## Abstract

In this paper, we investigate a possible energy scale dependence of the quantization rules and, in particular, from a phenomenological point of view, an energy scale dependence of an effective (reduced Planck’s constant). We set a bound on the deviation of the value of at the muon scale from its usual value using measurements of the anomalous magnetic moment of the muon. Assuming that inflation has taken place, we can conclude that nature is described by a quantum theory at least up to an energy scale of about 10^{16} GeV.

## 1. Introduction

There are three constants of nature that are usually assumed to be fundamental and immutable. These are the speed of light *c*, Planck’s constant *h* and Newton’s constant *G*_{N}. The speed of light *c*=299 792 458 m s^{−1} [1] determines the velocity of a massless particle in vacuum and is introduced in special relativity as a conversion factor between time and space. It allows one to define an invariant length element d*s*^{2}=*c*^{2} d*t*^{2}−d*x*_{i} d*x*^{i}. Planck’s constant *h*=6.62606957(29)×10^{−34} J s [1] is the physical constant that enters quantum mechanics and its extension, quantum field theory. Planck’s constant allows one to convert the energy *E* of a particle into the frequency *ν* of the wave associated with this particle using the relation *E*=*hν*. The momentum of the particle *p* is related to de Broglie’s wavelength λ via *E*=*hc*/λ, where *c* is the speed of light in vacuum. It is common to introduce the reduced Planck’s constant, [1]. Note that there are different ways to measure this fundamental constant (e.g. [2]). Among these constants, Newton’s constant *G*_{N}=6.67×10^{−11} N (m kg)^{−2} [1] is on a slightly different footing since it fixes the strength of the gravitational interactions and it gets renormalized (e.g. [3]). It is thus dependent on the energy scale at which it is probed.

Planck’s constant plays an essential role in quantum mechanics and quantum field theory. It is this constant of nature that determines the amount of quantization for both the first quantization and the second quantization. The quantization rules in quantum mechanics and, for example, in quantum field theory both involve .

In this paper, we will show that both the first and second quantization rules, which consist in imposing non-trivial commutation relations for coordinates and momenta (in the case of the first quantization, i.e. quantum mechanics) and for fields and their canonical momenta (in the case of the second quantization, i.e. quantum field theory) can be modified by quantum gravity or modification of quantum mechanics itself.

For example, in some models such as string theory, the uncertainty relation of quantum mechanics gets modified. Amati, Ciafaloni and Veneziano (ACV) have shown [4–7] by studying the scattering of massless strings in the trans-Planckian regime that the usual uncertainty principle of quantum mechanics,
1.1generalizes to
1.2where *α*′=*G*_{N}/*g*^{2}, *G*_{N} is Newton’s constant and *g* is the string loop expansion parameter. This relation is called the generalized uncertainty principle.

While this specific relation was derived in string theory, similar ones appear in different models of quantum gravity; see e.g. [8] for a recent review. One expects such a relation in most theories attempting to unify quantum mechanics and general relativity [9]. Indeed, black holes are expected to form in the collisions of particles at very high energy. For concreteness, let us consider the collision of two particles. It can be shown that, even for a non-zero impact parameter, a black hole will form at centre-of-mass energies above the Planck scale [10]. Imagine that we collide these particles at energies just above the threshold for black hole formation, i.e. just above the Planck mass. The black hole created will have a mass of the order of the Planck mass and thus a size (Schwarzschild radius) of the order of the inverse of the Planck mass. Such a scattering experiment is thus not able to resolve distances shorter than the Schwarzschild radius, corresponding to a black hole with a minimal length of the order of the Planck scale. As we increase the centre-of-mass energy, so the mass of the black hole increases. The black hole becomes larger and the length that can be probed by the scattering experiment increases with energy as well. Thus, as in the case of ACV, increasing the centre-of-mass energy of the scattering experiment does not allow one to resolve shorter distances, as the *Δx* probed by the scattering experiment increases with the centre-of-mass energy. We thus expect generically a modification of the uncertainty relation of the type
1.3where the parameter *α* is positive. That is, if quantum mechanics remains valid up to all energy scales.

Generically speaking, one should be able to derive such generalized uncertainty relations from a modification of the commutation relationship between *x* and *p* [8]:
1.4with *f*(*k*)_{ν}=*p*_{ν} where *f*_{ν} are invertible functions such that their inverses are well defined. One way to think of such models is as an energy scale *μ*-dependent effective ,
1.5in analogy to renormalized coupling constants such as the fine-structure constant of quantum electrodynamics. Note that the non-commutation relations in equation (1.4) should, as we shall see, be derived from a modified Lie algebra. In this paper, we shall investigate whether is, as *G*_{N}, dependent on the energy scale at which it is measured. Such an energy dependence would be motivated not by the necessity to renormalize the theory, but rather by a possible modification of quantum mechanics at short distance or equivalently at very high energy.

