Arxiv:1801.07676V3 [Hep-Th] 31 Oct 2018 Neligteeiesteei E a Obekscale Break to Way New That, a Emphasise That Is We Symmetry There However, Ideas Authors [2])

Home , Scale invariance

arXiv:1801.07676v3 [hep-th] 31 Oct 2018 neligteeiesteei e a obekscale break to way new that, a emphasise that is we symmetry there However, ideas authors [2]). these previous see underlying the of (e.g. many above by listed studied been has which transmuta- dimensional [8]. of tion Coleman mechanism the (CW) via Weinberg potentials invariant be im- scale spontaneously, must for treated nature, is plemented like breaking in this symmetry, seen Often symmetries Weyl broken. flavour or the for of Scale symmetry many foundational [2][3][4][5][6][7]. fun- a for physics and context fields scale modern scalar a to provide conjunc- damental turn may in gravity [1] to with symmetry principles tion Weyl particular, organising authors In many new symmetry. led of has search symmetry, in custodial natural unaccompanied a field, scalar by point-like, fundamental, a of ietcneuneo h tutr fteWy scale Weyl the of breaking. is structure mechanism this call the This We of [7]. current. ref consequence to direct not prior a seems codified it to been have taken scale, to approaches Planck many the generating the spontaneously in implicit is mechanism ‡ † ∗ lcrncades [email protected] address: Electronic [email protected] address: Electronic lcrncades [email protected] address: Electronic nti ae efcso h elkonWy current Weyl known well the on focus we paper this In appearance the with boson Higgs the of discovery The yieta pnaeu cl ymtybekn,we breaking, symmetry scale spontaneous inertial By oeta.Ti sdcae ytesrcueo h elcur Weyl the of structure the symmet by spontaneous dictated “inertial is of This process potential. dynamical a a by ated evdWy urn.W ontrqiesaeivratregul invariant scale potential. acti Coleman-Weinberg require the invariant not in Weyl fields do a We of of VEV’s current. the Weyl to served is back theory referred quantum are renormalized conditions Einstein-Hilbert the in the invariance and Weyl expands exact universe the which during elivrattere fsaasadgaiycngenerate can gravity and scalars of theories invariant Weyl osntepo potential a employ not does .INTRODUCTION I. nrilsotnossaesymmetry scale spontaneous inertial er .Ferreira, G. Pedro nrilSotnosSmer Breaking Symmetry Spontaneous Inertial 3 uofPirsCnr o hoeia Physics, Theoretical for Centre Peierls Rudolf ..Bx50 aai,Ilni 01,USA 60510, Illinois Batavia, 500, Box P.O. n unu cl Invariance Scale Quantum and 2 1 em ainlAclrtrLaboratory Accelerator National Fermi nvriyo xod el Road Keble 1 Oxford, of University 1, srpyis eateto Physics of Department Astrophysics, nvriyo xod el Road Keble Oxford, of University ∗ hl this While . hitpe .Hill, T. Christopher Dtd oebr1 2018) 1, November (Dated: xodO13RH OX1 Oxford xodO13NP OX1 Oxford tosaesrthdott eoeacntn au of value constant a become fluctu- to scale out initial stretched distance are short ations all Essentially stant. ee”oe ntal,w xetlclflcutosin fluctuations local nonzero expect e.g., “disor- we scale, an break Initially, that from fields state one. the “ordered” dered” represents an (SSB) between where difference breaking parlance, matter symmetry condensed spontaneous the follow presently dniytecnevdWy current Weyl conserved the identify breaking. must symmetry parameter all order The and any system groundstate. capture spin the to the cools in it magnetization as constant and the space that like in parameter” constant time, “order macroscopically an or- become need an to we evolves represent that not For do state. they dered break but macro- do locally, they a symmetry and in temperature scale high spins at magnetic system spin local scopic to analogous are These htsaecag est ileov ozero, to evolve such will universe density expanding charge an scale in that evolve dynamically will ieo clrquantity, deriva- scalar a a always of is tive current this theory invariant Weyl any current, nel.” hsi utcvrac,lk h iuino conserved general during a field of magnetic dilution a expansion. or the density like charge covariance, electric just is This 1 hsi hoe,adi s’ adt e hti od in holds it that see to hard isn’t it and theorem, a is This elgaiygnrlztos u h ro sbyn h sc the beyond discussion. is present proof our the of but generalizations, gravity Weyl ntepeetppr okn na“odnfae we frame” “Jordan a in working paper, present the In tte olw httekernel, the that follows then It 1 2, † wn oti tutr ftecnevdWeyl conserved the of structure this to Owing ffcieato sfre.Maintaining formed. is action effective D no h hoy hc mle con- a implies which theory, the of on tagtowr hnrenormalization when straightforward n rhmG Ross G. Graham and tr.W lutaetecomputation the illustrate We ators. rent, l assae pnaeul,initi- spontaneously, scales mass all µ ybekn”ta osntinvolve not does that breaking” ry K µ K ,w r urnedta h system the that guaranteed are we 0, = µ n omlgclphase cosmological a and , K µ = FERMILAB-PUB–18–008T ∂ 3, µ ‡ K where φ K K i µ ( x efidta in that find We . or ) → K OUTP-17-17P ∂φ K ste“ker- the is scon- is , K i ( 0 x R ,etc. ), → 2 and ope 0. 2

K = K, and the scale symmetry is broken. K is the define the final theory and its symmetries. Though it order parameter of the SSB since it intimately connects may be convenient, one need not deploy a regulator that with the dynamics. At late times we have, for N-fields, is consistent with the symmetries of the renormalised the- K( φ )= K( φ ) exp(σ/f) . Here σ is the dilaton and ory. The nonexistence of a symmetry in the regulator { i} { i} K( φi ) is a constraint that reduces the N fields φi to does not imply the nonexistence of the symmetry in the N { 1} fields constrained to a locus in field space, such{ } as renormalised theory. Furthermore, physics should not the− ellipse of [2]. depend upon the choice of regulator [9]. We clearly see that K is the order parameter because In this view, Weyl symmetry is central and all mass the decay constant of the dilaton is precisely f = √2K, scales must emerge by way of random initial conditions in analogy to fπ in a chiral Lagrangian, or the VEV of the governing VEV’s (Vacuum Expectation Values) of fields Higgs in the standard model. The constraint of constant that are entirely contained within the action. Essentially K( φi ) gurantees that the dilaton fluctuation is orthog- there exist no fundamental mass scales, and the mass of onal{ in} the kinetic terms to the other N 1 constrained anything is defined only relative to field VEV’s in the fields and neatly factorizes. − theory. For this to be phenomenologically acceptable it We emphasize that this is a dynamical process. Just is necessary to explain how the spontaneous breaking of as steam can condense into water, a scale disordered Weyl symmetry can lead to a period of inflation followed phase can condense into a scale ordered