The Higgs mechanism

An explicit mass term $\tfrac12 m_B^2 B_\mu B^\mu$ for a gauge field is not allowed. It is not a matter of taste or convenience: such a term is not invariant under the local gauge transformation $B_\mu\to B_\mu-\tfrac1g\partial_\mu\alpha(x)$ that the whole edifice of gauge theory is built on, so writing it down destroys the very symmetry that makes the theory renormalizable and sensible. Yet the $W$ and $Z$ bosons that carry the weak force are heavy — some eighty and ninety times the mass of a proton. The Higgs mechanism is how nature has it both ways: the gauge symmetry is kept exactly, and the gauge bosons are massive anyway. The trick is to let the symmetry be hidden by the vacuum rather than broken in the Lagrangian. The ingredient is a scalar field with a potential $V(\phi)=\mu^2\phi^\dagger\phi+\lambda(\phi^\dagger\phi)^2$ shaped like the bottom of a wine bottle. When the mass-squared parameter $\mu^2$ is negative, the field cannot sit at zero; its lowest-energy configuration lies on a whole circle of minima, and choosing one point on that circle spontaneously breaks the symmetry. Expand about that point and two things happen at once. The radial excursion out of the trough costs energy — a genuine massive scalar, the Higgs boson. The excursion around the flat trough costs nothing — a massless Goldstone boson. And when the broken symmetry is a local gauge symmetry, that would-be Goldstone does not survive as a physical particle at all: it is absorbed as the longitudinal polarization the gauge boson needs in order to be massive. The boson "eats" the Goldstone. This article builds that story in three stages — global breaking, the Abelian gauged version, and the connection to the electroweak scale $v=246$ GeV — checking the degree-of-freedom bookkeeping at every step.

Key takeaways

A gauge-boson mass term $\tfrac12 m_B^2 B_\mu B^\mu$ is forbidden by local gauge invariance; the Higgs mechanism gives mass without ever writing one down.
A scalar potential $V(\phi)=\mu^2\phi^\dagger\phi+\lambda(\phi^\dagger\phi)^2$ with $\lambda>0$ but $\mu^2<0$ has its minimum off zero, on a circle of radius $v=\sqrt{-\mu^2/\lambda}$ .
Expanding about one chosen vacuum leaves a massive radial scalar (the Higgs, $m_H=\sqrt{2\lambda\,}\,v$ ) and a massless angular Goldstone boson in the flat direction.
Gauge the symmetry and the covariant derivative turns the vacuum value into a gauge-boson mass term $\tfrac12 g^2v^2 B_\mu B^\mu$ , so $m_B=gv$ .
A gauge transformation removes the Goldstone field entirely (unitary gauge); it reappears as the longitudinal polarization of the now-massive vector, so degrees of freedom are exactly conserved — 4 before, 4 after.
In the Standard Model $v=(\sqrt2\,G_F)^{-1/2}=246$ GeV and $m_H=125$ GeV fix the self-coupling $\lambda\approx0.13$ ; the same $v$ sets $M_W=\tfrac12 g_W v$ .
Key idea: mass for a gauge boson is forbidden by gauge invariance — but the invariance can be hidden rather than broken. A scalar field with a nonzero vacuum value supplies the missing longitudinal degree of freedom, and the gauge boson eats the would-be Goldstone boson to become massive.

Where this lives today

The Higgs mechanism stopped being a hypothesis on 4 July 2012, when ATLAS and CMS announced a new boson at $m_H=125$ GeV; the current world average is $m_H=125.25\pm0.17$ GeV. Since then the programme has turned from discovery to precision: measuring whether the new particle couples to every other particle in proportion to that particle's mass, the sharp signature of a field whose vacuum value is the common origin of mass. Couplings to the heaviest states — the top quark, the $W$ and $Z$ , the bottom quark and the tau — are confirmed at the expected strengths to $10$ – $20\%$ , and the rare decay to two muons, a second-generation coupling, has now been seen. The frontier is the Higgs self-coupling $\lambda$ : the trilinear $hhh$ vertex, which fixes the shape of the potential near its minimum, is accessible only through the very rare production of two Higgs bosons at once — a headline target for the High-Luminosity LHC. Deeper still is vacuum stability: feeding the measured $m_H=125$ GeV and top mass $m_t\approx173$ GeV into the running of $\lambda$ suggests our electroweak vacuum is metastable — long-lived, but perhaps not the true minimum — a result sitting provocatively close to the stable/unstable boundary and hinting at new physics between here and the Planck scale.

