal factor [639]
2 −1 -n0cs-- → -n0- = -1- ρ0-∝ ρ0. (158 ) Ο±0 + p0 ¯mcs m¯2 cs cs
Under what conditions is the fully general relativistic discussion of this section necessary? (The non-relativistic analysis is, after all, the basis of the bulk of the work in “analogue spacetimes”, and is perfectly adequate for many purposes.) The current analysis will be needed in three separate situations:

4.1.3 Shallow water waves (gravity waves)

A wonderful example of the occurrence of an effective metric in nature is that provided by gravity waves in a shallow basin filled with liquid [560Jump To The Next Citation Point]. (See Figure 10View Image.)18 If one neglects the viscosity and considers an irrotational flow, v = ∇ Ο•, one can write Bernoulli’s equation in the presence of Earth’s gravity as

1 p ∂tΟ• + -(∇ Ο• )2 = − --− gz − Vβˆ₯. (159 ) 2 ρ
Here ρ is the density of the fluid, p its pressure, g the gravitational acceleration and V βˆ₯ a potential associated with some external force necessary to establish an horizontal flow in the fluid. We denote that flow by βˆ₯ vB. We must also impose the boundary conditions that the pressure at the surface, and the vertical velocity at the bottom, both vanish. That is, p(z = hB ) = 0, and v⊥ (z = 0) = 0.

Once a horizontal background flow is established, one can see that the perturbations of the velocity potential satisfy

βˆ₯ δp- ∂tδ Ο• + vB ⋅ ∇ βˆ₯δΟ• = − ρ . (160 )
If we now expand this perturbation potential in a Taylor series
∞ n δΟ• = ∑ z-δΟ• (x, y), (161 ) n! n n=0
it is not difficult to prove [560Jump To The Next Citation Point] that surface waves with long wavelengths (long compared with the depth of the basin, λ ≫ hB), can be described to a good approximation by δΟ•0(x, y) and that this field “sees” an effective metric of the form
2 1-[ 2 βˆ₯2 2 βˆ₯ ] ds = c2 − (c − vB )dt − 2v B ⋅ dxdt + dx ⋅ dx , (162 )
where √ ---- c ≡ ghB. The link between small variations of the potential field and small variations of the position of the surface is provided by the following equation
2 βˆ₯ -d- δv ⊥ = − hB∇ βˆ₯δΟ•0 = ∂tδh + v B ⋅ ∇ βˆ₯δh = dt δh. (163 )
The entire previous analysis can be generalised to the case in which the bottom of the basin is not flat, and the background flow not purely horizontal [560Jump To The Next Citation Point]. Therefore, one can create different effective metrics for gravity waves in a shallow fluid basin by changing (from point to point) the background flow velocity and the depth, hB (x, y).
View Image

Figure 10: Gravity waves in a shallow fluid basin with a background horizontal flow.

The main advantage of this model is that the velocity of the surface waves can very easily be modified by changing the depth of the basin. This velocity can be made very slow, and consequently, the creation of ergoregions should be relatively easier than in other models. As described here, this model is completely classical given that the analogy requires long wavelengths and slow propagation speeds for the gravity waves. Although the latter feature is convenient for the practical realization of analogue horizons, it is a disadvantage in trying to detect analogue Hawking radiation as the relative temperature will necessarily be very low. (This is why, in order to have a possibility of experimentally observing (spontaneous) Hawking evaporation and other quantum phenomena, one would need to use ultra-cold quantum fluids.) However, the gravity wave analogue can certainly serve to investigate the classical phenomena of mode mixing that underlies the quantum processes.

It is this particular analogue model (and its extensions to finite depth and surface tension) that underlies the experimental [532Jump To The Next Citation Point] and theoretical [531Jump To The Next Citation Point] work of Rousseaux et al., the historically-important experimental work of Badulin et al. [17Jump To The Next Citation Point], and the very recent experimental verification by Weinfurtner et al. of the existence of classical stimulated Hawking radiation [682Jump To The Next Citation Point].