one. All of this by a reheat phase and transition in the infrared to a is tracked in a single frame, which begins as a Jordan theory describing the fundamental states of matter and frame. In this view the universe is a physical system their interactions with an hierarchically large difference that starts in one phase, which has no scale ordering, between the Planck scale and the electroweak breaking and ends up in another in which the scale SSB defined scale. by K. This is treated in one set of “frame variables” Remarkably it has been shown in a simplified model with an Friedman-Roberston-Walker metric. In a sense involving two scalar fields that this structure is possible the approach of K K is just the relaxation of the dila- [2, 3, 7]. The model has a scale invariant scalar potential ton σ 0, though→ the dilaton can only be defined in the and non-minimal coupling of the scalar fields to the Ricci broken→ phase of the theory. scalar. When the fields develop VEVs the Planck scale Here we need not do the Weyl transformation along is generated spontaneously in the Brans-Dicke manner. the way and the SSB materializes dynamically. How- For a wide range of the non-minimal couplings and scalar ever, at late times the dynamically generated K can then interactions, there is an initial period of “slow-roll” in- be matched to the scale quantities, MP , Λ, etc., in an flation that can give acceptable values for the slow-roll Einstein-Hilbert action. Once these scales are identified, parameters. This is followed by a “reheat” phase and a then it is useful to make a Weyl transformation, e.g., flow of the field VEVs to an infrared fixed point at which using K, to isolate the dilaton. This guarantees that the ratio of the scalar field VEVs are determined by the the dilaton factorizes and alleviates any putative messy dimensionless couplings of the theory. Thus it is possible kinetic term mixing issues. (in fact, this permits the to arrange an hierarchically large ratio for the VEVs and, dilaton to be “eaten” by a Higgs mechanism if we in- interpreting the second scalar as modelling the Standard troduce Weyl’s photon, Aˆµ as in Section II.D, allowing Model Higgs boson, this large ratio corresponds to the the Weyl photon and dilaton to decouple as very heavy ratio between the Planck scale and the electroweak scale. states.) There does then remain a mixing issue amongst In section II we discuss the mechanism of inertial spon- the remaining N 1 constrained fields, and these must taneous symmetry breaking and conservation of the Weyl be diagonalized to− apply the low energy dynamics. current in a toy model, and general N-scalar models. As The advantage of phrasing things in terms of conserved it does not involve a potential the mechanism opens a currents is that the results are model independent. We new pathway to generating spontaneous scale symme- never have to actually construct and solve difficult non- try breaking and the associated spontaneous breaking of linear partial differential equations of motion to see this; other symmetries. As such it may be useful for novel as- this will happen automatically, and the resulting mass pects of model building. We also discuss a general feature scales, including the Planck mass, are generated sponta- of this mechanism, the origin of the dilaton and its inti- neously, controlled by the Weyl current. This mechanism mate relationship to the current, We also briefly consider, does not depend upon a potential, (though the particu- as an aside, locally Weyl invariant models in which the lar final vacuum state is dictated by a potential). The dilaton will be eaten by a “Weyl photon,” Aˆµ, to give it statements we make are general and model independent, mass, i.e., the inertial symmetry breaking thus becomes similar to those of any traditional “current algebra.” a Weyl symmetry Higgs mechanism, and the dilaton dis- A crucial aspect of this mechanism is that quantum appears from low energy physics [10]. theory should not break scale symmetry. We believe this In section III we discuss how Weyl invariance is main- is generally possible. To understand this, it is important tained at the quantum level and thus preserves the in- that one does not conflate the procedure of regularisa- ertial spontaneous symmetry breaking mechanism. As tion, which generally introduces arbitrary mass scales, a result the logarithmic corrections that normally break with renormalisation, which introduces counter-terms to the scale invariance now automatically depend only on 3 physically relevant ratios of field VEVs which preserve At first this doesn’t look too hard. Indeed, if φ starts the underlying Weyl invariance of the theory. We com- out in some very high-temperature phase, where the en- pare this procedure to previous proposals for scale in- ergy density is large compared to λφ4 then we expect the variant regularisation that require an arbitrary choice of scale factor will increase in a scale invariant way, a(t) t. regulator, a function of the scalar fields. This follows by intuiting that the Hubble constant satis-∼ Finally, in section IV, we present a summary of our fies H2 T 4/φ2, where the φ2 factor in the denominator ∼ 2 2 2 results and the conclusions to be drawn. replaces MP . In thermal equilibrium we expect φ T a˙ 1 ∼ and thus H = a T t . Therefore, a(t) t. As the universe∼ cools,∼ we expect φ(x)∼ to settle into II. INERTIAL SPONTANEOUS SYMMETRY some spatially constant VEV φ . However, our intuition BREAKING. h2 i from conventional Einstein MP R gravity tells us that this VEV will slow-roll in the potential, with φ eventually A. A Toy Example becoming zero. In eq.(3) this would then implyh i a vanishing MP , and the details of the solution become less clear. Consider a real scalar field theory action together with It is plausible that the increasing strength of gravity will Einstein gravity and a cosmological constant (our metric increase the Hubble damping, and halt the relaxation of signature convention is (1, 1, 1, 1)): φ , perhaps leading to a nonzero cosmological constant h i − − − λ φ 4. If true, this would then match the cosmological h i 1 µν 1 2 constant case of eq.(1), and it would imply a sponta- S = √ g g ∂µσ∂ν σ Λ+ M R . (1) − 2 − 2 P neous breaking of scale symmetry. We could resort to a Z numerical solution, but how can we see what happens in This action provides a caricature of the cosmological a simple and intuitive way, without having to puzzle over world we live in. the solutions of coupled nonlinear differential equations? We imagine an initial, ultra-high-temperature phase in Indeed, from eq.