The Quantum Vacuum; Chiral Symmetry and the Mass of the Pion

The pion is not light by accident. It is what a broken symmetry leaves behind, and the small mass it does have measures how approximate that symmetry was to begin with.

🎯 Why this matters

This is also why nuclear forces reach as far as they do. The lightest hadron sets the longest range, and the pion is light for a reason that has nothing to do with how strong the interaction is.

§6.7 said 99 % of the proton’s mass is field energy. This page is about the field it is energy of — the vacuum — and then about the one particle that escapes the pattern by being far too light.

§6.8 The vacuum is not empty

§5.8 had virtual pairs appearing around a charge. Remove the charge and they are still there. A pair of fermions can borrow enough energy to exist provided it pays the loan back fast enough, Eq. (6.66):

\htmlClass{t-dt}{\Delta t} \lesssim \frac{1}{\htmlClass{t-m}{2m}}

(6.66)

How long a particle–antiparticle pair may exist without anyone being able to object. In natural units (ħ = 1) an inverse mass is a time, which is the whole content of the inequality.

Every symbol, one at a time

Hover or tap a symbol above — it lights up in the equation and its meaning, units and type appear here.

Figs. 6.29–6.31 — vacuum polarization with nothing to polarize

Click a vertex or an internal line.

The book's Figs. 6.29, 6.30 and 6.31 are the same object at successive orders. The third line is the one QED cannot draw: gluons coupling to gluons with no fermion anywhere, and massless, so unconstrained in scale.

how big a vacuum fluctuation is

hbar, c = 6.582119569e-22, 2.99792458e23     # MeV s,  fm/s

print("Eq. (6.66): a pair of total mass 2m lives dt < 1/2m, inside a region c*dt")
for m2, lab in [(1.0, "e+e-        (2m = 1 MeV)"),
                (9.3, "d dbar      (2m = 9.3 MeV)"),
                (211.3, "mu+mu-      (2m = 211 MeV)"),
                (280.0, "a pion pair (2m = 280 MeV)")]:
    dt = hbar / m2
    print(f"  {lab:26s} dt = {dt:.2e} s,  c dt = {dt*c:8.2f} fm")
print()
print("The book's electron numbers check out exactly: 6.6e-22 s and 200 fm.")
print("The d quark is ONE order of magnitude heavier, so its excursion is ONE")
print("order shorter and ONE order smaller: 21 fm, matching the book's stated")
print("'10 fm radius maximum'.  See the erratum on the page for what the book")
print("says about the lifetime.")
print()
print("And note the trend: the heavier the fluctuation, the smaller its world.")
print("At the 1 fm scale of a hadron the vacuum is churning with pion-mass")
print("fluctuations -- which is why a hadron cannot be described as a few quarks")
print("in an otherwise empty box.")

prints

Eq. (6.66): a pair of total mass 2m lives dt < 1/2m, inside a region c*dt
e+e-        (2m = 1 MeV)   dt = 6.58e-22 s,  c dt =   197.33 fm
d dbar      (2m = 9.3 MeV) dt = 7.08e-23 s,  c dt =    21.22 fm
mu+mu-      (2m = 211 MeV) dt = 3.12e-24 s,  c dt =     0.93 fm
a pion pair (2m = 280 MeV) dt = 2.35e-24 s,  c dt =     0.70 fm

The book's electron numbers check out exactly: 6.6e-22 s and 200 fm.
The d quark is ONE order of magnitude heavier, so its excursion is ONE
order shorter and ONE order smaller: 21 fm, matching the book's stated
'10 fm radius maximum'.  See the erratum on the page for what the book
says about the lifetime.

And note the trend: the heavier the fluctuation, the smaller its world.
At the 1 fm scale of a hadron the vacuum is churning with pion-mass
fluctuations -- which is why a hadron cannot be described as a few quarks
in an otherwise empty box.

Erratum — “10 orders of magnitude less”

The book writes: “a positron–electron pair with $2m \approx 1$ MeV typically lives $\Delta t \approx 6.6\times10^{-22}$ s, hence in a region smaller than $c\Delta t \approx 200$ fm, while $d\bar d$ pairs, with masses of an order of magnitude larger, live 10 orders of magnitude less, in volumes within a 10 fm radius maximum.”