The physical picture

Picture the scalar field as a marble in a potential shaped like the punt at the bottom of a wine bottle — a central hump ringed by a circular valley. When $\mu^2>0$ the bottle has an ordinary bowl and the marble rests at the centre, $\phi=0$ , respecting the symmetry. When $\mu^2<0$ the centre becomes a hilltop and the marble must roll down into the valley. Every point around the circular trough is equally deep, so the field is forced to pick one — and the moment it picks, the rotational symmetry of the Lagrangian is no longer visible in the state of the system, even though the Lagrangian itself never changed. The first figure is that potential, with the two kinds of motion it allows and the real numbers the electroweak vacuum actually takes.

Figure

The scalar potential V(φ) with μ² negative. A cross-section through the wine-bottle shape shows two degenerate minima at φ = ±v; the field must choose one. Motion up the wall (radial, η) costs energy and is the massive Higgs, m_H = √(2λ) v = 125 GeV; motion around the flat trough (angular, ξ) costs nothing and is the massless Goldstone. The Standard-Model vacuum sits at v = 246 GeV with self-coupling λ = m_H²/2v² ≈ 0.13

The radial and angular motions are the whole story of what particles survive. Climb the wall of the trough and you feel a restoring force — that curvature is the mass of the physical Higgs boson. Slide around the flat bottom and there is no restoring force at all — that is the massless Goldstone boson demanded by Goldstone's theorem whenever a continuous global symmetry is spontaneously broken. The decisive twist comes when the symmetry is instead local: then a gauge field is present, the Goldstone is not gauge-invariant, and it can be rotated away entirely. It is not lost — it becomes the extra polarization a massive vector must have. The second figure is that bookkeeping, and it is the punchline of the mechanism: the count of degrees of freedom balances exactly, and the same vacuum value $v$ that hides the symmetry sets the gauge-boson mass.

Figure

Counting degrees of freedom: the Goldstone is eaten. Before breaking, a massless gauge boson carries 2 transverse polarizations and the complex scalar carries 2 real fields (η, ξ) — total 4. After breaking, the gauge boson becomes massive with 3 polarizations (2 transverse plus 1 longitudinal, the eaten ξ) and one physical Higgs remains — total 4. The same vacuum value drives the mass: M_W = ½ g_W v = ½ × 0.653 × 246 GeV = 80.4 GeV

The core equations

Everything starts with the scalar potential. It is the most general renormalizable, gauge-invariant potential for a complex field, and the sign of the quadratic coefficient $\mu^2$ is what decides whether the symmetry is realized normally or hidden. With $\lambda>0$ (so the potential is bounded below) the interesting case is $\mu^2<0$ .

V(\phi)=\mu^2\,\phi^\dagger\phi+\lambda\,(\phi^\dagger\phi)^2,\qquad \lambda>0 \qquad (1)

where: $\phi=\tfrac1{\sqrt2}(\phi_1+i\phi_2)$ — a complex scalar field; $\mu^2$ — the mass-squared parameter (its sign, not its value, is what matters); $\lambda$ — the dimensionless quartic self-coupling. Read it as a shape: $\mu^2>0$ is a bowl with its bottom at the origin, $\mu^2<0$ is a wine-bottle punt with a ring of minima away from the origin. Minimizing $V$ with respect to $\phi^\dagger\phi$ locates the vacuum. For $\mu^2<0$ the minimum is not at zero but on a circle, and its radius defines the vacuum expectation value $v$ — the single most important number in the whole construction.

\frac{\partial V}{\partial(\phi^\dagger\phi)}=0\ \Rightarrow\ \phi^\dagger\phi=-\frac{\mu^2}{2\lambda}\equiv\frac{v^2}{2},\qquad v=\sqrt{-\mu^2/\lambda} \qquad (2)

where: $v$ — the vacuum expectation value (vev), so $\langle\phi\rangle=v/\sqrt2$ ; the minimum is degenerate around the whole circle $\phi_1^2+\phi_2^2=v^2$ . Choosing one point (say $\phi_1=v$ , $\phi_2=0$ ) is the act of spontaneous symmetry breaking — the Lagrangian is untouched, only the expansion point is fixed. Expand the field about the chosen vacuum in a radial part $\eta$ and an angular part $\xi$ . Substituting into (1) and keeping the quadratic terms gives the spectrum directly: the radial field is massive, the angular field is massless.

