Neutrino mass & the seesaw

The Standard Model, taken at face value, says the neutrino is exactly massless — and it says so for a good reason. A Dirac mass term stitches a left-handed field to a right-handed one, but the Model contains no right-handed neutrino, and the only mass term you could build from the left-handed field alone is a Majorana term that the electroweak gauge symmetry forbids. So masslessness is not an accident of small numbers; it is a structural prediction. Then the experiments arrived. Atmospheric and solar neutrinos change flavour in flight, and flavour change is impossible unless the mass eigenstates differ — unless neutrinos have mass. The Standard Model, in its minimal form, is simply wrong about neutrinos. What replaces it must explain not just that neutrinos are massive but that they are absurdly light: at most a few hundredths of an electron-volt, more than a million times lighter than the electron. The seesaw mechanism is the most elegant answer. Add a heavy right-handed neutrino — a gauge singlet, so its own Majorana mass $M$ can sit far above the electroweak scale — and let it talk to the ordinary neutrino through an ordinary Dirac coupling $m_D$ . Diagonalising the resulting two-state mass matrix leaves a light neutrino of mass $m_\nu\simeq m_D^2/M$ : the heavier the new scale, the lighter the observed neutrino. A Dirac mass of electroweak size and a Majorana mass near the grand-unification scale then produce exactly the sub-eV mass the oscillation experiments see. This article builds that story from the chirality of the mass term up through the seesaw matrix to the model-independent Weinberg operator — the single dimension-five window through which the Standard Model first shows its seams.

Key takeaways

In the minimal Standard Model the neutrino is massless by construction: there is no right-handed neutrino for a Dirac mass, and a Majorana mass for the left-handed field is forbidden by the gauge symmetry.
Neutrino oscillations prove neutrinos have mass — but oscillations measure only the mass-squared differences $\Delta m^2$ , not the absolute scale.
A right-handed neutrino is a gauge singlet, so it can carry a Majorana mass $M$ disconnected from the electroweak scale, alongside an ordinary Dirac mass $m_D=y_\nu v/\sqrt2$ .
Diagonalising the $2\times2$ mass matrix gives the seesaw: a light state $m_\nu\simeq m_D^2/M$ and a heavy state $\simeq M$ — pushing $M$ up drives $m_\nu$ down.
With $m_D\sim100$ GeV and $M\sim2\times10^{14}$ GeV, $m_\nu\simeq0.05$ eV — exactly the atmospheric oscillation scale, with no tiny Yukawa needed.
Model-independently, neutrino mass is the dimension-five Weinberg operator $c\,v^2/\Lambda$ — the lowest-dimension window onto physics beyond the Standard Model.
Key idea: the Standard Model forbids a neutrino mass, so any nonzero mass is new physics; the seesaw turns the neutrino's extreme lightness into evidence for a very heavy Majorana scale, since a light state $m_\nu\simeq m_D^2/M$ is small precisely because $M$ is large.

The core insight

The minimal Standard Model forbids a neutrino mass, so the measured mass is itself physics beyond it; the seesaw explains the neutrino's extreme lightness by pairing an ordinary Dirac mass $m_D$ with a very heavy Majorana mass $M$ , leaving a light state $m_\nu\simeq m_D^2/M$ — small precisely because $M$ is large.

Where this lives today

Neutrino mass is one of the most active frontiers in all of physics, precisely because it is the one laboratory-confirmed crack in the Standard Model. Oscillation experiments now map the mixing with precision: Super-Kamiokande, T2K and NOvA study atmospheric and accelerator beams, reactor experiments (Daya Bay, JUNO) pin the splittings, and the flagship DUNE and Hyper-Kamiokande programmes aim to fix the mass ordering and measure the leptonic CP-violating phase $\delta_{CP}$ — the analogue of the quark sector's CP violation, and a possible key to why the universe contains matter at all. The absolute scale is squeezed from two sides: the KATRIN experiment measures the tritium beta-decay endpoint (current bound $m_\nu<0.8$ eV, heading below $0.3$ eV), while cosmological surveys of large-scale structure constrain the sum $\sum m_\nu<0.12$ eV. The sharpest test of the seesaw's Majorana premise is neutrinoless double beta decay: experiments such as KamLAND-Zen, GERDA/LEGEND and nEXO hunt for a nuclear decay that emits two electrons and no neutrinos, violating lepton number by two units — a signal that would prove the neutrino is its own antiparticle and directly implicate a Majorana mass. And the heavy right-handed neutrinos themselves, though inaccessible at colliders if they sit near the GUT scale, may have left the deepest imprint of all: through leptogenesis, their out-of-equilibrium CP-violating decays in the early universe can generate a lepton asymmetry that electroweak sphalerons convert into the baryon asymmetry we are made of.

