Spin-Statistics Theorem

The theorem is under permanent experimental audit, because any violation would falsify quantum field theory itself. The cleanest tests are statistics tests: spectroscopy on ⁴He and ³He — one a boson, one a fermion, their wildly different low-temperature fates dissected in Bosons and Fermions — confirms the composite rule; searches for Pauli-forbidden atomic transitions (VIP-2 at Gran Sasso listens for X-rays from copper electrons dropping into filled shells) keep returning null. On the theory side, the dimensional fine print became a field: in two-dimensional electron gases the braid-group loophole is real, and the fractional-quantum-Hall anyons that exploit it were directly braided in 2020-era interferometry, with topological quantum computing betting on their non-Abelian cousins. The $(-1)^{2j}$ ingredient itself is routine engineering — every qubit lab that drives a $2\pi$ rotation and watches the sign flip re-measures the theorem's key input (the fringe experiment lives in Rotation Operators and SU(2)).

Picture a quantum field as a machine distributed over space that can create or destroy one particle at each point. Relativity imposes a non-negotiable discipline: an intervention here must not disturb any measurement there if "there" lies outside the light cone. In operator language, observables built from the field at spacelike-separated points must commute — and whether they do depends on a single sign choice at quantization (commutators or anticommutators), which the field's spin decides.

Think of the two failure modes as two cliffs with one ridge between them. Cliff one is energy: quantize the wrong way and filling certain modes lowers the total energy without limit — no ground state to stand on. Cliff two is causality: the field's correlations refuse to vanish outside the light cone. Integer spin falls off cliff one under anticommutators; half-integer spin falls off cliff two (and one) under commutators. The single consistent path is the theorem, and the 2×2 audit is Derivation B below. The next figure is the article's punchline, and it is fully quantitative. For a free scalar field of mass $m$ we can compute both candidate equal-time correlation structures in closed form. The commutator function $\Delta(r,0)$ — the one Bose quantization actually uses — vanishes identically at any spatial separation $r>0$: we prove the zero symbolically below. The anticommutator function $\Delta_1(r,0)$ — the structure a hypothetical Fermi scalar would be built on — comes out as $mK_1(mr)/2\pi^2 r$: exponentially small beyond the Compton wavelength, but never zero. The wrong choice does not fail loudly; it fails by a whisper that never dies, and a whisper across the light cone is enough to wreck relativistic causality.

Crucially, the theorem is relativistic. Nonrelativistic quantum mechanics is perfectly consistent with either statistics for either spin — Permutation Symmetry and the Exchange Operator built both sectors without ever asking about spin. Experiment picks the pairing, and only relativistic field theory explains the pick. The rule you have been using all chapter is the shadow of a light-cone argument.

Before any equation, fix what is being claimed. The theorem is a statement about fields — the operator machines of relativistic quantum field theory — and every hypothesis is load-bearing. In a quantum field theory with (i) Lorentz-invariant dynamics, (ii) microcausality — observables built from fields at spacelike-separated points are compatible, (iii) a Hamiltonian bounded below with a stable vacuum, and (iv) a positive-norm state space, the following pairing is forced:

where: $s$ — the spin of the field, i.e. which representation of the rotation (Lorentz) group it transforms under; Bose/Fermi — whether mode operators obey $[a,a^\dagger]=1$ or $\{a,a^\dagger\}=1$, exactly the symmetric/antisymmetric sector split of Symmetric and Antisymmetric States. Each hypothesis has a job: drop Lorentz invariance and nonrelativistic mechanics happily hosts symmetric electrons; drop microcausality and the Fermi scalar below becomes legal; drop the energy floor and the Bose–Dirac disaster is admissible; drop positive norms and ghosts fake any statistics. In two spatial dimensions the exchange topology itself changes and the conclusion genuinely relaxes — that loophole is Anyons and Fractional Statistics' story.