The sentence contradicts itself. Eq. (6.66) gives $\Delta t \lesssim 1/2m$ , so a mass one order larger gives a time one order shorter — and the book’s own closing clause confirms it: 200 fm ÷ 10 = 20 fm, which is the quoted “10 fm radius maximum” to within the roundings being used.

Ten orders of magnitude would give $\Delta t \approx 10^{-31}$ s and a region of $2\times10^{-8}$ fm — a hundred million times smaller than a proton, and flatly inconsistent with the radius in the same sentence.

It reads like “10 times less” expanded into “10 orders of magnitude less”. The electron numbers, and the radius, are all correct.

💡 What this really says — what “the vacuum is alive” actually buys you

It is easy to file this under quantum weirdness. It is better filed under this is why the previous three sections worked:

the screening of §5.8 and the antiscreening of §6.5 are both this, with a charge dropped in;
the 99 % of §6.7 is the energy of this medium, reorganised by the presence of three quarks;
and §6.9, below, needs the vacuum to be an active medium with a structure of its own — because the symmetry that gets broken is broken by the vacuum, not by the Lagrangian.

The one genuinely new item here is the gluon-only fluctuation. Photons cannot do this; the QED vacuum needs a charged pair to fluctuate at all. The QCD vacuum churns with pure glue, at every scale, because the gluon is both massless and self-coupled. That is the medium the next section’s symmetry breaking happens in.

§6.9 Why the pion is so light

Set the light quark masses to zero. Then chirality equals helicity (§2.9), and the QCD Lagrangian splits into a left-handed part and a right-handed part that never talk to each other. Each is separately invariant under SU(3), so the symmetry is SU(3)_L × SU(3)_R — twice as large as the flavour SU(3) of §4.6. That is chiral symmetry .

If it held exactly, the meson spectrum would contain pseudoscalar and scalar states of equal mass. It does not, and it is not close.

📐 Physics you need first — spontaneous symmetry breaking, with no field theory

The hard idea is that a symmetric law can have an asymmetric outcome, and it needs no quantum mechanics at all.

Balance a pencil on its point. The situation is perfectly symmetric under rotation about the vertical. Let go, and it falls — in some direction. The law did not pick a direction; the pencil did. Ask “why that direction?” and there is no answer, and asking reveals a misunderstanding: the symmetry is still there, in the fact that every direction was equally available.

The consequences are what matter:

The symmetry is invisible in the ground state. Look at the fallen pencil and you would never guess the law was rotationally symmetric.
There is a free direction. Sliding the fallen pencil around the circle of equivalent fallen positions costs nothing. That costless motion, in a field theory, is a massless particle — a Goldstone boson.
Tilt the table slightly and the free direction is no longer quite free. The particle acquires a small mass: a pseudo-Goldstone boson .

Two distinct kinds of breaking are in play and they must not be conflated. Explicit breaking means the law itself is not symmetric (the quark masses are not zero). Spontaneous breaking means the law is symmetric and the ground state is not (the QCD vacuum). The pion is light because the second kind dominates, and non-zero because the first exists.

a symmetric law with an asymmetric ground state

μ² = -1.00 — symmetry spontaneously brokenexplicit tilt = 0.00 — the analogue of a quark mass

ground state at φ =

-0.707

across the valley — massive

3.99

around it — the Goldstone mode

0.00

Broken — and notice the law did not change. The potential is still perfectly symmetric; the two minima prove it. But the ball must sit at one of them, and once it does, the symmetry is invisible in the ground state. Motion around the valley floor costs nothing: that flat direction is a massless Goldstone boson. Motion across it costs: that is the massive partner. In QCD the flat directions are the pseudoscalar octet and the massive one is the f₀.

Slide μ² through zero and watch the ground state stop being symmetric while the potential stays symmetric. Then add the tilt — the analogue of a non-zero quark mass — and see the flat direction acquire a small curvature. That curvature is m_π², which is why m_π² is proportional to the quark mass rather than to anything else.