The physical picture

Why call it a seesaw? Because on the axis that matters — a logarithmic axis of mass — the Dirac mass sits exactly at the balance point between the two physical states. The relation $m_\nu\,M\simeq m_D^2$ means $\log m_\nu+\log M=2\log m_D$ , so $m_D$ is the geometric mean of the light neutrino and the heavy partner. Fix the pivot $m_D$ at an ordinary electroweak value and the two ends are locked in opposition: raise the heavy Majorana scale $M$ and the light neutrino mass $m_\nu$ must sink by the same factor on the other side of the plank. The first figure is that literal seesaw, drawn to scale with the numbers that reproduce the atmospheric oscillation data.

Figure

The seesaw as a lever on a logarithmic mass axis. The Dirac mass m_D sits at the geometric mean of the light and heavy states, m_D = √(m_ν M), so raising the heavy Majorana scale M drives the light neutrino mass m_ν down by the same factor. With m_D = 100 GeV and M = 2×10¹⁴ GeV the light mass is m_ν ≈ m_D²/M ≈ 0.05 eV, matching the atmospheric oscillation scale √Δm²_atm

The seesaw answers how to make a neutrino light without a fine-tuned coupling — but it starts from a puzzle that the second figure makes visceral. Lay every elementary fermion mass on one logarithmic ruler and the charged particles crowd the top few decades, from the top quark at 173 GeV down to the electron at half an MeV. Then comes a yawning gap of at least seven orders of magnitude, and only at the very bottom does the neutrino appear, near $0.05$ eV. If neutrino mass came from an ordinary Higgs Yukawa coupling like everyone else's, that coupling would have to be $y_\nu\simeq3\times10^{-13}$ — ten million times smaller than the electron's already tiny $y_e\simeq3\times10^{-6}$ . Nature could be that fine-tuned, but the seesaw offers something better: the gap is not a tiny coupling, it is a large scale.

Figure

The elementary-fermion mass ladder on a logarithmic scale, from the top quark at 173 GeV down to the neutrino near 0.05 eV. The charged fermions span the top decades; below the electron opens a gap of more than seven orders of magnitude before the neutrino. A pure Dirac mass would demand a Higgs Yukawa y_ν ≈ 3×10⁻¹³, ten million times below the electron's y_e ≈ 3×10⁻⁶ — the naturalness problem the seesaw sidesteps

The core equations

A fermion mass term couples a left-handed field to a right-handed one. The Standard Model neutrino has only a left-handed component in a weak doublet, so the first thing new physics must supply is a right-handed partner $\nu_R$ — a gauge singlet — from which an ordinary Yukawa coupling to the Higgs builds a Dirac mass when $\phi$ acquires its vacuum value $\langle\phi\rangle=v/\sqrt2$ .

\mathcal{L}_D=-y_\nu\,\overline{L}\,\tilde\phi\,\nu_R+\text{h.c.}\ \xrightarrow{\ \langle\phi\rangle\ }\ -m_D\,\overline{\nu_L}\,\nu_R+\text{h.c.},\qquad m_D=\frac{y_\nu v}{\sqrt2} \qquad (1)

where: $L$ — the left-handed lepton doublet; $\nu_R$ — a right-handed neutrino, a Standard-Model gauge singlet; $\tilde\phi=i\sigma_2\phi^*$ — the conjugate Higgs doublet; $y_\nu$ — the neutrino Yukawa coupling; $v=246$ GeV. Read it as: a Dirac neutrino is possible, but it needs a brand-new field and a Yukawa coupling of $\sim10^{-13}$ to fit the data. Because $\nu_R$ carries no gauge charge at all, nothing forbids a second, entirely different mass term built from $\nu_R$ alone — a Majorana mass that pairs the field with its own charge conjugate. This term is a gauge singlet, so its scale $M$ is a free parameter unrelated to $v$ , and it violates lepton number by two units.

\mathcal{L}_M=-\tfrac12 M\,\overline{\nu_R^{\,c}}\,\nu_R+\text{h.c.} \qquad (2)

where: $\nu_R^{\,c}=C\overline{\nu_R}^{\,\top}$ — the charge conjugate of the right-handed field; $M$ — the Majorana mass, unconstrained by the electroweak scale and typically taken very large. Read it as: this is the ingredient the Standard Model lacks and the reason a heavy scale can enter at all. With both terms present, the neutral-lepton mass Lagrangian is a quadratic form in the two-component basis $(\nu_L,\ \nu_R^{\,c})$ , and the physics lives entirely in the resulting symmetric $2\times2$ mass matrix.