Independently of our theoretical arguments, it is an interesting empirical question to investigate whether Planck’s constant has an energy dependence. We have seen that most unifications of general relativity and quantum mechanics are likely to lead to an effective Planck’s constant that increases with energy. There is, however, another framework where the opposite is expected. This framework is that advocated by ’t Hooft, who has been investigating the possibility that quantum mechanics could be an emergent phenomenon [11,12]. While this proposal faces potential difficulties in explaining the Bell inequalities, it is worth looking at it from a phenomenological point of view. If quantum mechanics is an emergent phenomenon, in the sense that our world appears quantum mechanical while the underlying theory at short distance would be classical, must effectively depend on the energy at which we probe it. On quantum scale, takes its usual value, while at the fundamental level, tends to zero, and one recovers a form of deterministic physics, maybe in the form of cellular automata as envisaged by ’t Hooft. It is worth noting that the limit does not imply that loop diagrams, in quantum field theory necessarily all go to zero. It has been pointed out that some parts of loop diagrams correspond to purely classical physics [13,14]. The Feynman diagram expansion, which leads to loop diagrams, is an expansion in the coupling constant of the theory and not in . One should therefore be careful when considering this limit.

We will first show that the assumption that Lorentz invariance is the correct symmetry of nature leads to surprisingly little freedom in terms of how we can modify the value of . We will then show that it is universal in a sense to be defined below.

## 2. Modification of the quantization rules: universality of

It is important to realize that the quantization rules cannot be modified in an arbitrary fashion. As we shall explain shortly, the quantization rules are not axioms of the theory but rather are derived from the foundations of quantum mechanics and quantum field theory. In particular, we shall show that must be universal. The universality of is widely accepted; here we provide a proof.

It is very well understood that the speed of light is universal; its value is the same in all laws of physics that are Lorentz invariant. Similarly, Newton’s constant is universal because of the equivalence principle. The reduced Planck’s constant appears both in quantum mechanics and in quantum field theory. In quantum mechanics, one could naively have a different in each of the quantization rules:
2.1Requiring that *x* and *p* do not commute is sometimes called the first quantization. Note that as usual [*x*_{μ},*x*_{ν}]=0 and [*p*_{μ},*p*_{ν}]=0. The second quantization rule is relevant to quantum field theory, where, for example, for a real scalar field one imposes
2.2with where the dot stands for a time derivative. Furthermore, one has [*ϕ*(*x*),*ϕ*(*y*)]=0 and [*Π*_{μ}(*x*),*Π*_{ν}(*y*)]=0. Why should the in these expressions be identical, and in quantum field theory, why should be universal for all particle species? In other words, why should be a universal constant of nature? It is important to remember that the quantization rules, whether one is looking at the first or the second quantization procedure, are not axioms of quantum mechanics or quantum field theory but rather are derived from the axioms of these theories. Note that, while we will work in the Schrödinger picture, the same reasoning obviously applies to the Heisenberg picture. We will show that there is no freedom at all and that, ultimately because of the structure of the Lie algebra for the generators of the Lorentz group, must be universal.

Let us first consider quantum mechanics. We will use the derivation of the commutator relation presented in Weinberg’s book [15]. We will show that translation invariance implies that in the momentum operator *P* (i.e. in the Schrödinger equation) is the same as in the commutator relationship between the operators *X* and *P*.

As usual, one assumes the existence of an operator *X*_{n} corresponding to the observable *x*_{n}, i.e. the position of a particle, where *n* labels the individual particles. We shall consider translation invariance (physics should not depend on a shift of, for example, the origin of the coordinate system). In other words, transformations *X*_{n}→*X*_{n}+** a**, where

**is an arbitrary three-vector, should not modify the physics. Translation invariance is generated by a unitary operator**

*a**U*(

**): 2.3For an infinitesimal translation, the unitary operator can be written as , where**

*a***is an operator that generates the translation. This operator is called the momentum and we thus set . The constant at this stage is introduced to make dimensionless and its value is only determined experimentally (like**

*T**c*or

*G*

_{N}). It must carry the dimensions m

^{2}kg s

^{−1}. One thus finds 2.4The requirement (2.3) implies that 2.5for any infinitesimal

**. One thus has 2.6where**

*a**X*

_{i}is the

*i*th component of

**and**

*X**P*

_{j}is the

*j*th component of

**, which is the total momentum of the system (if there are several particles in the problem).**

*P*We can now see that if we look at a one-particle state *Φ*(*x*), and define its position by
2.7the commutator (2.6) implies that
2.8

In principle, we could have a different for the different components of *P*_{j}. We need to assume rotational invariance, which will lead to a universal . The requirement of time translation invariance leads to the concept of energy and thus Hamiltonian, and thus to the Schrödinger equation:
2.9Note that at this stage the of *P* and that of *H* could be different. It is only when one requests invariance under Galilean transformation that one finds that the two must be the same. Galilean invariance (and there is a similar one when Lorentz invariance is imposed) implies that
2.10where ** K** is the boost generator. This relation forces one to assume that
2.11where

*V*is a function called the potential, which depends only on the differences of the particle coordinate vectors. The constant is thus universal in quantum mechanics.