4.1.4 More general water waves

If one moves beyond shallow-water surface waves the physics becomes more complicated. In the shallow-water regime (wavelength λ much greater than water depth d) the co-moving dispersion relation is a simple linear one ω = csk, where the speed of sound can depend on both position and time. Once one moves to finite-depth (λ ∼ d) or deep (λ β‰ͺ d) water, it is a standard result that the co-moving dispersion relation becomes

∘ ---------- ∘ ------------ tanh(kd) ω = gktanh (kd) = csk ---------. (164 ) kd
See, for instance, Lamb [370Jump To The Next Citation Point] §228, p. 354, Equation (5). A more modern discussion in an analogue spacetime context is available in [643Jump To The Next Citation Point]. Adding surface tension requires a brief computation based on Lamb [370] §267 p. 459, details can be found in [643Jump To The Next Citation Point]. The net result is
∘ ----------∘ ---------- ω = c k 1 + k2βˆ•K2 tanh-(kd). (165 ) s kd
Here K2 = g ρβˆ•σ is a constant depending on the acceleration due to gravity, the density, and the surface tension [643Jump To The Next Citation Point]. Once one adds the effects of fluid motion, one obtains
∘ ---------- ∘ -----2---2- tanh(kd-) ω = v ⋅ k + csk 1 + k βˆ•K kd . (166 )
All of these features, fluid motion, finite depth, and surface tension (capillarity), seem to be present in the 1983 experimental investigations by Badulin et al. [17Jump To The Next Citation Point]. All of these features should be kept in mind when interpreting the experimental [532Jump To The Next Citation Point] and theoretical [531] work of Rousseaux et al., and the very recent experimental work by Weinfurtner et al. [682Jump To The Next Citation Point].

A feature that is sometimes not remarked on is that the careful derivation we have previously presented of the acoustic metric, or in this particular situation the derivation of the shallow-water-wave effective metric [560Jump To The Next Citation Point], makes technical assumptions tantamount to asserting that one is in the regime where the co-moving dispersion relation takes the linear form ω ≈ csk. Once the co-moving dispersion relation becomes nonlinear, the situation is more subtle, and based on a geometric acoustics approximation to the propagation of signal waves one can introduce several notions of conformal “rainbow” metric (momentum-dependent metric). Consider

⌊ 2 2 i j | ⌋ |-−--{c-(k--) −-δijvv-}|+vj-| gab(k2) ∝ | | | , (167 ) ⌈ + vi +hij ⌉ |
and the inverse
⌊ | ⌋ − 1 | +vj ab 2 || -----|---------------|| g (k ) ∝ ⌈ + vi c2(k2)hij − vivj⌉ . (168 ) | |
At a minimum we could think of using the following notions of propagation speed
( | cphase(k2); |||| cgroup(k2); 2 { 2 2 c(k ) → csound = lki→m0 cphase(k ) provided this equals kli→m0 cgroup(k ); (169 ) |||| csignal = lim cphase(k2). |( k→ ∞
Brillouin, in his classic paper [92], identified at least six useful notions of propagation speed, and many would argue that the list can be further refined. Each one of these choices for the rainbow metric encodes different physics, and is useful for different purposes. It is still somewhat unclear as to which of these rainbow metrics is “best” for interpreting the experimental results reported in [17Jump To The Next Citation Point, 532Jump To The Next Citation Point, 682Jump To The Next Citation Point].

4.1.5 Classical refractive index

The macroscopic Maxwell equations inside a dielectric take the well-known form

∇ ⋅ B = 0, ∇ × E + ∂tB = 0, (170 ) ∇ ⋅ D = 0, ∇ × H − ∂tD = 0, (171 )
with the constitutive relations H = μ −1 ⋅ B and D = πœ– ⋅ E. Here, πœ– is the 3 × 3 permittivity tensor and μ the 3 × 3 permeability tensor of the medium. These equations can be written in a condensed way as
( μανβ ) ∂α Z F νβ = 0 (172 )
where Fνβ = A [ν,β] is the electromagnetic tensor,
k F0i = − Fi0 = − Ei, Fij = πœ€ijkB , (173 )
and (assuming the medium is at rest) the non-vanishing components of the 4th-rank tensor Z are given by
0i0j 0ij0 i0j0 i00j 1-ij Z = − Z = Z = − Z = − 2 πœ– ; (174 ) 1 Zijkl =--πœ€ijm πœ€klnμ −m1n; (175 ) 2
supplemented by the conditions that Z is antisymmetric on its first pair of indices and antisymmetric on its second pair of indices. Without significant loss of generality we can ask that Z also be symmetric under pairwise interchange of the first pair of indices with the second pair – thus Z exhibits most of the algebraic symmetries of the Riemann tensor, though this appears to merely be accidental, and not fundamental in any way.