(3) we can directly obtain the Einstein which the massless scalar σ has the dominant energy den- equation: sity, ρ T 4. Consider a Friedman-Robertson-Walker σ ∝ (FRW) metric: 1 3 α 3 2α αφ2G = − ∂ φ∂ φ g − ∂µφ∂ φ 6 αβ 3 α β − αβ 6 µ 2 2 2 a˙ gµν = [1, a (t), a (t), a (t)] H = . (2) 1 − − − a + α g φD2φ φD D φ + g V (φ). (4) 3 αβ − β α αβ In this theory the universe initially expands in a FRW phase, with the temperature red-shifting as T 1/a(t), The trace of the Einstein equation becomes: and the scale factor growing as a(t) √t. Eventually∼ ∼ 1 2 µ 2 the σ thermal energy becomes smaller than the cosmo- αφ R = (α 1)∂ φ∂µφ + αφD φ +4V (φ). (5) −6 − logical constant, ρσ < Λ, and we then enter a deSitter 2 phase with exponential growth, a(t) et√Λ/3MP . We We also have the Klein-Gordon (KG) equation for φ: ∼ can model the thermal phase as a pre-inflationary era, δ 1 and the cosmological constant then represents a poten- 0= φD2φ + φ V (φ)+ αφ2R. (6) δφ 6 tial energy that drives inflation. In any case, the intuition that allows us to readily understand how this works We can combine the KG equation, eq.(6), and trace equa- is well-honed. tion, eq.(5), to eliminate the αφ2R term, and obtain: Now consider a different action: 2 µ 0 = (1 α)φD φ + (1 α)∂ φ∂µφ 1 λ α − − S = √ g gµν ∂ φ∂ φ φ4 φ2R . (3) δ − 2 µ ν − 4 − 12 +φ V (φ) 4V (φ) . (7) Z δφ − This action is scale invariant, having no cosmological con- This can be written as a current divergence equation: stant or Planck scale. These two theories are classically equivalent, provided ∂ DµK =4V (φ) φ V (φ) . (8) α < 1. This equivalence follows from a Weyl transfor- µ − ∂φ mation. But, our question then becomes, given that, from our accumulated experience in inflationary cosmol- where ogy we understand the dynamics of eq.(1) so well, then K = (1 α)φ∂ φ (9) how could we directly understand the dynamics of the µ − µ Weyl equivalent eq.(3) without performing a Weyl trans- is the “Weyl current.” For the scale invariant potential, formation into eq.(1)? What happens in the pure evo- 4 V (φ) φ , the rhs of eq.(8) vanishes and the Kµ current lutionary dynamics intrinsic to eq.(3) that produces the is then∝ covariantly conserved: physical mass scales of MP and Λ, as well as all other µ scales in nature? D Kµ =0. (10) 4

N 2 We see that this is an “on-shell” conservation law, i.e., In particular, with F ( φj )= i αiφi the kernel takes it assumes that the gravity satisfies eq.(4). This is the the form [7]: { } global Weyl current and it can be derived by a Noether P N variation of the action under a Weyl transformation. 1 2 K = (1 αi)φi . (16) Note that the Weyl current, Kµ, is the derivative of a 2 − i=1 scalar, Kµ = ∂µK, where: X In this case the N scalar fields will evolve such that their 1 K = (1 α)φ2. (11) values will ultimately be constrained to lie on the N- 2 − dimensional locus by eq.(15) with K K, in particular → We refer to K as the “kernel.” Using the conserved K- an ellipsoid in the special case of eq.(16). current with its kernel, we can easily understand the dy- Here we are “launching” the theory in an effective Jor- namics of this theory. dan frame, with arbitrary initial values of the fields and The form of the conservation law is DµK = D2K = 0, their time derivatives φ , φ˙ . The initial expansion will µ { i j } and this holds in any frame. If we take φ to be func- be scale invariant, a(t) t, but as K K, the Planck tions of time t only, and consider a Friedman-Robertson- scale becomes dynamically∼ established,→ and we enter an 2 2 2 Walker universe (gµν = [1, a (t), a (t), a (t)]) the effective Einstein frame where all mass scales are √K, − − − ∝ current conservation equation implies: and the expansion becomes deSitter, a(t) exp √Kt). ∼ a˙ In a 2-scalar model discussed in ref.[7] we have checked K¨ +3 K˙ =0. (12) numerically that the initial rate of approach to the ellip- a soid is rapid and thereafter the fields precisely track the This can be readily solved to give: ellipsoid corresponding to constant K. This is true for a

t wide range of initial conditions and readily allows for an dt′ inflationary period to commence. Since K has dimension K(t)= c1 + c2 3 , (13) 2 0 a (t ) of (mass) , a constant vacuum value of K implies a spon- Zt ′ taneous breaking of the scale symmetry in the theory has where c are constants. Therefore in an expanding uni- 1,2 occurred. Note that this phase does not employ a poten- verse K will evolve to a constant value, K K. → tial but is driven solely by the initial conditions, and K In the single scalar case, as K K constant, the initial is the order parameter of inertial spontaneous symmetry Jordan frame theory flows to an→ effective final Einstein- 2 breaking. λK 2 Hilbert theory with parameters Λ = (1 α)2 ,MP = In multi–scalar theories the flow K K does not fix − → αK , f 2 = 2K (dilaton decay constant, see II.B) the relative values of the scalar field VEVs, which initially − 3(1 α) end up at some random point on the locus (e.g. ellipse). [7]. The− equivalence between the theories is achieved dy- It is here that the potential becomes important. In the namically, without having performed a Weyl transforma- infrared (IR), the fields constrained to the locus, flow tion, and it follows from the Weyl current algebra, and towards an IR fixed point in which the ratios of the field does not rely upon the solutions of complicated nonlinear VEVs are determined by the potential terms alone [7]. differential equations of motion. For the case that the potential has a flat direction, the This is robust. If we consider a set of N scalars, φ , j vacuum energy vanishes at the minimum, corresponding with action given by:2 { } to vanishing cosmological constant. The IR fixed point is N then the intersection of the potential’s flat direction with 1 S = d4x√ g ∂ φ ∂µφ W ( φ ) the locus. The ratios of the VEV’s is then determined − 2 µ i i − { j } " i by the scalar potential couplings, but constrained by the Z X 1 requirement the fields lie on the N-dimensonal ellipsoid. F ( φ )R . (14) −12 { j } For the case that the potential is positive definite, the IR fixed point corresponds to an eternally inflating de- where we maintain scale invariance (i.e., F ( φk ) and Sitter solution in which the ratio of the field VEV’s is de- 2ǫ { } W ( φk ) transform respectively as F e F and W termined by the scalar potential couplings together with 4ǫ { } → → e W under global Weyl transformations, as defined be- the couplings, αi, of the scalars to the Ricci scalar. low in eq.(20)). The conserved Noether current kernel then generalises to: B. General Discussion N 1 K = φ2 F ( φ ) . (15) 2 i − { k} Inertial spontaneous symmetry breaking can be re- " i=1 ! # X sponsible for triggering the spontaneous breaking of symmetry in all sectors of the theory. As such it opens new possibilities for model building. 2 It is straightforward to extend this effective Lagrangian to matter In summary, we found that the expansion of the uni- and gauge fields [2, 10, 13]. verse in a pre-inflationary phase drives the current charge 5 density, K , to zero. The global Weyl current, K , is al- space orthogonally to the N 1 φ′ fields. The dilaton 0 µ − { i} ways the derivative of a scalar, Kµ = ∂µK, and in partic- decouples in the action from everything except gravity ular K0 = ∂0K, where K is the kernel. Hence, as the Kµ (this holds true for fermions and gauge bosons as well; current density is diluted away, K0 0, the kernel K decoupling implies that there are no direct couplings in therefore evolves as K K constant.