A similar reasoning applies to quantum field theory. Following the derivation of [16], we consider one-particle states. They are defined by
2.12where *P*^{μ} is the four-momentum operator, *p*^{μ} is the corresponding eigenvalue and *σ* stands for any other quantum number. Under translations, the one-particle state transforms as
2.13Again here is introduced to compensate the mass dimensions of *p*⋅*a*.

The creation operator *a*^{†} is defined by the way it acts on a multiple-particle state:
2.14The creation operator adds a particle with quantum number *q* to the multiple-particle state. The annihilation operator *a*(*q*) is defined by
2.15with a + and − sign for bosons and fermions, respectively. One can easily show that
2.16where the upper sign applies to bosons and the lower one to fermions. One can then define creation and annihilation fields :
2.17The requirement of invariance under a translation leads to the requirements that
2.18Fields are classified according to the representation *SO*(3,1). The properties of Lorentz invariance as well as C, P, T transformations lead to relationships between *u*_{l} and *v*_{l} which depend on the representation of the Wigner little group under consideration. In any case, these relations ensure that the same appears in both expressions.

We will now specialize to the case of scalar fields, but the reasoning can be trivially extended to any other fields. For a causal scalar field, one finds:
2.19and
2.20The scalar field is then defined by
2.21One finds
2.22with
2.23One can show that
2.24where the dot above a variable now denotes the derivative with respect to *x*^{0}. One then verifies
2.25One may now wonder whether should be the same for all fields and the same in quantum mechanics and quantum field theory. The underlying reason for the universality of is that it appears in the quantum Lorentz algebra.

An infinitesimal Lorentz transformation is given by
2.26One may worry that the appearing in the transformation could be different, but one always has the freedom to normalize the transformations *w*_{μν} and *ϵ*_{σ} in such a way that the same appears in both terms of the equation. Let us calculate
2.27By comparing the coefficients of *w* and *ϵ*, one finds
2.28and
2.29Let us now consider . One finds
2.30and
2.31which leads to the Lie algebra of the Lorentz group:
2.32We see that the appearing in the algebra must be the same in the first two lines.

Let us now introduce the angular momentum operator ** J** given by

*J*

_{i}=

*ϵ*

_{ijk}

*M*

^{jk}and the boost operator given by

*K*

_{i}=

*M*

^{i0}: 2.33

Thus, must be universal to quantum mechanics and quantum field theory if we consider that quantum mechanics is the low-velocity limit of relativistic quantum mechanics. Furthermore, if quantum gravity introduces an energy dependence of , this energy dependence must be universal and affect all processes. Note that, if Lorentz symmetry is violated at high energies, then it is conceivable for to be non-universal at these energies. Let us now study the phenomenology of an energy-dependent assuming that Lorentz invariance is a valid symmetry of nature.

## 3. What do we know about the energy dependence of ?

In this section, we shall investigate modifications of the quantization rules which can be parametrized phenomenologically by an energy scale-dependent Planck’s constant. We thus adopt
3.1in quantum mechanics and, for eaxmple,
3.2for a scalar field, with an obvious generalization for fields carrying a different spin. This parametrization enables us to consider both frameworks. That is, we can consider both cases, where the theory is becoming more classical as short distances are probed, i.e. if decreases with increasing energy scale *μ*, or more quantum in the sense that increases with increasing energy scale. We shall consider two experiments. The first one is the anomalous magnetic moment of the muon, which is a low-energy experiment but one of the most sensitive ones done to date. The second one is a cosmological observation, namely the cosmic microwave background, which probes inflation. While there is a strong model dependence in the second one, this is a very high-energy process (the scale of inflation is close to the Planck scale), and if one assumes that the fluctuations in temperature in the cosmic microwave background are due to fluctuations of the inflaton, one can rule out models where would tend to zero at energy scales close to the Planck scale.