The cleanest laboratory is the free real scalar field of mass $m$ — spin 0, no indices. Expand it in modes (the standard construction, quoted as the starting point, $\hbar=c=1$):

where: $a_{\mathbf p},a_{\mathbf p}^\dagger$ — annihilation/creation operators for momentum $\mathbf p$; $p\cdot x = E_p t-\mathbf p\cdot\mathbf x$ — the Minkowski phase. Physically: $\phi(x)$ superposes "destroy a particle here" and "create a particle here" — the combination relativity itself forces, since a purely positive-frequency field cannot be local. Everything about statistics now sits in one question: does the field-pair bracket hand you the difference or the sum of the two frequency exponentials? Bose commutators give the difference — the commutator function; Fermi anticommutators give the sum. At equal times ($t_x=t_y$, spatial separation $r$) they are:

where: $\Delta$ — the structure $[\phi(x),\phi(y)]$ delivers under Bose quantization; $\Delta_1$ — the structure $\{\phi(x),\phi(y)\}$ delivers under the (hypothetical) Fermi quantization; equal times is the most spacelike slice, so it is the decisive test. Read the sign: the difference can cancel, the sum cannot. That one sign is the entire causality half of the theorem for spin 0, and Derivation A does both integrals exactly. The sum evaluates in closed form to $\Delta_1(r,0)=mK_1(mr)/2\pi^2 r$ (Derivation A, verified symbolically and numerically), with $K_1$ the modified Bessel function and $mr=r/\lambda\!\!\bar{}_C$ — the punchline curve of the figure above. It is positive for every $r$: exponentially suppressed beyond the Compton wavelength ($\sim e^{-mr}$), yet never zero. A Fermi-quantized scalar would carry irreducible operator structure connecting spacelike-separated points — measurements outside each other's light cones could not be independently prescribed.

The second cliff needs even less machinery: a single (particle, antiparticle) mode pair of energy $E$. Write the raw, unordered Hamiltonian each field theory delivers before any statistics choice is imposed. For the scalar, the energy density is a sum of squares, so both sectors enter positively; for the Dirac field, the antiparticle spin-sum flips the sign of its term — we quote that sign with its source (the $v$-spinor completeness relation; Peskin & Schroeder §3.5) rather than re-deriving spinor algebra here:

where: $a$ — particle mode, $b$ — antiparticle mode; the ordering $b\,b^\dagger$ is what the mode expansion literally produces. Everything now hinges on the reordering rule the statistics choice supplies: Bose gives $b\,b^\dagger=b^\dagger b+1$, Fermi gives $b\,b^\dagger=1-b^\dagger b$. Substituting each rule into each raw Hamiltonian generates the four-entry audit of the second figure — two constants shift harmlessly, one spectrum keeps a floor, and two lose it. Derivation B runs the table exactly. The mode-level table is exact and honest about its scope: it demonstrates the mechanism by which the wrong pairing destroys the energy floor, mode by mode. The general theorem — that no rescue is possible for any interacting, any-spin variant — is the axiomatic result of Fierz, Pauli, and the 1950s formalization; we cite, not derive.

The two derivations settle the physics; one line settles the bookkeeping that makes the theorem matter for this chapter. Fermionic creation operators anticommute, $\{\hat a_i^\dagger,\hat a_j^\dagger\}=0$, so two-particle states are automatically antisymmetric, $\hat a_i^\dagger\hat a_j^\dagger|0\rangle=-\hat a_j^\dagger\hat a_i^\dagger|0\rangle$, and coincident labels give $(\hat a_i^\dagger)^2=0$ — no double occupancy, the seed Pauli Exclusion Principle grows into the periodic table and the white dwarf. Bosonic operators commute and populate the symmetric sector the same way. The relativistic sign choice thereby selects between the two sectors that Permutation Symmetry and the Exchange Operator proved were the only options — it never re-derives them, and the nonrelativistic postulate (fermions antisymmetric, bosons symmetric) is inherited, not assumed. (Verified: $(a^\dagger)^2=0$ and $n^2=n$ are exact matrix identities in the verify script.)

Example 1 — how big is the causality leak of a Fermi-quantized electron-mass scalar? Task: a hypothetical spin-0 field with the electron's mass is quantized with anticommutators. Quantify $\Delta_1(r,0)$ at 1, 2 and 5 reduced Compton wavelengths ($\lambda\!\!\bar{}_C=\hbar/m_ec=0.386\,\mathrm{pm}$), in units of $m^2$.