🔢 Worked example — the pion against its scalar partner, and against the kaon

Against its partner. A complete SU(2)_L × SU(2)_R multiplet needs four members: the pseudoscalar isotriplet (the pions) and a scalar isosinglet. The scalar exists — it is the $f_0(1370)$ , identified in 1966 in $\bar p n \to 5\pi$ by a group including Bettini himself. Compare the squared masses:

m_\pi^2 = 0.019\ \text{GeV}^2, \qquad m_{f_0}^2 = 1.88\ \text{GeV}^2

A factor of 96 — two orders of magnitude. If chiral symmetry were exact these would be equal. The pion is not light because something made it light; it is light because it is a would-be Goldstone boson, and the $f_0$ is a perfectly ordinary hadron of ordinary mass.

Against the kaon. Because $m^2 \propto m_q$ rather than $m \propto m_q$ , the ratio of pion to kaon squared masses should track the ratio of quark masses:

\htmlClass{t-lhs}{\frac{m_\pi^2}{m_K^2}} \approx \htmlClass{t-rhs}{\frac{m_u + m_d}{m_u + m_s}}

(GMOR)

The Gell-Mann–Oakes–Renner relation as a ratio, so that every unknown constant cancels. Two measured hadron masses on the left, two quark masses on the right, and nothing to adjust in between.

Every symbol, one at a time

Hover or tap a symbol above — it lights up in the equation and its meaning, units and type appear here.

Left side: 0.0799. Right side: 0.0715. They agree to 12 % — with no free parameter. This is the Gell-Mann–Oakes–Renner relation, and it is the sharpest evidence that the pseudoscalar octet really are the Goldstone bosons of a broken chiral symmetry.

squared masses track quark masses, not masses

mpi, mK, mf0 = 0.13957, 0.49368, 1.370          # GeV
mu, md, ms = 0.00216, 0.00467, 0.0934            # MS-bar at 2 GeV, GeV

print("the pion against its chiral partner, the f0(1370):")
print(f"  m_pi^2  = {mpi**2:.5f} GeV^2")
print(f"  m_f0^2  = {mf0**2:.4f} GeV^2      the book says 'about 2 GeV^2'")
print(f"  ratio   = {mf0**2/mpi**2:.0f}x -- two orders of magnitude.")
print("  Exact chiral symmetry would make these EQUAL.")
print()
print("the Gell-Mann-Oakes-Renner relation, m^2 proportional to quark mass:")
lhs, rhs = (mpi/mK)**2, (mu+md)/(mu+ms)
print(f"  (m_pi/m_K)^2            = {lhs:.4f}")
print(f"  (m_u+m_d)/(m_u+m_s)     = {rhs:.4f}")
print(f"  agreement: {abs(lhs-rhs)/rhs*100:.0f}%, with no free parameter")
print()
print("note it is the SQUARE that is proportional to the quark mass.  A pion")
print("built the ordinary way -- mass proportional to its constituents -- would")
print("weigh about 7 MeV.  It weighs 140.  And a pion whose mass came out of the")
print("colour field like the proton's would weigh ~1 GeV.  It is neither, because")
print("it is a pseudo-Goldstone boson and obeys a different rule entirely.")

prints

the pion against its chiral partner, the f0(1370):
m_pi^2  = 0.01948 GeV^2
m_f0^2  = 1.8769 GeV^2      the book says 'about 2 GeV^2'
ratio   = 96x -- two orders of magnitude.
Exact chiral symmetry would make these EQUAL.

the Gell-Mann-Oakes-Renner relation, m^2 proportional to quark mass:
(m_pi/m_K)^2            = 0.0799
(m_u+m_d)/(m_u+m_s)     = 0.0715
agreement: 12%, with no free parameter

note it is the SQUARE that is proportional to the quark mass.  A pion
built the ordinary way -- mass proportional to its constituents -- would
weigh about 7 MeV.  It weighs 140.  And a pion whose mass came out of the
colour field like the proton's would weigh ~1 GeV.  It is neither, because
it is a pseudo-Goldstone boson and obeys a different rule entirely.

m² ∝ m_q — the pseudo-Goldstone prediction
the pseudoscalar octet
the vector mesons, for contrast — no such rule

The pseudoscalars (π, K, η) sit on a line through the origin: m² grows in proportion to the quark masses, which is what a pseudo-Goldstone boson does. The vector mesons are shown quartered so they fit on the same axes at all — the ρ and K* differ by only 15 % in mass while their quark content differs by a factor of fourteen. That is what an ordinary hadron looks like: its mass comes from the colour field (§6.7) and barely notices the quarks. The pseudoscalars are the exception, and the reason is that their mass has a completely different origin.