→ In a Weyl invari- the action to other fields). ant theory this implies→ that scale symmetry is broken, This result is elegantly simple. There is no messy ki- and the Planck mass is generated dynamically. netic term mixing problem of the dilaton with the re- K plays the role of the symmetry breaking order pa- maining φ′ fields, as some authors have alluded to. In- rameter. While a potential may then be needed to en- deed, there is nontrivial mixing amongst the φ′ that are gineer the final vacuum, and determine the ratios of in- subject to the constraint, but the dilaton is neatly factor- dividual fields φi , it plays no direct role in the inertial ized and does not mix with these other fields kinetically. Weyl symmetryh breakingi phenomenon. We further see that the current written in the uncon- With a little thought, one might have guessed the strained fields is equivalent to one written in the con- 2σ/f structure of the order parameter K. Consider a set of N strained fields by: Kµ = ∂µK(φ) = ∂µ(K(φ′)e ). scalar fields φi . If the fields are non-minimally coupled Hence in the broken phase (Einstein frame) limit { } 2 to gravity as ( 1/12) αiφ R(g), then if any of the φi K(φ′) K constant, K 2K∂ σ/f = f∂ σ. where − i i → µ → µ µ should develop a VEV, we would expect scale breaking, f = √2K. This is as we expect for a Nambu-Goldstone P and a nonzero K. Hence, we expect that the order pa- boson, e.g., the axial current of the pion takes the anal- rameter takes the form, K c φ2. However, if any φ ∼ i i i ogous form fπ∂µπ. This implies that K is the order pa- has αi = 1, then we can remove it from the action by a rameter of Weyl spontaneous symmetry breaking. P local Weyl transformation, absorbing it into the metric. Why is this formulation important? Results follow- 2 We therefore expect K = c′ (1 α )φ . Indeed, we i − i i ing from the “current algebra” of Weyl invariant theo- found that Kµ = ∂µK, with c′ = 1/2, combining both P ries are general statements that are true, independent of the trace of the Einstein and KG equations, or by the the specific structure of the Lagrangian. The particular Noether variation of the Jordan frame theory under a structure of Kµ and K is independent of the form of any Weyl transformation, thus confirming our guess. scale invariant potential, but the detailed structure of K does depend upon the choice of the non-minimal couplings, e.g., F (φi) in eq.(15) (and also any higher deriva- C. Factorization of the Dilaton tive gravitational terms can modify the simple forms we just discussed). The behavior of the current algebra will We’ve seen the result that K K constant as the remain intact, since K = ∂ K is conserved, but the con- → µ µ universe expands implies that N 1 fields φ′ will ul- − { i} straint defined by K could become a more general locus timately satisfy a constraint, such as in eq.(16), K = such as a hyperbola, etc., (such effects result from the 2 (1/2) (1 α )φ′ . Here the constrained fields, φ′ , i − i i { i} renormalization group [7]). The survival of the general lie on an ellipsoid in field space, but the constraint could feature of inertial breaking with a stable goundstate, e.g., be moreP general as in eq.(15) with F ( φ ), and the el- { i} a stable MP lanck, requires that the quantum theory does lipsoid could be a more general locus in field space. not break Weyl symmetry through loops, as we discuss In any case, there remains one field unconstrained that in Section III. becomes the dilaton. This is intimately related to the Kµ current. Let us perform a Weyl field redefinition on the N original fields, D. Local vs. Global Weyl invariance; Eating the Dilaton φ = exp(σ/f)φ′ g = exp( 2σ/f)g′ . (17) i i µν − µν We thus find the Weyl invariant action becomes: Our main discussion is based upon globally Weyl invariant theories. However, we include the present section S(φ, g) = S(φ′,g′) to indicate how it may be possible to promote these to locally Weyl invariant theories by introduction of Weyl’s µ 2 + g′ ∂µK(φ′)∂ (σ/f)+ K(φ′)(∂σ/f) (18) gauge field, i.e., “Weyl’s photon.” It is interesting that − Z inertial symmetry breaking now becomes a Higgs mecha- p Now using the constraint that K constant, and integrat- nism, since the Weyl photon will “eat” the massless dila- ing by parts, we have: ton and thus remove it from the low energy spectrum, where it becomes the longtudinal degree of freedom of a 1 2 massive Weyl photon. Hence, in this case the issue of long S(φ, g)= S(φ′ ,g′)+ g (∂σ) (19) i 2 − ′ range 5th force limits becomes moot. The present sec- Z p tion is classical, but it would be of interest to develop the Here we identify f 2 = 2K so the dilaton is canonically full quantum (renormalization group) behavior of Weyl’s normalized. From this we see that the dilaton, σ, de- photon. scribes a dilation of the ellipse, and fluctuates in field Weyl’s original idea was that, since coordinates are 6 merely numbers invented by humans to account for absence of a factor of i in the coefficient of A˜µ: QED events in space-time, they should not carry length scale gauges phase, while the Weyl photon gauges scale). [1]. Rather, the concept of length should be relegated Armed with this we can construct another local Weyl to the (covariant) metric, and (contravariant) coordinate invariant: differentials are scale free. Therefore, under a local Weyl µν ˜ ˜ scale transformation we would have: √ gg Dµφ(x)Dν φ(x). (24) − 2ǫ(x) µν 2ǫ(x) µν 3 gµν (x) e− gµν (x) g (x) e g (x) This is a locally Weyl invariant kinetic term. → → 4ǫ(x) ǫ(x) We can combine this with the previous invariants to √ g e− √ g φ(x) e φ(x) (20) − → − → define an action in which the Weyl symmetry is local, Weyl transformations are distinct from coordinate dif- yet the nonminimal coupling of scalars to R is arbitrary: feomorphisms that define scale transformations on coordinates, as δxµ = ǫ(x)xµ, which we discuss below. 1 λ S = d4x √ g (1 α)gµν D˜ φD˜ φ φ4 The global Weyl symmetry corresponds as usual to − 2 − µ ν − 4 ǫ =(constant in spacetime). Z α φ4R(φ2g) It is straightforward to construct a list of local Weyl −12 invariants: 4 1 µν α 2 λ 4 = d x √ g g ∂µφ∂ν φ φ R(g) φ 2 2 µν 4 φ (x)g (x); φ− (x)g (x); √ g(x)φ (x); − 2 − 12 − 4 µν − Z 2 2 3 µ 1 R(φ g ) = φ− R(g )+6φ− D (∂ φ) µ 2 µ 2 µν µν µ (1 α) A˜ ∂µ(φ ) A˜ A˜µφ . (25) 4 2 2 µ −2 − − √ gφ R(φ gµν ) = √ g φ R(gµν )+6φD (∂µφ) − − ... (21) Now, we want to pass to the Weyl broken phase. We 2 write: Note that the computation of R(φ gµν ) above requires that any Christoffel symbols used in the definition of R φ(x) f exp(σ(x)/f) be evaluated in the metric φ2g . Using these identities → µν g (x) exp( 2σ(x)/f)g (x) (26) we can construct an action that is locally Weyl invariant: µν → − µν 1 λ Note we do not at this stage do a gauge tranformation, S = d4x √ g φ4R(φ2g) φ4 (22) − −12 − 4 Aµ Aµ + ∂µσ(x)/f. We obtain, Z → 4 1 µ 1 2 λ 4 1 = d x √ g ∂µφ∂ φ φ R(g) φ S = √ g (1 α) gµυ∂ σ∂ σ (1 α)gµυA f∂ σ − 2 − 12 − 4 − − 2 µ ν − − µ ν Z Z where we substituted the relationship of eq.