\phi=\tfrac1{\sqrt2}\,(v+\eta+i\xi),\qquad m_\eta=\sqrt{2\lambda v^2}=\sqrt{-2\mu^2},\qquad m_\xi=0 \qquad (3)

where: $\eta$ — the radial (Higgs) excitation, with mass fixed by the curvature $V''=2\lambda v^2$ at the bottom of the trough; $\xi$ — the angular (Goldstone) excitation, massless because the trough is flat in that direction. The massless $\xi$ is Goldstone's theorem in action for the global symmetry. Now gauge the symmetry: promote $\partial_\mu\to D_\mu=\partial_\mu+igB_\mu$ so the Lagrangian is invariant under local phase rotations. The kinetic term $|D_\mu\phi|^2$ , evaluated at the vacuum, contains a term quadratic in $B_\mu$ — a mass term the gauge principle would otherwise never permit.

|D_\mu\phi|^2\ \supset\ \tfrac12\,g^2 v^2\,B_\mu B^\mu\ \Rightarrow\ m_B=g\,v \qquad (4)

where: $g$ — the gauge coupling; $B_\mu$ — the gauge field. The vacuum value $v$ has done what an explicit mass term could not: it has given the gauge boson a mass $m_B=gv$ while the Lagrangian stays exactly gauge-invariant. The photon-like massless case is recovered only if the vacuum is uncharged ( $v=0$ ). The same expansion also produces a cross-term mixing the gauge field with the Goldstone field, the tell-tale sign that $\xi$ is not a physical particle. A gauge transformation (the unitary gauge) sets $\xi=0$ and removes it, and it resurfaces as the longitudinal polarization of the massive $B$ .

|D_\mu\phi|^2\ \supset\ g\,v\,B_\mu\,\partial^\mu\xi\ \xrightarrow{\ \text{unitary gauge}\ }\ 0,\qquad \phi=\tfrac1{\sqrt2}(v+h) \qquad (5)

where: $h$ — the single surviving physical (Higgs) scalar; the mixing term $gvB_\mu\partial^\mu\xi$ is exactly cancelled by the field redefinition $B_\mu\to B_\mu+\tfrac1{gv}\partial_\mu\xi$ , which absorbs the Goldstone. This is the "eating": one scalar degree of freedom becomes the vector's longitudinal mode. Finally, the connection to the real world. In the Standard Model the vev is fixed not by choice but by the measured strength of the weak interaction, and together with the measured Higgs mass it pins down the self-coupling $\lambda$ .

v=\big(\sqrt2\,G_F\big)^{-1/2}=246\ \text{GeV},\qquad m_H=\sqrt{2\lambda}\,v\ \Rightarrow\ \lambda=\frac{m_H^2}{2v^2}\approx0.13 \qquad (6)

where: $G_F=1.166\times10^{-5}\ \text{GeV}^{-2}$ — the Fermi constant; $m_H=125$ GeV — the measured Higgs mass. Read it as a closed loop: the weak scale fixes $v$ , the discovery of the Higgs fixes $m_H$ , and between them the last free parameter of the scalar potential, $\lambda$ , is determined.

Derivation ASpontaneous breaking of a global U(1) and the Goldstone boson

The goal is to show, from the potential of equation (1) alone, that a complex scalar with $\mu^2<0$ has a nonzero vacuum value $v=\sqrt{-\mu^2/\lambda}$ , and that expanding about it produces exactly one massive real scalar of mass $\sqrt{2\lambda v^2}$ together with one strictly massless scalar — the Goldstone boson mandated whenever a continuous global symmetry is spontaneously broken. We locate the minimum, expand in radial and angular fields, and read the masses off the quadratic terms.

Step 1. Find the vacuum. Writing $\phi^\dagger\phi=\rho$ , the potential $V=\mu^2\rho+\lambda\rho^2$ is minimized at $\mathrm{d}V/\mathrm{d}\rho=\mu^2+2\lambda\rho=0$ , i.e. $\rho=-\mu^2/2\lambda$ . For $\mu^2<0$ this is positive and real, giving the circle $\phi_1^2+\phi_2^2=-\mu^2/\lambda\equiv v^2$ . Why: the sign of $\mu^2$ is doing all the work — a positive $\mu^2$ would force $\rho=0$ and a symmetric vacuum, whereas a negative one pushes the field off the origin onto a degenerate ring.