Step 1. Enumerate. The closed form is $\Delta_1(n\lambda\!\!\bar{}_C,0)=m^2K_1(n)/2\pi^2n$; everything reduces to Bessel values $K_1(1)=0.60191$, $K_1(2)=0.13987$, $K_1(5)=4.0446\times10^{-3}$.

Step 2. Apply. At $n=1$: $\Delta_1=0.60191/(2\pi^2)=\mathbf{0.030493\,m^2}$. At $n=2$: $0.13987/(4\pi^2)=\mathbf{0.0035430\,m^2}$. At $n=5$: $4.0446\times10^{-3}/(10\pi^2)=\mathbf{4.098\times10^{-5}\,m^2}$.

Step 3. Boundary number — the suppression ladder. Relative to one Compton wavelength, the leak falls to $11.6\%$ at $2\lambda\!\!\bar{}_C$ and $0.134\%$ at $5\lambda\!\!\bar{}_C$. At atomic distances ($r\approx137\,\lambda\!\!\bar{}_C$) the leak is $\sim e^{-137}$: hopeless to measure, yet nonzero — and a single nonzero value anywhere outside the cone already contradicts the causality axiom. The wrong theory is not approximately right; it is exactly wrong.

Example 2 — the energy books for one mode pair, all four ways. Task: a particle–antiparticle mode pair has $E=1$ (units of $mc^2$). Tabulate the exact spectra of the four spin-statistics pairings and identify each verdict.

Step 1. Enumerate the raw Hamiltonians and reordering rules: $H^{\text{raw}}_{\text{scalar}}=E(a^\dagger a+bb^\dagger)$, $H^{\text{raw}}_{\text{Dirac}}=E(a^\dagger a-bb^\dagger)$; Bose $bb^\dagger=n_b+1$, Fermi $bb^\dagger=1-n_b$.

Step 2. Apply each rule to each raw form:

Step 3. Boundary number. For the wrong-Dirac case the floor at bosonic occupation $n_b=20$ is already $\mathbf{-21E}$; nothing stops $n_b=10^6$. For the wrong-scalar case, two modes with $E_1=1,E_2=3$ sit $\mathbf{4E}$ below their own $b$-vacuum — and the real field has a continuum of modes. Both wrong cells have no ground state; both right cells do.

References. Pauli, "The Connection Between Spin and Statistics", Phys. Rev. 58, 716 (1940) · Fierz, Helv. Phys. Acta 12, 3 (1939) · Streater & Wightman, PCT, Spin and Statistics, and All That (1964) · Weinberg, The Quantum Theory of Fields Vol. I, Ch. 5 · Peskin & Schroeder, An Introduction to Quantum Field Theory, Sec. 3.5 · Duck & Sudarshan, Pauli and the Spin-Statistics Theorem (1997) · Griffiths, Introduction to Quantum Mechanics, Ch. 5 · Sakurai & Napolitano, Modern Quantum Mechanics, Ch. 7.

Provenance. Field expansions, (anti)commutators, and spin sums follow Peskin & Schroeder Sec. 3.5 and Weinberg QFT Vol. I Ch. 5. A verifier should confirm: the scalar commutator $\Delta(x-y)$ vanishes for spacelike separation while the anticommutator does not; the Dirac spin sums $\sum u\bar u=\cancel{p}+m$, $\sum v\bar v=\cancel{p}-m$; and the energy-positivity argument requiring anticommutators for spin-1/2. Historical attribution to Fierz (1939) and Pauli (Phys. Rev. 58, 716, 1940) per Duck & Sudarshan (1997). Nonrelativistic consequences cross-checked against Griffiths Ch. 5 and Sakurai Ch. 7. (This lane's verify script executes the first and third confirmations exactly — the symbolic spacelike zero, the K₁ closed form to 3×10⁻⁹, and the four mode-level spectra; the spinor completeness sums are quoted with their source above.)

Builds on: Indistinguishability of Identical Particles · Permutation Symmetry and the Exchange Operator · Symmetric and Antisymmetric States · Bosons and Fermions. Leads to: Pauli Exclusion Principle · Slater Determinants · Aufbau Principle and Electron Configurations · The Helium Atom.