\mathcal{L}=-\tfrac12\begin{pmatrix}\overline{\nu_L^{\,c}}&\overline{\nu_R}\end{pmatrix}\begin{pmatrix}0&m_D\\ m_D&M\end{pmatrix}\begin{pmatrix}\nu_L\\ \nu_R^{\,c}\end{pmatrix}+\text{h.c.} \qquad (3)

where: the top-left $0$ is the forbidden left-handed Majorana mass; the off-diagonal $m_D$ is the Dirac coupling of (1); the bottom-right $M$ is the singlet Majorana mass of (2). Read the matrix's structure — a zero on one diagonal, a huge entry on the other — as the entire content of the seesaw. Diagonalising (3) in the hierarchical limit $M\gg m_D$ produces two Majorana mass eigenstates: one nearly all $\nu_L$ and extremely light, one nearly all $\nu_R$ and extremely heavy. The light eigenvalue is the seesaw formula, and the mixing between the two states is tiny.

m_{\text{light}}\simeq\frac{m_D^2}{M},\qquad m_{\text{heavy}}\simeq M,\qquad \tan2\theta=\frac{2m_D}{M}\simeq2\theta \qquad (4)

where: $m_{\text{light}}=m_\nu$ — the observed light-neutrino mass; $m_{\text{heavy}}$ — a heavy sterile Majorana neutrino $N$ ; $\theta\simeq m_D/M$ — the active–sterile mixing angle. Read it as the payoff: the product $m_{\text{light}}\,m_{\text{heavy}}\simeq m_D^2$ is fixed at the electroweak scale, so lightness at one end demands heaviness at the other. The same physics can be written without ever naming $\nu_R$ . Integrating out any heavy states leaves a unique operator of dimension five — the lowest-dimension term the Standard Model fields admit beyond the renormalizable Lagrangian — and after electroweak breaking it is a Majorana neutrino mass suppressed by the new-physics scale $\Lambda$ .

\mathcal{L}^{(5)}=\frac{c_{ij}}{\Lambda}\,\big(\overline{L_i^{\,c}}\,\tilde\phi\big)\big(\tilde\phi^{\top}L_j\big)+\text{h.c.}\ \xrightarrow{\ \langle\phi\rangle\ }\ m_\nu=\frac{c\,v^2}{\Lambda} \qquad (5)

where: $c_{ij}$ — dimensionless coefficients; $\Lambda$ — the scale of the new physics; the operator is dimension five, hence the single power of $1/\Lambda$ . Read it as the model-independent statement: neutrino mass is the leading footprint of whatever lies beyond the Standard Model, and the seesaw is one way of generating it with $\Lambda\sim M$ . None of these masses is measured directly. What oscillation experiments measure is the interference of mass eigenstates, which depends only on the differences of squared masses — two of which are now known to good precision.

\Delta m^2_{21}\simeq7.5\times10^{-5}\ \text{eV}^2,\qquad |\Delta m^2_{31}|\simeq2.5\times10^{-3}\ \text{eV}^2 \qquad (6)

where: $\Delta m^2_{21}$ — the solar splitting; $|\Delta m^2_{31}|$ — the atmospheric splitting, whose square root $\sqrt{|\Delta m^2_{31}|}\simeq0.05$ eV sets the minimum mass of the heaviest state. Read it as the reason the absolute scale, and the Dirac-versus-Majorana question, remain open even though the splittings are pinned down.

Derivation AWhy the Standard Model forbids a neutrino mass

The goal is to show, from the gauge quantum numbers of the lepton fields alone, that the minimal Standard Model cannot give the neutrino a mass of either type: a Dirac mass is impossible because no right-handed neutrino exists, and a Majorana mass for the left-handed neutrino is forbidden because the required operator is not invariant under the electroweak gauge group. Masslessness is therefore a prediction of the Model's field content, not a tunable input, which is exactly why the measured mass counts as physics beyond it.

Step 1. Recall what a Dirac mass needs. A term $-m\,\overline{\psi_L}\psi_R+\text{h.c.}$ is the only Lorentz-invariant renormalizable mass, and it explicitly links a left-handed field to a right-handed one of the same charges. Why: chirality is the crux — a mass term flips handedness, so it can exist only if both handedness partners are present in the spectrum, which for the neutrino they are not.

Step 2. Count the neutrino's fields. In the Standard Model the neutrino appears only inside the left-handed doublet $L=(\nu_L,\ e_L)$ ; there is no $\nu_R$ . With no right-handed partner, the Dirac term $\overline{\nu_L}\nu_R$ cannot even be written down. Why: this is a statement about particle content, not dynamics — the field you would need simply is not in the theory, so the mass is zero identically.