If we compare this to the Lagrangian for electromagnetism in curved spacetime

√ --- β„’ = − ggμαgνβF μνFαβ (176 )
we see that in curved spacetime we can also write the electromagnetic equations of motion in the form (172View Equation) where now (for some constant K):
√ ---{ } Z μναβ = K − g gμαgνβ − gμβgνα . (177 )
If we consider a static gravitational field we can always re-write it as a conformal factor multiplying an ultra-static metric
g = Ω2 {− 1 ⊕ g } (178 ) μν ij
then
Z0i0j = − Z0ij0 = Zi0j0 = − Zi00j = − K √ −-ggij; (179 ) ijkl √--- { ik jl iljk} Z = K − g g g − g g . (180 )

The fact that Z is independent of the conformal factor Ω is simply the reflection of the well-known fact that the Maxwell equations are conformally invariant in (3+1) dimensions. Thus, if we wish to have the analogy (between a static gravitational field and a dielectric medium at rest) hold at the level of the wave equation (physical optics) we must satisfy the two stringent constraints

√--- 1 K − ggij = --πœ–ij; (181 ) 2 √--- { ik jl il jk} 1-ijm kln −1 K − g g g − g g = 2πœ€ πœ€ μmn. (182 )
The second of these constraints can be written as
√ --- { ik jl} −1 K − gπœ€ijm πœ€kln g g = μmn. (183 )
In view of the standard formula for 3 × 3 determinants
πœ€ πœ€ {XikXjl } = 2 detXX −1, (184 ) ijm kln mn
this now implies
2K √-gij- = μ−1, (185 ) − g ij
whence
1 √ --- ---- − ggij = μij. (186 ) 2K
Comparing this with
√ --- 2K − ggij = πœ–ij, (187 )
we now have:
ij 2 ij πœ– = 4K μ ; (188 ) ij 4K2--ij ----1----- ij g = detπœ– πœ– = 4K2 detμ μ . (189 )
To rearrange this, introduce the matrix square root 1βˆ•2 ij [μ ], which always exists because μ is real positive definite and symmetric. Then
[ ] { 1βˆ•2 1βˆ•2}1 βˆ•2 ij gij = μ---πœ–μ---- . (190 ) det(μπœ–)
Note that if you are given the static gravitational field (in the form Ω, gij) you can always solve it to find an equivalent analogue in terms of permittivity/permeability (albeit an analogue that satisfies the mildly unphysical constraint πœ– ∝ μ).19 On the other hand, if you are given permeability and permittivity tensors πœ– and μ, then it is only for that subclass of media that satisfy πœ– ∝ μ that one can perfectly mimic all of the electromagnetic effects by an equivalent gravitational field. Of course, this can be done provided one only considers wavelengths that are sufficiently long for the macroscopic description of the medium to be valid. In this respect it is interesting to note that the behaviour of the refractive medium at high frequencies has been used to introduce an effective cutoff for the modes involved in Hawking radiation [523Jump To The Next Citation Point]. We shall encounter this model (which is also known in the literature as a solid state analogue model) later on when we consider the trans-Planckian problem for Hawking radiation. Let us stress that if one were able to directly probe the quantum effective photons over a dielectric medium, then one would be dealing with a quantum analogue model instead of a classical one.

Eikonal approximation:
With a bit more work this discussion can be extended to a medium in motion, leading to an extension of the Gordon metric. Alternatively, one can agree to ask more limited questions by working at the level of geometrical optics (adopting the eikonal approximation), in which case there is no longer any restriction on the permeability and permittivity tensors. To see this, construct the matrix