⚙️ Engineer’s bridge — a degenerate mode, and what lifts it

The structure here is one an engineer meets whenever a system has a continuous family of equivalent configurations.

A perfectly symmetric ring resonator has degenerate modes: rotate the excitation around the ring and nothing changes, so that motion costs no energy and shows up as a zero-frequency mode. Break the symmetry — a notch in the ring, a fabrication asymmetry — and the degeneracy lifts: the previously free direction acquires a small restoring force and the mode moves to a small but non-zero frequency, proportional to the size of the defect.

That is the whole of §6.9, term for term:

ring resonator	QCD
continuous rotational symmetry	chiral SU(3)_L × SU(3)_R
the excitation sits somewhere on the ring	the vacuum picks a chiral direction
zero-frequency mode around the ring	massless Goldstone boson
a notch in the ring	non-zero quark masses (explicit breaking)
small non-zero frequency ∝ defect	m_π² ∝ m_q

The engineering instinct transfers exactly: a soft mode is a signature of a nearly-exact symmetry, and its frequency measures how nearly. The pion is soft, so the symmetry is nearly exact, so the light quark masses are nearly zero — and the GMOR relation turns “nearly” into a number.

Where it breaks: a soft mode’s frequency measures the breaking only once you know the mode is the Goldstone, and identifying it is the whole difficulty. The pion qualifies; its chiral partner the $f_0$ does not, which is why §6.9 compares them rather than treating both as soft. The relation also has an exception that matters: the $\eta'$ is in the same nonet and is not light, because the axial anomaly gives it mass having nothing to do with quark masses at all. GMOR works to 12 % across the octet and fails completely for the ninth member — a reminder that “soft mode ⇒ nearly-exact symmetry” is a diagnosis, not a theorem.

💡 What this really says — Banks and Susskind’s picture of why the vacuum breaks it

The book gives a physical mechanism, and it is worth keeping because it connects back to §5.6’s helicity algebra.

A quark and antiquark in a pseudoscalar meson are in an S wave, so their motion is essentially an oscillation along a diameter. But at the end of each swing they would have to reverse direction — and for a massless quark that means reversing helicity, which the vertex cannot do.

So they do not turn around. The quark carries on and disappears into the sea of $q\bar q$ pairs inside the meson, while another of the opposite helicity emerges to take its place. The pseudoscalar meson is continuously exchanging its constituents with the vacuum, which is exactly why its properties are tied to the vacuum’s structure rather than to its nominal constituents.

The scalar mesons do not do this: in $J^{PC} = 0^{++}$ the pair is in a P wave, $l = 1$ , where the velocity has non-radial components and no such reversal is required. Hence they are ordinary, and heavy.

🔑 If you remember only three things

The vacuum is a medium, not an absence. Screening, antiscreening and confinement are all properties of what is present when nothing is there.
Massless was the prediction and nearly massless is the measurement. The gap between the two is exactly where the quark masses enter.
The kaon is the same argument with a heavier quark. How badly it fits says how far the approximation can be pushed before it stops being useful.

Where this goes next

§6.10 computes the whole light-hadron spectrum on a lattice, including the pion — the only method that handles a vacuum this structured.
§9.12 is the same mechanism again, applied to the electroweak gauge symmetry: the Higgs field’s potential is the widget above, and the would-be Goldstone bosons are eaten by the W and Z.
§2.9 is the chirality that becomes helicity in the massless limit — the assumption the whole section rests on.
§6.7 is the hadron mass this section’s pion conspicuously does not obey.

✅ Check yourself — the vacuum, chiral symmetry, and the pion

0/5 answered · 0 correct

1.What can the QCD vacuum do that the QED vacuum cannot?
2.In the widget, push μ² negative and the ground state stops being symmetric while the potential stays symmetric. What is the physically important consequence?
3.Why must explicit and spontaneous symmetry breaking be kept distinct?
4.The pion's squared mass is proportional to the quark mass, not its mass. Why does that distinction carry so much weight?
5.The book says a d̄d vacuum fluctuation lives '10 orders of magnitude less' than an e⁺e⁻ one. How would you check that without looking anything up?