(21) and in- 1 µυ 2 1 2 µ + (1 α)g AµAν f αf R (27) tegrated by parts using the divergence rule DµV = 2 − − 12 1 µ √ g− ∂µ(√ gV ). Here we obtain the famous locally Weyl− invariant− theory in which the nonminimal coupling Note that the Weyl transformation cancelled the original 1 µυ of scalars to gravity is fixed by the coefficient 1/12, 2 αg ∂µφ∂ν φ piece since it was local. What is left is a needed to canonically normalize the φ kinetic term. This perfect square; is a special and somewhat degenerate theory, since we 2 1 2 µυ 2 1 2 can revert to the metricg ˆµν = φ gµν and φ disappears S = √ g f (1 α)g (A ∂ σ/f) αf R − 2 − µ − µ − 12 from the action. The theory has a vanishing Weyl current Z [11]. 1 1 = √ g f 2(1 α)gµυB B αf 2R (28) We note that covariant gauge fields, such as the elec- − 2 − µ ν − 12 Z tromagnetic vector potential, Aµ, do not transform under the local Weyl transformation, since they are associated where we redefine B = A ∂ σ/f which is a massive µ µ − µ with derivatives ∂µ ieAµ which, like coordinates, do not spin one field of mass m = f (1 α). The dilaton has transform. The electromagnetic− fields that have the usual − engineering scale (mass)2, E~ and B~ , are contained in p the field strength with∼ one covariant and one contravari- ν 0 3 ant index, F , e.g., E~i = F . Here there is a subtlety, as we must define the derivative of µ i − We can construct a covariant derivative of a scalar any conformal field as a commutator: [Dµ, Φ] = ∂µΦ Aµ[W, Φ] where [W, φ] = wφ and w is the conformal charge of Φ. Hence field under local Weyl transformations by introducing the µν w = 1 for φ. We also require w = −2 for gµν , w = +2 for g , “Weyl photon,” A˜ , as µ w = −4 for det −g, etc. Note that [Dµ,gρσ]= Dµgρσ + 2A˜µgρσ = A˜ g since D g = 0. This insures the invariance of the D˜ φ = ∂ φ A˜ φ (23) µ ρσ µ ρσ µ µ − µ action with the Weyl covariant derivative under integration by parts. Note that we can alternatively define a restricted “pure ǫ(x) where φ(x) e φ(x) and A˜µ(x) A˜µ(x)+ ∂µǫ(x) A ∂ χ χ → → gauge theory” with µ = µ ln( ), where is any massless scalar (note the major difference from electrodynamics in the field. 7 been eaten by the Weyl photon to become its longitudi- by diffeomorphisms are identical to those in K for the nal mode, and the massless dilaton has thus disappeared scalar sector of the theory. from the spectrum of the theory. Scale and Weyl symmetry of a theory appears ab ini- We can always have a kinetic term for Aµ with tio to be broken by quantum loops. Loop divergences are subtle, however, and are often confused with physics. F = [D ,D ]= ∂ A ∂ A = ∂ B ∂ B (29) µν − µ ν µ ν − ν µ µ ν − ν µ Here we adopt an operating principle that has been es- and poused by W. Bardeen [9]: The allowed symmetries of a renormalised quantum field heory are determined by 1 1 1 anomalies, (or absence thereof). Quantum loop diver- S = √ g F F µν + m2gµυB B αf 2R − −4 µν 2 µ ν − 12 gences are essentially unphysical artefacts of the method Z (30) of calculation. The equation of motion for Bµ is Weyl or scale symmetry is permitted if the renormalised theory has no Weyl anomalies. Since trace µν 2 ν ν µ 2 ν ∂µF = D B ∂ (DµB )= m B (31) anomalies come from triangle diagrams they are necessar- − ily associated with dimension-4 operators. Hence there This is mathematically analogous to a superconductor or is no Weyl anomaly in the Standard Model of the form the Standard model Higgs mechanism. A gas of Bµ will 2 H†H where the Higgs mass is m H†H. Thus there are freeze out and redshift away like matter once the temper- no Weyl anomalies associated with quadratic or quartic ature redshifts below m. It is also interesting to note that divergences in quantum field theory in four dimensions. if we have N φ fields, the inertial symmetry breaking will i Another way of saying this is that divergent terms and yield the N 1 φi′ fields and the dilaton which is again counter terms are not separately measurable, only the eaten to become− the longitudinal component of Bµ, but µ renormalised mass is physical. In a variation of the Stan- we then find that the gauge field B decouples from the dard Model with no gravity, no grand unification and no φ ! It also has even charge conjugation and presumably i′ Landau poles in the far UV the Higgs mass would be decouples from fermions and gauge fields as well, and it technically natural with no hierarchy problem! cannot decay to a pair of gravitons (this is a variation on Yang’s theorem which forbids decay of a vector meson to µ a photon pair). Therefore, relic B fields are stable and A. The origin of Weyl anomalies could constitute a dark matter candidate if they are not inflated away. Our problem of maintaining Weyl symmetry requires From the action of eq.(25) we see that the Weyl current that we build a theory that has no anomaly in K . To is easily obtained: µ understand this problem, and its solution, we turn to the

1 δS 2 CW potential. In computing CW potentials for mass- Kµ = = (1 α) φ∂µφ A˜µφ −√ g δA˜µ − − less scalar fields we encounter an infrared divergence that − must be regularised [8, 14]. To do so we often introduce = (1 α)φD˜ φ. (32) − µ explicit “external” mass scales into the theory by hand. These are mass scales that are not part of the defining This still has the general form Kµ = DµK, where Dµ is a covariant Weyl derivative. action of the theory, and essentially define the RG tra- jectories of coupling constants. These externally injected By setting A˜µ = 0 we obtain a globally invariant theory, and this current becomes the conserved Noether cur- mass scales lead directly to the Weyl anomaly. rent for the global Weyl invariant theory: We can see this in eq.(3.7) of CW [8] where, to renormalise the quartic scalar coupling constant, λ, in an ef- Kµ = (1 α)φ∂µφ. (33) fective potential at one loop level, W (φ), they introduce − a mass scale M. Once one injects M into the theory, one has broken scale and Weyl symmetry, and the effective III. QUANTUM SCALE INVARIANCE AND φ potential in the large M limit then takes the form REGULARISATION β φ W (φ)= 1 φ4 ln (35) 4! M Up to now our discussion has been confined to the classical action. For the scenario of inertial spontaneously Here β1 is the one-loop renormalisation group coefficient, broken scale symmetry to work, and lead to a stable dλ(µ)/dµ = β1. The manifestation of this is seen in Planck mass, it is essential the that Weyl current be iden- the trace of the improved stress tensor [13], and in the tically conserved at the quantum level [5] : divergence of the Kµ current: µ D Kµ =0. (34) δ β ∂µK =4W (φ) φ W (φ)= 1 φ4 (36) µ δφ 4! In what follows we will refer to nonzero contributions − − coming from loops to the rhs of eq.