Step 2. Break the symmetry by choosing a point. The Lagrangian is invariant under $\phi\to e^{i\alpha}\phi$ , so every point on the ring is physically equivalent; but the field must sit at one of them. Choose $\phi_1=v$ , $\phi_2=0$ and expand $\phi=\tfrac1{\sqrt2}(v+\eta+i\xi)$ , with $\eta$ the radial and $\xi$ the tangential fluctuation. Why: the symmetry is not destroyed — it still relates the different vacua — but it is no longer manifest once one vacuum is selected, which is the precise meaning of "spontaneous".

Step 3. Extract the scalar masses. Substituting into $V$ and keeping terms quadratic in the fluctuations gives $V=\text{const}+\lambda v^2\eta^2+O(\eta^3,\xi^3,\dots)$ , with no $\xi^2$ term at all. Comparing to the canonical $\tfrac12 m^2\varphi^2$ yields $m_\eta^2=2\lambda v^2$ and $m_\xi^2=0$ . Why: the potential curves upward along the radial direction (out of the trough) but is dead flat around the trough, so the radial mode is massive and the angular mode costs no energy — it is the Goldstone boson.

Step 4. Identify the Goldstone as the symmetry direction. The flat direction $\xi$ is exactly the direction in which the U(1) rotation moves the vacuum; a symmetry transformation carries one degenerate vacuum into another at no energy cost, which is why the associated excitation is massless. Why: this makes Goldstone's theorem geometric — one massless boson appears for each broken continuous symmetry generator, because each generator maps the vacuum along a flat valley. *As a consistency check: the minimum sits at $\phi^\dagger\phi=-\mu^2/2\lambda$ ; the radial mass satisfies $m_\eta^2=2\lambda v^2=-2\mu^2>0$ ; and $m_\xi^2=0$ identically.*

Go deeper · Derivation B · 1 minThe Abelian Higgs mechanism: a gauge boson acquires mass

Derivation BThe Abelian Higgs mechanism: a gauge boson acquires mass

The aim is to gauge the U(1) symmetry and show that the very same vacuum value $v$ that produced a Goldstone boson now produces a gauge-boson mass term $\tfrac12 g^2v^2 B_\mu B^\mu$ , so that $m_B=gv$ , while simultaneously generating a mixing term $gvB_\mu\partial^\mu\xi$ whose appearance is the warning that $\xi$ is unphysical. We promote the derivative to a covariant one, expand the kinetic term about the vacuum, and collect the quadratic terms in the gauge and scalar fields.

Step 1. Gauge the symmetry. Replace $\partial_\mu\phi\to D_\mu\phi=(\partial_\mu+igB_\mu)\phi$ and add the gauge kinetic term $-\tfrac14 F_{\mu\nu}F^{\mu\nu}$ , so the Lagrangian is invariant under the local transformation $\phi\to e^{i\alpha(x)}\phi$ , $B_\mu\to B_\mu-\tfrac1g\partial_\mu\alpha$ . Why: local invariance is what forbids an explicit mass term $\tfrac12 m_B^2 B_\mu B^\mu$ in the first place, so any mass the boson ends up with must come entirely from the covariant derivative, not from a term put in by hand.

Step 2. Expand the covariant derivative at the vacuum. With $\phi=\tfrac1{\sqrt2}(v+\eta+i\xi)$ , compute $D_\mu\phi=\tfrac1{\sqrt2}\big[(\partial_\mu\eta-gB_\mu\xi)+i(\partial_\mu\xi+gB_\mu(v+\eta))\big]$ . Why: the vacuum value $v$ rides inside the covariant derivative, so when the derivative is squared it multiplies the gauge field by $v$ — this is the structural reason a nonzero vev talks directly to the gauge sector.

Step 3. Collect the mass term. Squaring and keeping quadratic pieces, $|D_\mu\phi|^2\supset\tfrac12 g^2 v^2 B_\mu B^\mu$ , a Proca mass term with $m_B=gv$ ; the transverse kinetic term $-\tfrac14 F_{\mu\nu}F^{\mu\nu}$ is untouched. Why: a term $\tfrac12 m^2 B_\mu B^\mu$ with $m=gv$ has appeared without spoiling gauge invariance, because it was generated by the invariant object $|D_\mu\phi|^2$ evaluated at a symmetry-breaking vacuum — precisely the loophole the mechanism exploits.