Step 3. Try a Majorana mass from $\nu_L$ alone. One can imagine $-\tfrac12 m\,\overline{\nu_L^{\,c}}\nu_L$ , which needs no new field. But $\nu_L$ lives in a weak doublet with hypercharge $Y=-\tfrac12$ , so this bilinear transforms as a weak triplet with $Y=-1$ — it is not a gauge singlet, and no renormalizable gauge-invariant term can produce it. Why: the gauge symmetry, not Lorentz invariance, is what kills this option; only after adding a Higgs field of the right quantum numbers, or a higher-dimension operator, can the charges be soaked up.

Step 4. Conclude and locate the loophole. Both mass types are blocked in the minimal Model, so $m_\nu=0$ exactly. The two escapes are exactly the two ingredients of this article: add a singlet $\nu_R$ (opening Dirac and singlet-Majorana terms), or accept a non-renormalizable dimension-five operator that pairs two doublets with two Higgs fields to make a gauge singlet. Why: every viable model of neutrino mass is, at bottom, one of these two ways of evading Steps 2 and 3 — which is why neutrino mass is unavoidably a beyond-the-Standard-Model phenomenon. *Verified: the $\nu_L\nu_L$ bilinear carries weak isospin 1 and $Y=-1$ , forbidding a renormalizable left-handed Majorana mass; Dirac mass requires an added singlet $\nu_R$ — standard result, Thomson §13, Schwartz Ch. 29.*

Go deeper · Derivation B · 2 minDiagonalising the seesaw mass matrix

Derivation BDiagonalising the seesaw mass matrix

The aim is to take the symmetric $2\times2$ mass matrix of equation (3), with a zero upper-diagonal entry, a Dirac off-diagonal $m_D$ , and a large Majorana entry $M$ , and show by exact diagonalisation that its two eigenvalues are, in the hierarchical limit $M\gg m_D$ , a light state $m_\nu\simeq m_D^2/M$ and a heavy state $\simeq M$ , with an active–sterile mixing angle $\theta\simeq m_D/M$ . This is the mechanism in three lines of algebra, and the light eigenvalue is the entire reason neutrinos can be so light without a tiny coupling.

Step 1. Write the eigenvalue equation. For $\mathcal{M}=\left(\begin{smallmatrix}0&m_D\\ m_D&M\end{smallmatrix}\right)$ the characteristic polynomial is $\det(\mathcal{M}-\lambda\mathbb{1})=\lambda^2-M\lambda-m_D^2=0$ , so $\lambda_\pm=\tfrac12\big(M\pm\sqrt{M^2+4m_D^2}\,\big)$ . Why: the physical masses are the magnitudes of these eigenvalues, so everything reduces to expanding this square root in the small quantity $m_D/M$ .

Step 2. Expand in the hierarchy. With $M\gg m_D$ , use $\sqrt{M^2+4m_D^2}=M\sqrt{1+4m_D^2/M^2}\simeq M\big(1+2m_D^2/M^2\big)$ . Then $\lambda_+\simeq M+m_D^2/M\simeq M$ and $\lambda_-\simeq-m_D^2/M$ . Why: the whole seesaw lives in this one Taylor expansion — the large root barely shifts from $M$ , while the small root is pushed down to the tiny value $m_D^2/M$ that we identify with the observed neutrino.

Step 3. Read off the masses. The heavy eigenvalue is $m_{\text{heavy}}\simeq M$ ; the light one has magnitude $m_\nu=|\lambda_-|\simeq m_D^2/M$ (the sign of $\lambda_-$ is an unobservable Majorana phase, removed by a field redefinition). Their product is $\lambda_+\lambda_-=-m_D^2$ , so $m_\nu\,m_{\text{heavy}}\simeq m_D^2$ exactly — the seesaw relation. Why: the invariant $\det\mathcal{M}=-m_D^2$ is fixed at the electroweak scale, so the two masses are forced into inverse proportion — lightness of one is heaviness of the other.

Step 4. Find the mixing. The angle that diagonalises $\mathcal{M}$ satisfies $\tan2\theta=2m_D/(M-0)=2m_D/M$ , so $\theta\simeq m_D/M\ll1$ . The light eigenstate is almost pure $\nu_L$ (active), the heavy one almost pure $\nu_R$ (sterile). Why: the tiny mixing is what makes the heavy partner nearly invisible — it barely couples to the weak interaction, so a GUT-scale $N$ leaves essentially no low-energy trace except through the light mass it generates. *Verified: eigenvalues of $\left(\begin{smallmatrix}0&m_D\\ m_D&M\end{smallmatrix}\right)$ are $\tfrac12(M\pm\sqrt{M^2+4m_D^2})$ ; light $\simeq m_D^2/M$ , heavy $\simeq M$ , $\det=-m_D^2$ , $\tan2\theta=2m_D/M$ — reproduced numerically for $m_D=100$ GeV, $M=2\times10^{14}$ GeV giving $m_\nu=0.05$ eV.*