(34) as “Weyl anoma- Of course, there is nothing wrong with the CW poten- lies.” The trace anomalies of the scale current determined tial, or with this procedure, if one is only treating the 8 effective potential as a subsector of the larger theory. If, neglecting graviton loops) but we emphasise that the pro- however, Weyl symmetry is to be maintained as an exact cedure immediately generalises to the case with arbitrary invariance of the world, then M must be replaced by an N and indeed to the inclusion of fundamental fermions internal mass scale that is part of action, i.e. M must and vectors. then be the VEV of a field, χ, or some combination of We start with the action given in eq.(14) with N = 2. the fields, appearing in the extended action. We would The Weyl invariance of the theory is spontaneously bro- then have the Coleman-Weinberg potential: ken by the VEVs of the fields giving a massless Goldstone boson, the dilaton, σ. It was shown in [10] that the dila- β φ W (φ, χ)= 1 φ4 ln (37) ton decouples and so, of the two initial scalar degrees of 4! χ freedom, only one interacting one remains. To see how this happens in practice, we change variables to: and, because we now have no external mass scales, the current divergence vanishes: σ/f φi = e− φî 2σ/f δW (φ, χ) δW (φ, χ) gµν = e gˆµν (39) ∂µK = 4W (φ, χ) φ χ =0. µ − δφ − δχ where φˆ are constrained to lie on the ellipse given by: (38) i N This defines the basic idea for maintaining scale symme- 2K = (1 α )φˆ2 = f 2 (40) try in the quantum theory. It simply implements the − i i i=1 notion that there are no fundamental mass scales, and X masses are determined only as dimensionless ratios in- where f 2 is a constant. It is important to note that f volving VEV’s of scalar fields. In the next section we is invariant under scale transformations as the dilaton illustrate this through a calculation of the one loop cor- dependence of the original fields has been factored out. rection to the scalar potential arising from the quartic To illustrate the regularisation procedure it is sufficient scalar interaction. Of course there will be further gravi- λ 4 to calculate the CW potential resulting from the 4! φ1 tational corrections but their calculation lies beyond the term in the potential. We first re-parameterise the fields scope of this paper. by:

f f φˆ1 = sin θ, φˆ2 = cos θ (41) B. Weyl Invariant Coleman-Weinberg Calculation √1 α √1 α − 1 − 2 How might we derive such a result as in eq.(37) from After scaling out the dilaton, the relevant terms of eq.(14) first principles? We do so via a computation of a become: Coleman-Weinberg (CW) effective potential. It is impor- 1 cos2 θ sin2 θ tant to realise that CW effective potentials themselves S = d4x gˆ f 2 + ∂ θ∂µθ must have the full symmetry of the underlying theory. − 2 (1 α ) (1 α ) µ Z − 1 − 2 The symmetry is then broken spontaneously by the min- p λ sin4 θ imum of the potential. f 4 (42) − 4 (1 α )2 In fact it is straightforward to show that the usual − 1 regularisation procedure applied to the Weyl invariant Performing the further redefinition Θ = F (θ) where: theory of eq.(14) does have a Weyl invariant form. For the simple two scalar case, N = 2, with fields φ = φ1 θ 2 2 and χ = φ2, it reduces to that of eq.(37) when the ratio cos θ′ sin θ′ F (θ)= + dθ′ (43) of VEV’s is small, but the general form is applicable for 0 s(1 α1) (1 α2) arbitrary values of the ratio. Z − − the action becomes:

1 λ sin4 F 1(Θ) 1. The two scalar action S = d4x gˆ f 2∂ Θ∂µΘ f 4 − . − 2 µ − 4! (1 α )2 Z − 1 The case, N = 2, is the simplest model with “realistic” p (44) phenomenological properties. For reasonable parameter For the case θ is small the action approximates to the choices and initial conditions it can have an initial infla- simpler form: tionary period followed by a “reheat” phase and subse- 1 λ quent evolution to an IR stable fixed point in which the S d4x gˆ ∂ Φ∂µΦ Φ4 . (45) ≈ − 2 µ − 4! ratio of the field VEVs is determined by the fudamental Z couplings of the theory. We will illustrate the regular- p where Φ = fΘ and Θ θ . isation procedure applied to this model (in the limit of √1 α1 ≈ − 9

2. The CW potential In terms of the original fields Φ = fΘ,Θ θ and √1 α1 ≈ − θ φˆ1/φˆ2, the potential is given by: Here we demonstrate the derivation of the Weyl invari- ≈ φ1 ant CW potential for the case 1, starting with the 2 ˆ4 ˆ2 φ2 ≪ λ ˆ4 λ φ1 Cφ1c 25 action of eq.(45). Adding a classical source term, JΦ, W φ1 + 2 ln (53) ≈ 4! 256π φˆ2 − 6 to the Lagrangian induces a shift in the Φ field: − 2c ! ! 2 f 1 ~1/2 ˆ 2 Φ=Φc + Φ (46) where C = M 1 α2 is a constant invariant under scale changes. This is the− Weyl invariant CW potential written where Φˆ is the small fluctuation about the classical min- in terms of the variables (φˆ1, φˆ2) which are constrained imum. Thus the potential has the form: by eq.(41). In addition there is a dilaton, σ, with an iso- lated kinetic term. By performing a Weyl transformation λ λ W (Φ) = Φ4 + ~ Φ2Φ2 + ... (47) that is the inverse of eq.(39), we can relax the constraint 4! c 4 c eq.(41) and obtain, where the linear term cancels due tob the classical source λ λ2φ4 Cφ2 25 term. Treating the quadratic term in Φˆ as an interaction W φ4 + 1 ln 1c (54) ≈ 4! 1 256π2 φ2 − 6 the 1-loop potential with Φˆ the propagating field is given 2c by: which is Weyl invariant, and the the fields (φ1, φ2) = 1 n exp( σ/f)(φˆ , φˆ ) are independent variables. d4k ∞ 1 λΦ2 1 2 W = Ω+ i 2 c The− reason Weyl invariance has been preserved is be- eff 4 2n k2 + iε (2π) n=1 cause the inertial spontaneous symmetry breaking has Z X 4 2 introduced the mass scale, f, that compensates for the 1 d k λΦc = Ω+ ln 1+ appearance of the renormalisation scale M under the log, 2 (2π)4 2k2 Z leaving the logarithmic terms invariant. Note that the 2 2 4 1 2 2 λΛ λ Φ λΦc +Λ usual renormalisation group equations still apply as a = Ω+ Φ2 c ln 2 2 c 2 1 2 change in the renormalisation scale M (a change in C in 128π − 256π 2 λΦc eq.(53)) is compensated by a change in the couplings and Λ4 1 λΦ2 +Λ2 + ln 2 c (48) wave function factors in the usual way. 64π2 Λ2 where: 3. Scale invariant regularisation λ 1 λ Ω= Φ4 B Φ2 C Φ4 (49) 4! c − 2 c − 4! c The standard regularisation described above clearly preserves Weyl invariance even away from the small φ1 Note, at the intermediate stage the UV divergences are φ2 limit because, on dimensional grounds, the spontaneous regulated by introducing a cut-off, Λ2, when performing scale breaking factor, f, always compensates for the the k2 integration. Thus, in the Λ limit, we have renormalisation scale factor to give an overall constant the CW result: → ∞ under the log, together with a function of the scale in- λΛ2 λ2Φ4 λΦ2 1 variant field Θ = fΦ. W =Ω+ Φ2 + c ln c (50) eff 64π2 c 256π2 2Λ2 − 2 Expanding eq.(42) beyond leading order leads to higher order terms in θ but these non-renormalisable Following CW, the renormalisation conditions are: terms are small. The reason is that Planck scale is pre- dominantly due to the VEV of φ2 whereas the VEV of d2W d4W φ , which models the SM Higgs, is at the electroweak eff =0, eff = λ, Z =1 1 dΦ2 dΦ4 |Φc=M scale so that the non-renormalisable terms are Planck c Φc=0 c Φc=M (51) suppressed. In order to generate the hierarchy in the