Step 4. Notice the mixing. The same expansion contains the cross-term $gv\,B_\mu\partial^\mu\xi$ , linear in both the gauge field and the Goldstone. Why: a bilinear kinetic mixing between $B_\mu$ and $\xi$ signals that neither is the true propagating field — the physical massive vector is a combination of them, and this term is exactly what the unitary gauge (next derivation) will remove. *As a consistency check: expanding $|D_\mu\phi|^2$ yields $\tfrac12 g^2v^2 B_\mu B^\mu$ (so $m_B=gv$ ) and the mixing term $gvB_\mu\partial^\mu\xi$ , with the $F_{\mu\nu}$ kinetic term unchanged.*

Go deeper · Derivation C · 2 minUnitary gauge and the conservation of degrees of freedom

Derivation CUnitary gauge and the conservation of degrees of freedom

The goal is to make the Goldstone field disappear and to show that the total number of physical degrees of freedom is exactly preserved: four before breaking (two transverse gauge polarizations plus two real scalars) and four after (three polarizations of the massive vector plus one Higgs). We use the gauge freedom to rotate away $\xi$ , watch it reappear as the vector's longitudinal mode, and tally the count on both sides.

Step 1. Use the gauge freedom to remove $\xi$ . Because the phase of $\phi$ can be rotated locally, choose the gauge parameter $\alpha(x)$ so that $\phi$ is everywhere real: $\phi=\tfrac1{\sqrt2}(v+h)$ , with $\xi\equiv0$ . This is the unitary gauge. Why: the Goldstone is pure gauge — it can be set to zero by a gauge choice — which is the sharpest statement that it is not a physical particle in the gauged theory, unlike in the global case.

Step 2. Absorb the mixing. Equivalently, redefine $B_\mu\to B_\mu+\tfrac1{gv}\partial_\mu\xi$ ; this shift cancels the $gvB_\mu\partial^\mu\xi$ term exactly and eliminates $\xi$ from the Lagrangian. What remains is a massive vector $B_\mu$ and a single real scalar $h$ . Why: the Goldstone has not vanished into nothing — it has been repackaged into the gauge field, which is why the operation is a redefinition rather than a deletion.

Step 3. Count before. A massless gauge boson has 2 physical (transverse) polarizations; a complex scalar has 2 real components. Total: $2+2=4$ . Why: a massless spin-1 field is constrained to two helicities, so the pre-breaking count is unambiguous.

Step 4. Count after. A massive vector has 3 polarizations (2 transverse plus 1 longitudinal); the surviving Higgs $h$ is 1 real scalar. Total: $3+1=4$ . The extra longitudinal polarization is precisely the eaten Goldstone. Why: the books balance exactly — no degree of freedom was created or destroyed, only redistributed, which is the deepest consistency check on the whole mechanism and the reason it does not secretly add or lose states. *As a consistency check: the unitary gauge removes $\xi$ and the mixing term, and the degree-of-freedom count $2+2=4$ (before) equals $3+1=4$ (after).*

Limiting cases

Case	Result	Meaning
$\mu^2>0$	minimum at $\phi=0$ , $v=0$	symmetry unbroken; gauge boson stays massless, no Higgs vev, no mass generation
$\mu^2<0$	$v=\sqrt{-\mu^2/\lambda}\neq0$	symmetry hidden; gauge boson mass $m_B=gv$ , one physical Higgs $m_H=\sqrt{2\lambda}\,v$
Global symmetry (no gauge field)	massless Goldstone $\xi$ survives	Goldstone's theorem: one massless scalar per broken continuous generator
Local symmetry (gauged)	Goldstone eaten, no massless scalar	the would-be Goldstone becomes the longitudinal polarization of the massive vector
$g\to0$ (decoupling)	$m_B\to0$ , Goldstone reappears	the gauge boson decouples and the theory reverts to the global case
$\lambda\to0$	$m_H\to0$ , potential flattens	self-coupling vanishing sends the Higgs mass to zero and $v$ becomes ill-defined
Unitary gauge	$\xi=0$ , only $h$ and massive $B_\mu$	the physical spectrum made manifest; no unphysical Goldstone in the Lagrangian

Common trapSpontaneous symmetry breaking destroys the symmetry and loses a particle