Go deeper · Derivation C · 2 minThe Weinberg operator: neutrino mass as the first sign of new physics

Derivation CThe Weinberg operator: neutrino mass as the first sign of new physics

The goal is to show that, without committing to any specific heavy particle, the Standard Model's own fields admit exactly one operator of dimension five, that it generates a Majorana neutrino mass $m_\nu=c\,v^2/\Lambda$ after electroweak symmetry breaking, and that integrating a heavy right-handed neutrino out of the seesaw reproduces this operator with $\Lambda\sim M$ . The point is structural: neutrino mass is the lowest-dimension effective interaction beyond the renormalizable Standard Model, so it is generically the first place new physics announces itself.

Step 1. Enumerate the operators. Organise all interactions by mass dimension: the renormalizable Standard Model stops at dimension four, and gauge and Lorentz invariance allow no dimension-five term built purely from Standard-Model fields except $(\overline{L^{\,c}}\tilde\phi)(\tilde\phi^{\top}L)$ . Why: this uniqueness is remarkable — of everything one could add, the single leading correction is precisely a neutrino Majorana mass, so light neutrino masses are the effective theory's headline prediction.

Step 2. Break the symmetry. Replace each Higgs doublet by its vacuum value $\langle\phi\rangle=v/\sqrt2$ ; the two factors of $\tilde\phi$ supply $v^2/2$ , leaving $\tfrac12(c\,v^2/\Lambda)\overline{\nu_L^{\,c}}\nu_L$ , a Majorana mass $m_\nu=c\,v^2/\Lambda$ . Why: the operator was engineered to be a gauge singlet by using two Higgs fields to cancel the hypercharge of two lepton doublets — Step 3 of the previous derivation's obstruction, now overcome at dimension five.

Step 3. Match to the seesaw. In the full theory a heavy $\nu_R$ of mass $M$ mediates this operator: exchanging it between two $L\tilde\phi$ vertices gives $c/\Lambda\sim y_\nu^2/M$ , so $m_\nu\sim y_\nu^2 v^2/(2M)=m_D^2/M$ , recovering equation (4). Why: the effective and explicit descriptions agree, confirming that the seesaw is simply one ultraviolet completion of the Weinberg operator, with the heavy scale $\Lambda$ identified as the Majorana mass $M$ .

Step 4. Read the scale. Inverting with $c\sim1$ and $m_\nu\simeq0.05$ eV gives $\Lambda=v^2/m_\nu\simeq1.2\times10^{15}$ GeV — startlingly close to the grand-unification scale. Why: the mere smallness of the neutrino mass points, through the single suppression $1/\Lambda$ , to a new-physics scale near where the gauge couplings are expected to unify — arguably the strongest quantitative hint we have of physics at that energy. *Verified: the unique dimension-five SM operator is $(\overline{L^{\,c}}\tilde\phi)(\tilde\phi^{\top}L)/\Lambda$ ; it yields $m_\nu=cv^2/\Lambda$ ; matching gives $\Lambda\sim M/y_\nu^2$ ; $\Lambda=v^2/m_\nu=1.21\times10^{15}$ GeV for $c=1$ , $m_\nu=0.05$ eV.*

Limiting cases

Case	Result	Meaning
$M\to\infty$	$m_\nu\to0$ , $N$ decouples	the heavy state leaves the theory; neutrino becomes massless — the minimal Standard Model limit
$M\gg m_D$	$m_\nu\simeq m_D^2/M$ , $\theta\simeq m_D/M$	the genuine seesaw: one very light active state, one very heavy sterile state, tiny mixing
$M\sim m_D$	two comparable Majorana states	pseudo-Dirac regime; no hierarchy, no natural suppression of $m_\nu$
$M\to0$	pure Dirac neutrino, $m_\nu=m_D$	lepton number restored; needs the unnaturally tiny Yukawa $y_\nu\sim3\times10^{-13}$
$m_D\to0$	active neutrino massless, $N$ decoupled	a sterile Majorana state with no coupling to the doublet; no seesaw
$M\sim10^{15}$ GeV, $m_D\sim100$ GeV	$m_\nu\sim0.05$ eV	GUT-scale seesaw naturally reproduces the atmospheric oscillation scale