C μν = Z μανβkαkβ. (191 )
The dispersion relations for the propagation of photons (and therefore the sought for geometrical properties) can be obtained from the reduced determinant of C (notice that the [full] determinant of C is identically zero as Cμνk ν = 0; the reduced determinant is that associated with the three directions orthogonal to k = 0 ν). By choosing the gauge A = 0 0 one can see that this reduced determinant can be obtained from the determinant of the 3 × 3 sub-matrix ij C. This determinant is
det(Cij ) = 1det (− ω2πœ–ij + πœ€ikmπœ€jlnμ −1k k ) (192 ) 8 mn k l
or, after making some manipulations,
ij 1- [ 2ij − 1 ij kl im jl ] det(C ) = 8 det − ω πœ– + (det μ ) (μ μ kkkl − μ km μ kl) . (193 )
To simplify this, again introduce the matrix square roots 1βˆ•2 ij [μ ] and −1βˆ•2 [μ ]ij, which always exist because the relevant matrices are real positive definite and symmetric. Then define
&tidle;i 1βˆ•2ij k = [μ ] kj (194 )
and
[&tidle;πœ–]ij = det(μ)[μ −1βˆ•2πœ–μ −1βˆ•2]ij (195 )
so that
{ } det(Cij) ∝ det − ω2[&tidle;πœ–]ij + δij[δmn&tidle;km &tidle;kn ] − &tidle;ki&tidle;kj . (196 )
The behaviour of this dispersion relation now depends critically on the way that the eigenvalues of &tidle;πœ– are distributed.

3 degenerate eigenvalues:
If all eigenvalues are degenerate then πœ–&tidle;= &tidle;πœ–I, implying πœ– ∝ μ but now with the possibility of a position-dependent proportionality factor (in the case of physical optics the proportionality factor was constrained to be a position-independent constant). In this case we now easily evaluate

πœ– = tr(πœ–)μ and &tidle;πœ– = det μ-tr(πœ–)-, (197 ) tr(μ ) tr(μ )
while
ij 2{ 2 −1 &tidle; m&tidle;n }2 det(C ) ∝ ω ω − [&tidle;πœ– δmnk k ] . (198 )
That is
{ } det(Cij) ∝ ω2 ω2 − [gijkikj] 2, (199 )
with
ij 1- ij tr(μ-)[μ-]ij- -tr(πœ–)[πœ–]ij-- g = &tidle;πœ– [μ ] = tr(πœ–)det μ = tr(μ )det πœ–. (200 )
This last result is compatible with but more general than the result obtained under the more restrictive conditions of physical optics. In the situation where both permittivity and permeability are isotropic, (πœ–ij = πœ–δij and μij = μδij) this reduces to the perhaps more expected result
ij δij g = ---. (201 ) πœ–μ

2 distinct eigenvalues:
If &tidle;πœ– has two distinct eigenvalues then the determinant det(Cij) factorises into a trivial factor of ω2 and two quadratics. Each quadratic corresponds to a distinct effective metric. This is the physical situation encountered in uni-axial crystals, where the ordinary and extraordinary rays each obey distinct quadratic dispersion relations [82Jump To The Next Citation Point]. From the point of view of analogue models this corresponds to a two-metric theory.

3 distinct eigenvalues:
If &tidle;πœ– has three distinct eigenvalues then the determinant det(Cij) is the product of a trivial factor of ω2 and a non-factorizable quartic. This is the physical situation encountered in bi-axial crystals [82Jump To The Next Citation Point, 638Jump To The Next Citation Point], and it seems that no meaningful notion of the effective Riemannian metric can be assigned to this case. (The use of Finsler geometries in this situation is an avenue that may be worth pursuing [306Jump To The Next Citation Point]. But note some of the negative results obtained in [573Jump To The Next Citation Point, 574Jump To The Next Citation Point, 575Jump To The Next Citation Point].)

Abstract linear electrodynamics:
Hehl and co-workers have championed the idea of using the linear constitutive relations of electrodynamics as the primary quantities, and then treating the spacetime metric (even for flat space) as a derived concept. See [474, 276, 371, 277].

Nonlinear electrodynamics:
In general, the permittivity and permeability tensors can be modified by applying strong electromagnetic fields (this produces an effectively nonlinear electrodynamics). The entire previous discussion still applies if one considers the photon as the linear perturbation of the electromagnetic field over a background configuration

bg ph Fμν = F μν + f μν. (202 )
The background field F bg μν sets the value of πœ–ij(F bg), and μij(F bg). Equation (172View Equation) then becomes an equation for ph fμν. This approach has been extensively investigated by Novello and co-workers [465, 469, 170, 468, 466, 467, 464, 214].