It does neither, and mistaking this is the single most common error about the mechanism. Spontaneous symmetry breaking does not change the Lagrangian and does not destroy the symmetry — the Lagrangian remains exactly invariant, and the symmetry still relates the infinitely many degenerate vacua to one another. What happens is only that the field must sit at one vacuum, so the symmetry is no longer manifest in that state; it is hidden, not broken. (This is why "hidden symmetry" is the more accurate name.) Nor is a degree of freedom lost when the Goldstone boson disappears: in the gauged theory the Goldstone is not annihilated but absorbed, becoming the longitudinal polarization that a massive vector requires. The tally is exact — two transverse polarizations plus two scalars before, three polarizations plus one Higgs after, four in both cases. A related trap: the mechanism does not explain most mass in the universe. It gives mass to the gauge bosons and, via Yukawa couplings, to the elementary fermions — but a proton's mass is roughly $99\%$ QCD binding energy, not Higgs-generated. This is a theory of gauge-boson mass, not of mass in general.

Worked examples

1 · The Higgs potential from two measured numbers (fully worked)

The electroweak vacuum value is $v=246$ GeV and the Higgs mass is $m_H=125$ GeV. Determine the self-coupling $\lambda$ , the mass parameter $\mu$ , and the depth of the potential at its minimum — that is, reconstruct the entire scalar potential from the two things experiment hands us.

Step 1. Get $\lambda$ from the curvature. The Higgs mass is the curvature at the bottom of the well, $m_H^2=2\lambda v^2$ , so $\lambda=\dfrac{m_H^2}{2v^2}=\dfrac{(125)^2}{2\,(246)^2}=\dfrac{15625}{121032}=\mathbf{0.129}$ . A modest, perturbative coupling — the scalar sector is weakly interacting.

Step 2. Get $\mu^2$ from the vev. Since $v^2=-\mu^2/\lambda$ , we have $\mu^2=-\lambda v^2=-(0.129)(246)^2=-7810\ \text{GeV}^2$ , so $\mu=\sqrt{7810}=\mathbf{88.4\ \text{GeV}}$ (with the understanding that $\mu^2$ is negative — this is the "wrong-sign" mass term that triggers breaking). As a cross-check, $m_H=\sqrt{-2\mu^2}=\sqrt{15620}=125.0$ GeV. Consistent.

Step 3. Get the depth. At the minimum $\phi^\dagger\phi=v^2/2$ , so $V_{\min}=\mu^2\big(\tfrac{v^2}{2}\big)+\lambda\big(\tfrac{v^2}{2}\big)^2=-\tfrac14\lambda v^4=-\tfrac18 m_H^2 v^2$ . Numerically $V_{\min}=-\tfrac18(125)^2(246)^2=-\mathbf{1.19\times10^{8}\ \text{GeV}^4}$ . The two measured numbers $v=246$ GeV and $m_H=125$ GeV fix the whole potential: $\lambda=0.13$ , $\mu\approx88$ GeV, and a well $1.2\times10^8$ GeV⁴ deep. This is why the Higgs mass measurement in 2012 was so consequential — it was the last unknown parameter of the scalar potential, and with it the electroweak vacuum is completely specified.

2 · The W-boson mass from the vacuum value

The $SU(2)_L$ gauge coupling is $g_W=0.653$ . Using $M_W=\tfrac12 g_W v$ with $v=246$ GeV, find $M_W$ , and confirm the degree-of-freedom bookkeeping for the three weak bosons.

Step 1. Plug into the mass formula. The Standard-Model relation is $M_W=\tfrac12 g_W v$ (the factor $\tfrac12$ coming from the doublet normalization of the Higgs field, versus $m_B=gv$ for the minimal Abelian toy). Then $M_W=\tfrac12(0.653)(246)=\mathbf{80.3\ \text{GeV}}$ , right on the measured $80.4$ GeV.

Step 2. Check the counting. The Higgs of the Standard Model is a complex $SU(2)$ doublet: 4 real scalar degrees of freedom. After breaking, three are eaten — one each by $W^+$ , $W^-$ , and $Z$ , giving each its longitudinal polarization and hence its mass. Three Goldstones eaten, three massive vectors. The fourth scalar is left over.