Common trapNeutrino oscillations measure the neutrino masses

They do not. An oscillation experiment measures a phase that accumulates as different mass eigenstates propagate, and that phase depends only on the mass-squared differences $\Delta m^2_{ij}=m_i^2-m_j^2$ — never on any single mass. So all of the oscillation data, however precise, is blind to the absolute scale: the lightest neutrino could weigh $0.05$ eV or essentially nothing, and oscillations would look the same. Three separate frontiers are needed to pin the rest down. The absolute scale comes from the endpoint of tritium beta decay (KATRIN: $m_\nu<0.8$ eV) and from cosmology (the sum $\sum m_\nu<0.12$ eV from large-scale structure). The mass ordering — whether the third state is heaviest (normal) or lightest (inverted) — is being resolved by long-baseline and reactor experiments. And the Dirac-versus-Majorana question is invisible to oscillations entirely; only neutrinoless double beta decay, which requires lepton-number violation, can settle it. A related trap: the seesaw is a compelling framework, not an established fact — it accommodates the observed masses rather than predicting them, since it trades one unknown ( $m_\nu$ ) for two ( $m_D$ and $M$ ).

Worked examples

1 · The seesaw scale from the atmospheric splitting (fully worked)

Atmospheric oscillations give $|\Delta m^2_{31}|\simeq2.5\times10^{-3}$ eV², so the heaviest neutrino is at least $m_\nu\simeq\sqrt{|\Delta m^2_{31}|}\simeq0.05$ eV. Taking a Dirac mass of electroweak size, $m_D=100$ GeV (comparable to the top quark's), find the heavy Majorana scale $M$ the seesaw requires, and comment on what scale that is.

Step 1. Fix the light mass. The atmospheric splitting sets the mass of the heaviest state when the lightest is small: $m_\nu\simeq\sqrt{2.5\times10^{-3}\ \text{eV}^2}=\mathbf{0.05\ eV}$ . Convert to GeV for the seesaw formula: $m_\nu=0.05\times10^{-9}\ \text{GeV}=5\times10^{-11}$ GeV.

Step 2. Invert the seesaw. From $m_\nu\simeq m_D^2/M$ , solve for the scale: $M\simeq\dfrac{m_D^2}{m_\nu}=\dfrac{(100\ \text{GeV})^2}{5\times10^{-11}\ \text{GeV}}=\dfrac{10^4}{5\times10^{-11}}\ \text{GeV}=\mathbf{2\times10^{14}\ GeV}$ .

Step 3. Interpret. A Majorana mass of $2\times10^{14}$ GeV sits just below the canonical grand-unification scale of $\sim10^{16}$ GeV, and the corresponding Weinberg-operator scale is $\Lambda=v^2/m_\nu\simeq1.2\times10^{15}$ GeV — the same ballpark. A neutrino mass of $0.05$ eV, an ordinary Dirac coupling, and a heavy scale near GUT energies fit together with no fine-tuning: $M\simeq2\times10^{14}$ GeV. This numerical coincidence — that the observed neutrino mass points to the unification scale — is the single most cited motivation for the seesaw.

2 · The unnaturalness of a pure Dirac neutrino

Suppose instead there is no Majorana mass and the neutrino is purely Dirac, so $m_\nu=m_D=y_\nu v/\sqrt2$ . Compute the Yukawa coupling $y_\nu$ needed for $m_\nu=0.05$ eV, and compare it with the electron's Yukawa.

Step 1. Solve for the neutrino Yukawa. From $m_\nu=y_\nu v/\sqrt2$ with $v=246$ GeV, $y_\nu=\dfrac{\sqrt2\,m_\nu}{v}=\dfrac{\sqrt2\,(0.05\times10^{-9}\ \text{GeV})}{246\ \text{GeV}}=\mathbf{2.9\times10^{-13}}$ .

Step 2. Compute the electron Yukawa for comparison. With $m_e=0.511$ MeV $=0.511\times10^{-3}$ GeV, $y_e=\dfrac{\sqrt2\,m_e}{v}=\dfrac{\sqrt2\,(0.511\times10^{-3})}{246}=\mathbf{2.9\times10^{-6}}$ — itself already small, the lightest charged-fermion coupling.

Step 3. Take the ratio. $\dfrac{y_e}{y_\nu}=\dfrac{2.9\times10^{-6}}{2.9\times10^{-13}}\simeq\mathbf{10^{7}}$ . A pure-Dirac neutrino would need a coupling ten million times smaller than the electron's, spanning the same seven-order-of-magnitude gap the mass ladder shows. Nothing in the theory forbids such a coupling, but the seesaw is more satisfying: it replaces an inexplicably tiny number by a large scale, converting a naturalness embarrassment into a hint about high-energy physics.

3 · Active–sterile mixing of the heavy partner

For the seesaw of Example 1 ( $m_D=100$ GeV, $M=2\times10^{14}$ GeV), estimate the mixing angle between the light and heavy states, and comment on the heavy neutrino's detectability.

Step 1. Apply the mixing formula. From equation (4), $\theta\simeq m_D/M=100/(2\times10^{14})=\mathbf{5\times10^{-13}}$ .