Summary:
The propagation of photons in a dielectric medium characterised by 3 × 3 permeability and permittivity tensors constrained by πœ– ∝ μ is equivalent (at the level of geometric optics) to the propagation of photons in a curved spacetime manifold characterised by the ultra-static metric (200View Equation), provided one only considers wavelengths that are sufficiently long for the macroscopic description of the medium to be valid. If, in addition, one takes a fluid dielectric, by controlling its flow one can generalise the Gordon metric and again reproduce metrics of the Painlevé–Gullstrand type, and therefore geometries with ergo-regions. If the proportionality constant relating πœ– ∝ μ is position independent, one can make the stronger statement (189View Equation) which holds true at the level of physical optics. Recently this topic has been revitalised by the increasing interest in (classical) meta-materials.

4.1.6 Normal mode meta-models

We have already seen how linearizing the Euler–Lagrange equations for a single scalar field naturally leads to the notion of an effective spacetime metric. If more than one field is involved the situation becomes more complicated, in a manner similar to that of geometrical optics in uni-axial and bi-axial crystals. (This should, with hindsight, not be too surprising since electromagnetism, even in the presence of a medium, is definitely a Lagrangian system and definitely involves more than one single scalar field.) A normal mode analysis based on a general Lagrangian (many fields but still first order in derivatives of those fields) leads to a concept of refringence, or more specifically multi-refringence, a generalization of the birefringence of geometrical optics. To see how this comes about, consider a straightforward generalization of the one-field case.

We want to consider linearised fluctuations around some background solution of the equations of motion. As in the single-field case we write (here we will follow the notation and conventions of [45Jump To The Next Citation Point])

A A A πœ–2 A 3 Ο• (t,βƒ—x) = Ο• 0(t,βƒ—x) + πœ–Ο• 1(t,βƒ—x) + 2 Ο•2 (t,βƒ—x ) + O (πœ– ). (203 )
Now use this to expand the Lagrangian
[ ∂β„’ ∂β„’ ] β„’ (∂μΟ•A, Ο•A) = β„’ (∂μΟ•A0 ,Ο•A0) + πœ– -----A--∂μΟ•A1 + --A-Ο•A1 [ ∂(∂μΟ• ) ] ∂Ο• πœ–2 ∂β„’ A ∂β„’ A + -2 ∂-(∂--Ο•A)∂ μΟ•2 + ∂-Ο•AΟ• 2 [ μ πœ–2 ∂2β„’ A B + -- ------A-------B-∂μΟ• 1∂νΟ• 1 2 ∂(∂μΟ• )∂(∂νΟ• ) ] ∂2 β„’ ∂2β„’ +2 ------------∂ μΟ•A1Ο•B1 + --------Ο•A1Ο•B1 ∂ (∂ μΟ•A)∂ Ο•B ∂Ο•A ∂Ο•B 3 +O (πœ–). (204 )
Consider the action
∫ S[Ο•A] = dd+1xβ„’ (∂μΟ•A,Ο•A ). (205 )
Doing so allows us to integrate by parts. As in the single-field case we can use the Euler–Lagrange equations to discard the linear terms (since we are linearizing around a solution of the equations of motion) and so get
A A S[Ο• ] = S [Ο• 0] [ πœ–2 ∫ { ∂2β„’ } + -- dd+1x ------A------B-- ∂μΟ•A1∂νΟ•B1 2 ∂(∂μΟ• )∂(∂νΟ• ) { 2 } { 2 } ] +2 ----∂-β„’----- ∂ Ο•A Ο•B + --∂--β„’-- Ο•AΟ•B ∂(∂μΟ•A )∂Ο•B μ 1 1 ∂ Ο•A∂ Ο•B 1 1 3 + O (πœ– ). (206 )
Because the fields now carry indices (AB) we cannot cast the action into quite as simple a form as was possible in the single-field case. The equation of motion for the linearised fluctuations are now read off as
( { } ) ( ) ∂2β„’ B ∂2β„’ B ∂μ ------A------B-- ∂νΟ• 1 + ∂ μ ------A----BΟ•1 ∂(∂μΟ• )∂(∂νΟ• ) ( ) ∂(∂μΟ• )∂Ο• B ----∂2β„’----- --∂2β„’--- B − ∂μΟ•1 ∂(∂μ Ο•B)∂Ο•A − ∂Ο•A ∂Ο•B Ο• 1 = 0. (207 )
This is a linear second-order system of partial differential equations with position-dependent coefficients. This system of PDEs is automatically self-adjoint (with respect to the trivial “flat” measure dd+1x).