Step 3. Identify the survivor. That one uneaten real scalar is the physical Higgs boson $h$ , discovered in 2012. From a single doublet: 4 scalars → 3 eaten (the longitudinal $W^\pm$ , $Z$ ) + 1 physical Higgs, and $M_W=\tfrac12 g_W v=80.3$ GeV follows directly from the same vacuum value $v$ that hides the symmetry. The photon stays massless because one particular combination of the neutral fields never couples to the vacuum charge — the unbroken electromagnetic $U(1)$ .

Physics of the portrait

The mechanism carries one name by convention and many by right: it was published in 1964 by three independent groups. The portrait is Peter Higgs, whose paper alone predicted an observable consequence — a new massive scalar particle — and whose name the boson now bears.

Origins · 1 minPeter Higgs and how this was discovered

Peter Higgs

1929–2024, Edinburgh; the 1964 papers and the 2012 discovery · spontaneous gauge-symmetry breaking and the scalar boson

In 1964, three teams — Robert Brout and François Englert in Brussels; Peter Higgs in Edinburgh; and Gerald Guralnik, C. Richard Hagen and Tom Kibble in London — independently found the loophole in Goldstone's theorem: when the broken symmetry is local, the would-be massless Goldstone boson is not a physical particle but is absorbed by the gauge field, which becomes massive. The idea had a non-relativistic precursor in Philip Anderson's 1962 observation that a photon inside a superconductor is effectively massive (the Meissner effect), but the relativistic gauge-theory version was the 1964 achievement. Higgs's contribution stood slightly apart: where the others focused on the gauge-boson mass, his second paper explicitly noted that the mechanism must leave behind a massive scalar excitation — the quantum of the radial field, the particle we now call the Higgs boson. That prediction of a concrete new particle is why his name attached to it. Steven Weinberg and Abdus Salam wove the mechanism into the electroweak theory in 1967–68, giving mass to the $W$ and $Z$ while sparing the photon. For nearly half a century the boson remained hypothetical, until ATLAS and CMS at the LHC found it at 125 GeV in July 2012. Englert and Higgs shared the 2013 Nobel Prize; Brout had died in 2011, and the prize is not awarded posthumously.

History

1961

— Goldstone's theorem: breaking a continuous global symmetry forces a massless boson

1964

— Englert–Brout, Higgs, and Guralnik–Hagen–Kibble publish the relativistic mechanism

1967

— Weinberg and Salam apply it to give the W and Z mass in the electroweak theory

2012

— ATLAS and CMS discover the Higgs boson at about 125 GeV

2013

— Englert and Higgs receive the Nobel Prize in Physics

Going further

Texts. Thomson, Modern Particle Physics, §17.5 (17.5.1–17.5.3, mass generation for a real scalar, a gauge boson, and the electroweak sector) · Griffiths, Introduction to Elementary Particles, 2nd ed., Ch. 10, §10.7–10.9 · Halzen & Martin, Quarks and Leptons, Ch. 14 (spontaneous symmetry breaking and the Higgs mechanism) · Peskin & Schroeder, An Introduction to Quantum Field Theory, §20.1 (spontaneous breaking of gauge symmetries) · Schwartz, Quantum Field Theory and the Standard Model, Ch. 28. The key results used throughout this article are: (1) $v=\sqrt{-\mu^2/\lambda}$ and $m_\eta^2=2\lambda v^2$ , $m_\xi^2=0$ from the potential; (2) $|D_\mu\phi|^2$ gives $\tfrac12 g^2v^2 B_\mu B^\mu$ so $m_B=gv$ , plus the mixing $gvB_\mu\partial^\mu\xi$ ; (3) the degree-of-freedom count $2+2=4=3+1$ ; (4) $\lambda=m_H^2/2v^2=0.129$ , $\mu=88.4$ GeV, $V_{\min}=-\tfrac18 m_H^2 v^2$ for $v=246$ GeV, $m_H=125$ GeV; (5) $M_W=\tfrac12 g_W v=80.3$ GeV.

Prerequisites. Spontaneous Symmetry Breaking · Lagrangians and Gauge Invariance · The Gauge Boson Mass Problem · The Local Gauge Principle. Leads to. Higgs Couplings · Electroweak Unification · The Higgs Discovery.

Ask the tutor: Ask me to expand $|D_\mu\phi|^2$ term by term to show exactly where the gauge-boson mass and the eaten Goldstone come from, or to walk through why the proton's mass is almost entirely QCD and not the Higgs.

Does the Higgs field really give everything its mass?