Step 2. Interpret. The heavy partner is sterile to an extraordinary degree — its overlap with the weak-interacting doublet is $\sim10^{-13}$ , so its production rate through mixing is suppressed by $\theta^2\sim10^{-25}$ . A GUT-scale seesaw neutrino is inaccessible to direct production; its only imprint is the light mass it generates and, possibly, the matter–antimatter asymmetry it seeds through leptogenesis. This is why low-scale seesaws (with $M$ near the TeV scale and correspondingly small $m_D$ ) are studied — there the mixing can be large enough for collider searches.

Physics of the portrait

The seesaw has no single inventor — it surfaced in several papers around 1977–1980 — but its earliest and most complete statement was buried in a 1977 paper whose title asked a question about muon decay. The portrait is Peter Minkowski, who wrote the formula down first.

Origins · 1 minPeter Minkowski and how this was discovered

Peter Minkowski

b. 1941, Bern; the overlooked 1977 seesaw paper · the seesaw origin of light Majorana neutrino mass

In 1977 Peter Minkowski, a theorist at the University of Bern, published a paper with the unassuming title " $\mu\to e\gamma$ at a rate of one out of $10^9$ muon decays?" Inside it, almost in passing, was the first written statement of what we now call the seesaw mechanism: couple the ordinary neutrino to a heavy right-handed singlet with a large Majorana mass, and the light neutrino mass comes out as $m_D^2/M$ , naturally tiny. The paper was largely overlooked for years. The idea was rediscovered and sharpened at a cluster of 1979 workshops — by Murray Gell-Mann, Pierre Ramond and Richard Slansky in the context of grand unification, and independently by Tsutomu Yanagida — and given its enduring name, the "seesaw," for the way the light and heavy masses move in opposition. Rabindra Mohapatra and Goran Senjanović showed in 1980 how the mechanism arises organically in left–right symmetric theories, where the right-handed neutrino is not an optional extra but a required member of a right-handed doublet. The deep root reaches back to Ettore Majorana, whose 1937 "symmetric theory of the electron and positron" first showed that a neutral fermion can be its own antiparticle — the property that lets $M$ exist at all. When Super-Kamiokande proved in 1998 that neutrinos really do have mass, Minkowski's twenty-one-year-old formula was waiting.

History

1930

— Pauli postulates a neutral, nearly massless particle to save beta-decay energy conservation

1937

— Majorana shows a neutral fermion can be its own antiparticle, allowing a Majorana mass

1977

— Minkowski writes the seesaw relation m_ν ≈ m_D²/M in a paper on muon decay

1979

— Gell-Mann, Ramond, Slansky and Yanagida recast it in grand unification; Weinberg writes the dimension-five operator

1998

— Super-Kamiokande discovers atmospheric neutrino oscillations, proving neutrinos have mass

2015

— Kajita and McDonald share the Nobel Prize for the discovery of neutrino oscillations

Going further

Texts. Thomson, Modern Particle Physics, §13 (neutrino mass, oscillations and mixing) · Griffiths, Introduction to Elementary Particles, 2nd ed., §11.7 (neutrino oscillations and mass) · Halzen & Martin, Quarks and Leptons, Ch. 12–15 (weak interactions and the lepton sector) · Peskin & Schroeder, An Introduction to Quantum Field Theory, §20.2 (fermion masses beyond the Standard Model) · Schwartz, Quantum Field Theory and the Standard Model, Ch. 29 (neutrino masses, Majorana fermions, the seesaw).

Provenance. The chirality of the mass term and the Weyl/Majorana structure are recalled from aq-pp-weyl-chirality; the Higgs vacuum value $v=246$ GeV and the Yukawa mechanism $m=yv/\sqrt2$ from aq-pp-higgs-mechanism; the $SU(2)_L\times U(1)_Y$ quantum numbers that forbid the left-handed Majorana mass from aq-pp-electroweak-gauge-group. The Dirac-versus-Majorana distinction and neutrinoless double beta decay are owned by aq-pp-majorana-dirac-neutrinos and used here by name; the grand-unification scale $M\sim10^{15}$ – $10^{16}$ GeV by aq-pp-grand-unification. A verifier should confirm: (1) the $\nu_L\nu_L$ bilinear is a weak triplet with $Y=-1$ , forbidding a renormalizable left-handed Majorana mass; (2) the eigenvalues of $\left(\begin{smallmatrix}0&m_D\\ m_D&M\end{smallmatrix}\right)$ give $m_\nu\simeq m_D^2/M$ , $m_{\text{heavy}}\simeq M$ , $\theta\simeq m_D/M$ ; (3) $M=m_D^2/m_\nu=2\times10^{14}$ GeV for $m_D=100$ GeV, $m_\nu=0.05$ eV; (4) $y_\nu=\sqrt2\,m_\nu/v=2.9\times10^{-13}$ versus $y_e=2.9\times10^{-6}$ ; (5) $\Lambda=v^2/m_\nu=1.2\times10^{15}$ GeV; (6) $\sqrt{|\Delta m^2_{31}|}=0.05$ eV.