To simplify the notation we introduce a number of definitions. First

1 ( ∂2β„’ ∂2β„’ ) fμνAB ≡ -- ----------------+ ---------------- . (208 ) 2 ∂(∂μΟ•A )∂ (∂νΟ•B ) ∂(∂νΟ•A )∂ (∂μΟ•B )
This quantity is independently symmetric under interchange of μ, ν and A, B. We will want to interpret this as a generalization of the “densitised metric”, μν f, but the interpretation is not as straightforward as for the single-field case. Next, define
2 2 μ ----∂-β„’----- ----∂-β„’----- Γ AB ≡ + ∂(∂μΟ•A )∂Ο•B − ∂(∂μΟ•B )∂Ο•A ( 2 2 ) + 1∂ν ------∂-β„’-------− ------∂-β„’------- . (209 ) 2 ∂(∂νΟ•A )∂ (∂μΟ•B ) ∂(∂μΟ•A )∂(∂νΟ•B )
This quantity is anti-symmetric in A, B. One might want to interpret this as some sort of “spin connection”, or possibly as some generalization of the notion of “Dirac matrices”. Finally, define
2 ( 2 ) ( 2 ) KAB = − ---∂-β„’-- + 1∂μ ----∂-β„’----- + 1-∂μ ----∂--β„’---- . (210 ) ∂ Ο•A∂ Ο•B 2 ∂(∂μΟ•A )∂ Ο•B 2 ∂ (∂ μΟ•B)∂ Ο•A
This quantity is by construction symmetric in (AB ). We will want to interpret this as some sort of “potential” or “mass matrix”. Then the crucial point for the following discussion is to realise that Equation (207View Equation) can be written in the compact form
( μν B ) 1-[ μ B μ B ] B ∂ μ f AB ∂νΟ•1 + 2 ΓAB ∂μΟ•1 + ∂μ (Γ ABΟ• 1 ) + KAB Ο• 1 = 0. (211 )
Now it is more transparent that this is a formally self-adjoint second-order linear system of PDEs. Similar considerations can be applied to the linearization of any hyperbolic system of second-order PDEs.

Consider an eikonal approximation for an arbitrary direction in field space; that is, take

Ο•A(x) = πœ–A(x )exp[− iφ (x)], (212 )
with πœ–A (x ) a slowly varying amplitude, and φ(x ) a rapidly varying phase. In this eikonal approximation (where we neglect gradients in the amplitude, and gradients in the coefficients of the PDEs, retaining only the gradients of the phase) the linearised system of PDEs (211View Equation) becomes
μν μ B {f AB∂μφ (x)∂νφ (x ) + Γ AB∂ μφ(x) + KAB } πœ– (x) = 0. (213 )
This has a nontrivial solution if and only if πœ–A(x ) is a null eigenvector of the matrix
fμνABk μkν + Γ μABk μ + KAB, (214 )
where kμ = ∂μφ (x). Now, the condition for such a null eigenvector to exist is that
F(x, k) ≡ det{f μν k k + Γ μ k + K } = 0, (215 ) AB μ ν AB μ AB
with the determinant to be taken on the field space indices AB. This is the natural generalization to the current situation of the Fresnel equation of birefringent optics [82, 375]. Following the analogy with the situation in electrodynamics (either nonlinear electrodynamics, or more prosaically propagation in a birefringent crystal), the null eigenvector A πœ– (x) would correspond to a specific “polarization”. The Fresnel equation then describes how different polarizations can propagate at different velocities (or in more geometrical language, can see different metric structures). In the language of particle physics, this determinant condition F (x,k ) = 0 is the natural generalization of the “mass shell” constraint. Indeed, it is useful to define the mass shell as a subset of the cotangent space by
{ || } β„± (x) ≡ k μ||F (x,k ) = 0 . (216 )
In more mathematical language we are looking at the null space of the determinant of the “symbol” of the system of PDEs. By investigating F(x, k) one can recover part (not all) of the information encoded in the matrices μν f AB, μ Γ AB, and KAB, or equivalently in the “generalised Fresnel equation” (215View Equation). (Note that for the determinant equation to be useful it should be non-vacuous; in particular one should carefully eliminate all gauge and spurious degrees of freedom before constructing this “generalised Fresnel equation”, since otherwise the determinant will be identically zero.) We now want to make this analogy with optics more precise, by carefully considering the notion of characteristics and characteristic surfaces. We will see how to extract from the the high-frequency high-momentum regime described by the eikonal approximation all the information concerning the causal structure of the theory.