### (a) The anomalous magnetic moment of the muon

The non-relativistic form of the Dirac equation describing a spinor in an electromagnetic field is given by
3.3where ** L**=

**×**

*r***is the angular momentum and**

*p***is the spin of the spinor with eigenvalues . We see immediately that an energy dependence of would account for an energy dependence of the magnetic moment**

*S**g*of fermions, which was famously predicted by Dirac to be equal to 2. It is known that quantum gravitational effects modify the value of the magnetic moment. The anomalous magnetic moment is given by 3.4The anomalous magnetic moment of the muon is one of the best-measured quantities [1], 3.5and also one of the most precisely calculated quantities in the standard model of particle physics, 3.6The difference between experiment and theory, , allows us to set a limit on the energy dependence of . The relevant energy scale that is probed by the anomalous magnetic moment of the muon is that of the muon mass (105.7 MeV

*c*

^{−2}): 3.7At 105.7 MeV, the deviation between an that is energy-dependent and its standard value cannot be larger than 3×10

^{−9}. On the other hand, the small deviation between the standard model prediction and the experimental measurement of the anomalous magnetic moment of the muon, which currently represents 3.6

*σ*deviation, could be interpreted as the first sign that differs from its standard value at 105.7 MeV by 3×10

^{−9}eV s.

### (b) Inflation

While the previous example gives a bound on the scale dependence of at relatively low energy scale, inflation is a process that takes place at a very high energy scale close to the Planck scale. The issue is that we are now dealing with observations and not experiments. In particular, any bound on coming from observations in the cosmic microwave background will be dependent on the specific model of inflation. However, the energy scale of inflation is known to be close to 10^{16} GeV. A common feature of the inflationary model is the prediction of the existence of fluctuations of the temperature in the cosmic microwave background corresponding to quantum fluctuation of the inflaton. Assuming that inflation has really taken place, the observation of these fluctuations leads to the conclusion that, at an energy scale of 10^{16} GeV, physics is still described by a quantum theory. This implies that, at this energy scale, , as one would expect from a model of emergent quantum mechanics. Let us now study this quantitatively.

The standard normalization implies that the perturbation at the present Hubble scale *δ*_{H} is given by [17,18]
3.8unless the slow role parameter *ϵ* is anomalously small. The energy scale that is probed at the start of inflation is thus around 10^{16} GeV. When evaluating the slow role parameters
3.9and
3.10we should thus evaluate them using ). By measuring the slow parameters, we can thus probe the value of close to the Planck mass for a given model of inflation since the slow role parameters depend on the potential *V* and thus on the model. One can also consider the number of e-foldings, which is given by
3.11It is in the range 50–60 and used to extract *ϕ*_{N}. The value of the inflation field at the end of inflation *ϕ*_{e} is calculated from the requirement that *ϵ*>1 and *η*>1 and is also dependent on the sliding . In the limit of , the slow role parameters go to zero. Assuming that the temperature fluctuations in the cosmic microwave background are really due to the fluctuation of the inflaton, we can rule out from current observations that .

While small changes of will be tough to observe and strongly model-dependent, we can conclude that, at 10^{16} GeV, nature is still described by quantum mechanics. If quantum mechanics is emergent, as it would be in ’t Hooft’s model, we must assume that the underlying theory becomes relevant at a very high energy scale above 10^{16} GeV.

## 4. Conclusion

In this paper, we have investigated a possible energy scale dependence of the quantization rules and, in particular, from a phenomenological point of view, an energy scale dependence of an effective . Different models of quantum mechanics and quantum gravity lead to an that is effectively scale-dependent. We have shown that this effective must be universal since modern theories of the world are based on the Lorentz algebra. We have set a bound on the deviation of the value of at the muon scale from its usual value using measurements of the anomalous magnetic moment of the muon. Assuming that inflation has taken place, we can conclude that nature is described by a quantum theory at least up to an energy scale of about 10^{16} GeV. Finally, it is worth mentioning that an energy scale dependence of Planck’s constant has some interesting consequences for thought experiments probing a unification of general relativity and quantum mechanics. Firstly, it would lead to an energy scale dependence of the Planck mass, which is given by . Note that Newton’s constant fixes the strength of gravitational interactions. The limit thus does not imply that the strength of gravitation goes to zero. Secondly, the Planck length , which can be shown to be the minimal length [19–21] that can be measured given our current understanding of quantum mechanics and general relativity, goes to zero in the limit where goes to zero. This can be easily understood since the standard quantum limit goes to zero in that limit. In the limit , the minimal length or Planck length tends to zero. This would be a problem for a quantum mechanical description of general relativity. It is often conjectured that singularities in general relativity are cured or at least hidden by the minimal length. If the minimal length tends to zero, singularities may be observable.

## Competing interests

I declare I have no competing interests.

## Funding

This work is supported in part by the Science and Technology Facilities Council (grant no. ST/J000477/1).

## Footnotes

One contribution of 13 to a theme issue ‘New geometric concepts in the foundations of physics’.

- Accepted April 27, 2015.

- © 2015 The Author(s) Published by the Royal Society. All rights reserved.