No — this is the most common overreach. The Higgs mechanism gives mass to the electroweak gauge bosons ( $W$ and $Z$ ), and through separate Yukawa couplings it gives mass to the elementary fermions (quarks and charged leptons). But the mass of ordinary matter is dominated by protons and neutrons, and roughly $99\%$ of a proton's mass is the binding energy of the gluon field and the kinetic energy of its quarks — pure QCD, with no Higgs involved. If you switched off the Higgs field, electrons and quarks would become massless and atoms would fall apart, but a "proton" built of massless quarks would still weigh almost the same. The Higgs is the origin of gauge-boson mass and elementary-fermion mass, not of mass in general.

What is the difference between the Higgs field, the mechanism, the vev, and the boson?

They are four distinct things people routinely conflate. The Higgs field is the scalar field $\phi$ present everywhere in space. The Higgs mechanism is the process by which that field's nonzero vacuum value hides the gauge symmetry and gives the gauge bosons mass. The vacuum expectation value (vev), $v=246$ GeV, is the specific nonzero value the field settles into — it is the vev, not the boson, that does the mass generation. The Higgs boson is the quantum of the small oscillations of the field about that vacuum value, the radial excitation $h$ with $m_H=125$ GeV. The mechanism would work — bosons would still be massive — even if the boson were too heavy to produce; discovering the boson is what confirmed the field is really there.

Why does the mass term need μ² to be negative?

Because a positive $\mu^2$ is an ordinary mass term with the minimum at $\phi=0$ , and nothing interesting happens — the symmetry is exact and the gauge boson stays massless. The whole mechanism requires the field to not want to sit at zero, and that is what a negative $\mu^2$ arranges: it turns the origin from a valley into a hilltop, so the field rolls out to the ring of true minima at $|\phi|=v$ . The name "wrong-sign mass term" is a bit misleading — with $\mu^2<0$ , the parameter $\mu^2$ is no longer the mass of anything; the actual physical mass is the Higgs mass $m_H=\sqrt{-2\mu^2}$ , which is perfectly positive. The negative $\mu^2$ is best read as the instability of the symmetric point, not as an imaginary mass.

What does it actually mean for the gauge boson to "eat" the Goldstone?

A massless spin-1 particle has only two polarizations (transverse, like a photon), but a massive one must have three, including a longitudinal mode in which the field oscillates along its direction of motion. The Higgs mechanism has to supply that third polarization from somewhere, and it takes it from the scalar sector: the Goldstone degree of freedom, which in the ungauged theory would be a massless particle, is instead reinterpreted as the longitudinal oscillation of the vector. Concretely, in the unitary gauge you rotate the Goldstone field to zero, and the very same degree of freedom reappears inside the gauge field's now-three-component description. "Eating" is a vivid way of saying the bookkeeping transfers one scalar degree of freedom into the vector — nothing is created or destroyed.

If the Goldstone can be gauged away, was it ever real?

This is a genuinely subtle point. In the global theory (no gauge field) the Goldstone boson is a real, physical, massless particle — you cannot gauge it away because there is no gauge freedom. In the local (gauged) theory it is not physical: it can be set to zero by a gauge choice, which is the precise statement that it does not correspond to an observable particle. But its degree of freedom does not vanish — it is carried by the longitudinal polarization of the massive vector, which is observable. So the honest statement is: the Goldstone is not a physical particle in the gauged theory, but the degree of freedom it represents is physical and shows up as the vector's third polarization. Different gauges (unitary versus $R_\xi$ ) make different bookkeeping choices, but all physical predictions agree.

How does this connect to the full electroweak theory with W, Z, and photon?

The Abelian toy in this article breaks a single U(1); the Standard Model breaks the larger group $SU(2)_L\times U(1)_Y$ down to the electromagnetic $U(1)$ . The Higgs is then a complex $SU(2)$ doublet with four real scalar degrees of freedom. Three of them are Goldstones, eaten to become the longitudinal polarizations of the $W^+$ , $W^-$ , and $Z$ , which is why those three bosons are massive. The fourth is the physical Higgs boson. One particular combination of the neutral gauge fields never couples to the vacuum and stays massless — that is the photon. The mixing angle that sorts the neutral sector into a massless photon and a massive $Z$ , and the mass relation $M_W=M_Z\cos\theta_W$ , are developed in the electroweak-unification article; this one supplies the mechanism they rely on.