One of the key structures that a Lorentzian spacetime metric provides is the notion of causal relationships. This suggests that it may be profitable to try to work backwards from the causal structure to determine a Lorentzian metric. Now the causal structure implicit in the system of second-order PDEs given in Equation (211View Equation) is described in terms of the characteristic surfaces, and it is for this reason that we now focus on characteristics as a way of encoding causal structure, and as a surrogate for some notion of a Lorentzian metric. Note that, via the Hadamard theory of surfaces of discontinuity, the characteristics can be identified with the infinite-momentum limit of the eikonal approximation [265]. That is, when extracting the characteristic surfaces we neglect subdominant terms in the generalised Fresnel equation and focus only on the leading term in the symbol (μν f AB). In the language of particle physics, going to the infinite-momentum limit puts us on the light cone instead of the mass shell; and it is the light cone that is more useful in determining causal structure. The “normal cone” at some specified point x, consisting of the locus of normals to the characteristic surfaces, is defined by

{ || } 𝒩 (x) ≡ kμ|det (fμνABk μkμ) = 0 . (217 ) |

As was the case for the Fresnel Equation (215View Equation), the determinant is to be taken on the field indices AB. (Remember to eliminate spurious and gauge degrees of freedom so that this determinant is not identically zero.) We emphasise that the algebraic equation defining the normal cone is the leading term in the Fresnel equation encountered in discussing the eikonal approximation. If there are N fields in total then this “normal cone” will generally consist of N nested sheets each with the topology (not necessarily the geometry) of a cone. Often several of these cones will coincide, which is not particularly troublesome, but unfortunately it is also common for some of these cones to be degenerate, which is more problematic.

It is convenient to define a function Q (x,k) on the co-tangent bundle

Q (x,k) ≡ det (fμνAB (x )kμkμ). (218 )
The function Q(x,k ) defines a completely-symmetric spacetime tensor (actually, a tensor density) with 2N indices
Q (x,k) = Q μ1ν1μ2ν2⋅⋅⋅μNνN(x)k k k k ⋅ ⋅⋅k k . (219 ) μ1 ν1 μ2 ν2 μN νN
(Remember that f μνAB is symmetric in both μν and AB independently.) Explicitly, using the expansion of the determinant in terms of completely antisymmetric field-space Levi–Civita tensors
μ1ν1μ2ν2⋅⋅⋅μN νN -1- A1A2⋅⋅⋅AN B1B2⋅⋅⋅BN μ1ν1 μ2ν2 μN νN Q = N !πœ– πœ– f A1B1f A2B2 ⋅⋅⋅f ANBN . (220 )
In terms of this Q (x,k) function, the normal cone is
{ || } 𝒩 (x) ≡ kμ||Q(x,k ) = 0 . (221 )
In contrast, the “Monge cone” (aka “ray cone”, aka “characteristic cone”, aka “null cone”) is the envelope of the set of characteristic surfaces through the point x. Thus the “Monge cone” is dual to the “normal cone”, its explicit construction is given by (Courant and Hilbert [154, vol. 2, p. 583]):
{ || } β„³ (x) = tμ = ∂Q-(x,k)|kμ ∈ 𝒩 (x) . (222 ) ∂k μ |

The structure of the normal and Monge cones encode all the information related with the causal propagation of signals associated with the system of PDEs. We will now see how to relate this causal structure with the existence of effective spacetime metrics, from the experimentally favoured single-metric theory compatible with the Einstein equivalence principle to the most complicated case of pseudo-Finsler geometries [306Jump To The Next Citation Point].

The message to be extracted from this rather formal discussion is that effective metrics are rather general and mathematically robust objects that can arise in quite abstract settings – in the abstract setting discussed here it is the algebraic properties of the object fμνAB that eventually leads to mono-metricity, multi-metricity, or worse. The current abstract discussion also serves to illustrate, yet again,

  1. that there is a significant difference between the levels of physical normal modes (wave equations), and geometrical normal modes (dispersion relations), and
  2. that the densitised inverse metric is in many ways more fundamental than the metric itself.

  Go to previous page Go up Go to next page