Here CW renormalise at an “external” mass scale, M, VEV’s at the IR fixed point it is necessary that the only to avoid the IR singularity. Implementing these condi- large coupling is λ while the other couplings associated tions4 determines the counter terms and gives the final with the other scale invariant quartic interactions are hi- CW result: erarchically small and can be neglected when calculating the radiative corrections. λ λ2Φ4 Φ2 25 Of course there will be further terms when the gravi- W = Φ4 + c ln c (52) 4! c 256π2 M 2 − 6 tational interactions are included. Gravitational corrections require the addition of the Weyl tensor, W 2, and R2 terms, which are induced by matter loops and have log- arithmically running coefficients. An analysis of the full 4 There is no wave-function renormalisation at 1-loop order renormalization group equations appears in [15]. While 10 the Weyl tensor term is locally invariant, the R2 term is As a result the tree level potential introduced in eq.(55) only globally invariant. Hence we expect to maintain a has additional interactions of the form conserved current, Kµ′ , however the current will be modi- 2 W˜ (φ, χ) W (φ, χ)=(4 d) W (φ, χ) ln µ(φ, χ)+O(4 d)2. fied by the addition of a new term, Kµ′ = Kµ + c′∂µR/f0 − − − in the notation of [15]. We expect that this is a small (56) correction to the above scenario of a fixed ellipse, but Although these interactions vanish in 4 dimensions, may have some phenomenological implications that will they give a finite correction to Weff at 1-loop order be pursued elsewhere. because the underlying divergence in 4-dimensions can- Another potentially challenging consequence of the cels the 4 d factor in the additional term in eq.(56). Thus, due− to the additional interaction terms in eq.(56) gravitational corrections is that the λi become locked to the α by the renormalization group. This may necessi- that depend on the choice of µ(φ, χ), the scale invari- i ant d-dimensional theory is not the same as that defined tate some large fine-tunings to maintain a small cosmological constant and/or flat potentials. We feel that this purely in 4-dimensions. As a result the final regulated theory in 4-dimensional has additional terms that de- requires a more sophisticated fundamental analysis since the RG equations computed in flat geometries amount pend on the precise choice of the regulator µ(φ, χ). For the 2-scalar case with potential given by eq.(55) and the ot a “gauge choice” for the Weyl symmetry and do not admit analysis of the Weyl transformation. choice µ(φ, χ) = χ the additional term at one-loop is of the form φ6/χ2. While this is still scale invariant Finally, it is possible to maintain the local Weyl sym- it means the resulting 4-dimensional potential is differ- metry without choosing special values of the αi, but ent from that obtained by the regularisation procedure rather by introducing the Weyl vector potential. When discussed above. The origin of this discrepancy is that this is done, the dilaton is “eaten” to become the lon- the requirement that scale invariance be preserved in d- gitudinal part of a massive Weyl vector potential. The dimensions rather than regularisation ambiguity requires relationship of this to gravitational corrections and our such additional terms and defines a different theory. general framework is unexplored. In summary, we have shown that the standard regularisation procedure preserves scale invariance. It does not involve the introduction of an arbitrary regularisation function and, although it involves non-renormalisable in- 4. Scale invariant dimensional regularisation teractions, these are well defined. Of course it is possible to add additional non-polynomial terms to the theory Of course regularisation should not depend on the while preserving scale invariance but we see no reason to method used to control the intermediate divergences. Up do so. to now we have used a momentum space cut-off but it is straightforward to use dimensional regularisation. In this case one first continues the theory to d-dimensions IV. SUMMARY AND CONCLUSIONS and introduces an external mass scale, µ, to relate the 4- D dimensionless couplings to the dimension-full ones in We have discussed how inflation and Planck scale gen- d-dimensions. For the 2-scalar theory discussed above, eration can emerge from a dynamics associated with dimensional regularisation leads straightforwardly to the global Weyl symmetry and its current, Kµ. In the pre- form of eq.(53) with M replaced by µ. In this case the inflationary universe, the Weyl current density, K0 , is quartic and quadratic terms are automatically absent. driven to zero by general expansion. However, Kµ has The dependence on the mass parameter, µ, needed to a kernel structure, i.e., K = ∂ K and, as K 0, the µ µ 0 → continue away from four dimensions, will always appear kernel evolves as K K, constant. This resulting con- in the scale invariant ratio µ/f giving eq.(53) as before. stant K, that does not→ depend on the scalar potential, is the order parameter of the Weyl symmetry breaking; indeed, K directly defines the Planck mass. In N-multi-scalar-field theories K has the general form 5. Relation to previous regularisation proposals K = 1 (F ( φ ) N φ2 for nonminimal coupling − 2 { j } − i=1 i (1/12)F ( φj )R. The fields become constrained to − { } P Scale invariant dimensional regularisation that differs the manifold K K( φj ). In detail we have stud- from the one just described has been considered by sev- 1 →N { 2} ied F ( φj ) = 2 i=1 αiφi . This defines an ellipsoidal eral authors [5][6]. The method generally adopted to constraint{ } on the scalar field VEVs. An inflationary slow- maintain scale invariance in radiative order replaces µ by roll period is thenP associated with the field VEVs migrat- a function of the scalar fields, µ µ(φ ), with the appro- → i ing along the ellipse. Up to this point the fate of scale priate scaling behaviour. In this case the d-dimensional symmetry is entirely controlled by the inertial symmetry tree level potential V˜ has the form breaking, K K( φj ). A potential ultimately sculpts the ensuing slow→ roll{ on} the manifold to the IR, and de- 4 d V (φ, χ) µ(φ, χ) − V (φ, χ) . (55) fines the ultimate vacuum (together with any quantum ≡ e 11 effects that may distort the K ellipse [7]) This fixes the is technically natural with no hierarchy problem! relative value of the scalar field VEVs through quartic terms only. There is a harmless massless dilaton associated with the dynamical symmetry breaking which represents dilations of the ellipsoid. We emphasize that with Acknowledgements more general choices of F ( φj ), the constraint manifold can become a more general{ manifold} in the field space, We thank W. Bardeen, D. Ghilencea, A. Salvio, and and it would be of interest to explore the possibilities in A. Strumia for discussions. PGF acknowledges support this case. from STFC, the Beecroft Trust and the ERC. Part of this Any Weyl symmetry breaking effect at the quantum work was done at Fermilab, operated by Fermi Research level is intolerable and will show up as a nonzero diver- Alliance, LLC under Contract No. DE-AC02-07CH11359 with the United States Department of Energy. gence in the Kµ current. We showed how, due to the decoupling of the dilaton, these quantum effects actually preserve the Weyl symmetry using the normal momentum space cut-off or dimensional regularisation schemes. The potential scale dependence introduced by the “external” mass scale needed to regulate the logarithmic divergences is cancelled by the scale invariant order parameter responsible for spontaneous breaking of the Weyl symmetry. It would be of interest to study the local Wetyl invariant theories that involve the Weyl photon, as in Section II.C, in great detail. This provides an example of an inertial Higgs mechanism, and the dilaton is eaten and completely removed from the low energy spectrum. A strong motivation for considering such Weyl invariant theories is to provide a solution to the hierarchy problem of the Standard Model. In the absence of gravity or very massive states associated with the Landau pole of the Standard Model or of an extension of the Standard Model such as Grand or string unification, the Standard Model is natural in the sense that the quadratic divergence found in radiative corrections to the Higgs mass is unphysical and is cancelled by the mass counter term. Requiring scale invariance ensures that the Higgs is massless but, of course, some mechanism to spontaneously break the scale symmetry is needed. If gravity is included via the Weyl invariant extension discussed here, then the Standard Model plus gravity is natural in the sense just discussed. Of course it is still necessary that there be no massive states strongly coupled to the Higgs with masses much larger than the electroweak scale. Moreover the scale symmetry is now automatically spontaneously broken by the inertial mechanism. To obtain the hierarchy between the Planck scale and the electroweak breaking scale it is necessary to have hierarchically large ratios of the dimensionless couplings of the scalar potential. In the absence of gravitational radiative corrections, these ratios are only multiplicatively changed by radiative corrections and thus are natural. This may be seen from the underlying shift symmetry of the Weyl invariant Higgs potential. This shift symmetry is broken by the Higgs coupling to the Ricci scalar. To determine whether the hierarchy is ultimately preserved requires a calculation of the gravitational radiative corrections which is beyond the scope of the present paper. In a Weyl invariant variation of the Standard Model with no gravity, no grand unification and no Landau poles in the far UV the Higgs mass 12

[1] Hermann Weyl, Raum, Zeit, Materie (Space, Time, Mat- (2013) no.9, 096001 doi:10.1103/PhysRevD.87.096001 ter), “Lectures on General Relativity,” (in German) [arXiv:1212.4148 [hep-ph]]. D. Ghilencea, “Manifestly Berlin, Springer (1921). scale-invariant regularization and quantum effective op- [2] C. Wetterich, Nucl. Phys. B 302, 668 (1988); M. Sha- erators,” Phys. Rev. D 93 (2016) no.10, 105006; D. Ghi- poshnikov and D. Zenhausern, Phys. Lett. B 671, 162 lencea, “One-loop potential with scale invariance and (2009); D. Blas, M. Shaposhnikov and D. Zenhausern, effective operators,” PoS CORFU 2015 (2016) 040 Phys. Rev. D 84, 044001 (2011); [arXiv:1605.05632 [hep-ph]]; D. Ghilencea, Z. Lalak and [3] J. Garcia-Bellido, J. Rubio, M. Shaposhnikov and P. Olszewski, “Standard Model with spontaneously bro- D. Zenhausern, Phys. Rev. D 84, 123504 (2011); ken quantum scale invariance”, Phys. Rev. D 96 (2017) [4] R. Kallosh and A. Linde, JCAP 1310 (2013), no.5, 055034; D. Ghilencea, Z. Lalak and P. Olszewski, 033; J. J. M. Carrasco, R. Kallosh and A. Linde “Two-loop scale-invariant scalar potential and quantum [arXiv:1506.00936[hep-th]]; I. Quiros, arXiv:1405.6668 effective operators,” Eur. Phys. J. C 76 (2016) no.12, 656; [gr-qc]; arXiv:1401.2643 [gr-qc]; E. Guendelman, F. Gretsch and A. Monin, “Perturbative conformal sym- E. Nissimov and S. Pacheva, arXiv:1709.03786 metry and dilaton,” Phys. Rev. D 92 (2015) no.4, 045036; [gr-qc]; G. K. Karananas and M. Shaposhnikov, D. M. Ghilencea, “Quantum implications of a scale in- arXiv:1708.02220 [hep-th]; M. Kurkov, Eur. Phys. J. variant regularisation,” arXiv:1712.06024 [hep-th]. C 76, no. 6, 329 (2016); A. Karam, T. Pappas and [7] P. G. Ferreira, C. T. Hill and G. G. Ross, Phys. Rev. D K. Tamvakis, Phys. Rev. D 96, no. 6, 064036 (2017); 95, no. 4, 043507 (2017); Phys. Lett. B 763, 174 (2016). J. Rubio and C. Wetterich, Phys. Rev. D 96, no. 6, [8] S. R. Coleman and E. J. Weinberg, Phys. Rev. D 7, 1888 063509 (2017); G. K. Karananas and M. Shaposh- (1973). nikov, Phys. Lett. B 771, 332 (2017); M. Shimon, [9] W. A. Bardeen, “On naturalness in the standard model,” arXiv:1702.08472 [astro-ph.CO]; K. Kannike, M. Raidal, Fermilab-Conf-95-391-T; “Beyond Higgs,” Fermilab- C. Spethmann and H. Veerme, JHEP 1704 (2017) Conf-08-118-T and private communication. 026 [arXiv:1610.06571 [hep-ph]]. M. B. Einhorn and [10] P. G. Ferreira, C. T. Hill and G. G. Ross, Phys. Rev. D D. R. T. Jones, Phys. Rev. D 96, no. 12, 124025 (2017). 95, no. 6, 064038 (2017); [5] F. Englert, C. Truffin and R. Gastmans, Nucl. Phys. B [11] R. Jackiw and S. Y. Pi, Phys. Rev. D 91, no. 6, 067501 117 (1976) 407; C. Wetterich, Nucl. Phys. B 302 (1988) (2015). 668 ; M. Shaposhnikov and D. Zenhausern, Phys. Lett. [12] R. Jackiw, C. Nunez and S.-Y. Pi, Phys. Lett. A 347, 47 B 671 (2009) 187; M. E. Shaposhnikov and F. V. Tka- (2005). chov, “Quantum scale-invariant models as effective field [13] K. Allison, C. T. Hill and G. G. Ross, Nucl. Phys. B 891, theories,” arXiv:0905.4857 [hep-th]. 613 (2015); Phys. Lett. B 738, 191 (2014); [6] R. Armillis, A. Monin, M. Shaposhnikov, “Sponta- K. Kannike, G. Haetsi, L. Pizza, A. Racioppi, M. Raidal, neously Broken Conformal Symmetry: Dealing with the A. Salvio and A. Strumia, JHEP 1505, 065 (2015). Trace Anomaly,” JHEP 1310 (2013) 030; C. Tamarit, [14] C. T. Hill, Phys. Rev. D 89, 073003 (2014). “Running couplings with a vanishing scale anomaly,” [15] A. Salvio and A. Strumia, JHEP 1406, 080 (2014). JHEP 1312 (2013) 098 F. Bezrukov, G. K. Karananas, J. Rubio and M. Shaposhnikov, Phys. Rev. D 87