Prerequisites. The Higgs Mechanism · Weyl Spinors and Chirality · The Electroweak Gauge Group. Leads to. Majorana versus Dirac Neutrinos · Neutrino Mass Beyond the Standard Model · Grand Unification.

Ask the tutor: Ask me to diagonalise the seesaw matrix step by step for any $m_D$ and $M$ you choose, or to show why a left-handed Majorana mass is gauge-forbidden while the right-handed singlet's is not.

If the Standard Model says neutrinos are massless, how do we know they have mass?

Through oscillations. A neutrino produced as one flavour — say an electron neutrino from the Sun — can be detected some distance later as a different flavour. This flavour change is quantum interference between mass eigenstates, and it is mathematically impossible unless those eigenstates have different masses. Super-Kamiokande saw it in atmospheric neutrinos in 1998, and SNO confirmed it for solar neutrinos by 2002; the 2015 Nobel Prize recognised both. So the evidence for neutrino mass is indirect but airtight: we have never weighed a neutrino directly, yet we know at least two of them are massive because they oscillate. What oscillations cannot tell us is how massive — only the differences of squared masses, which is why the absolute scale remains an open experimental question.

What is the difference between a Dirac and a Majorana neutrino?

A Dirac neutrino is like every other fermion: the particle and antiparticle are genuinely distinct, and a mass term links a left-handed field to a separate right-handed one. A Majorana neutrino is its own antiparticle — the field and its charge conjugate are the same object — which is only possible for a truly neutral particle, and the neutrino is the only known fermion that qualifies. The distinction is not academic: a Majorana mass violates lepton number by two units, whereas a Dirac mass conserves it. The seesaw uses a Majorana mass for the heavy singlet and delivers Majorana light neutrinos. Experimentally the two cases are almost identical except in one place — neutrinoless double beta decay — which can occur only if neutrinos are Majorana. That single experiment is the cleanest way to tell them apart.

Why is it called a "seesaw"?

Because the light and heavy masses move in strict opposition around a fixed pivot. The relation $m_\nu\,M\simeq m_D^2$ says the product of the two masses is locked to the square of the Dirac mass, which sits at an ordinary electroweak scale. So if you push the heavy Majorana mass $M$ up, the light neutrino mass $m_\nu$ must come down by the same factor — one end rises as the other falls, exactly like a child's seesaw balanced on a fulcrum. On a logarithmic axis, the Dirac mass $m_D$ is the geometric mean of the two, sitting precisely at the balance point. The name captures the mechanism's whole appeal: you do not need a tiny number to make the neutrino light, you need a large number for its heavy partner, and the seesaw does the rest.

Does the seesaw predict the neutrino mass?

No — and this is an important honesty. The seesaw does not predict $m_\nu$ ; it relates it to two quantities we do not independently know, the Dirac mass $m_D$ and the Majorana scale $M$ . Give the seesaw the measured $m_\nu\simeq0.05$ eV and it tells you the combination $m_D^2/M$ , but not either one separately — the same light mass follows from $m_D=1$ GeV with $M=2\times10^{10}$ GeV, or $m_D=100$ GeV with $M=2\times10^{14}$ GeV. What the seesaw provides is not a prediction but an explanation of naturalness: it shows how a sub-eV mass can arise from perfectly ordinary couplings plus a high scale, rather than from an unnaturally tiny Yukawa. Turning it into a prediction requires an additional theory — a grand-unified or flavour model — that fixes $m_D$ and $M$ .

What is a right-handed (sterile) neutrino, and why haven't we seen it?

It is the missing partner the Standard Model never included: a neutrino field that is right-handed and carries no gauge charge — no electric charge, no weak isospin, no colour. Because it is a complete gauge singlet, it does not couple to the $W$ , $Z$ , photon or gluons at all; the only way it interacts is through its tiny Dirac mixing $\theta\simeq m_D/M$ with the ordinary neutrino, hence the name "sterile." In a GUT-scale seesaw that mixing is around $10^{-13}$ , so the heavy partner is produced with a probability of $\sim10^{-25}$ — utterly hopeless at any conceivable collider. Lighter sterile neutrinos, at the keV or GeV scale, are actively searched for as dark-matter candidates and in beam-dump and reactor experiments, and would have larger, potentially observable mixings. So "haven't seen it" depends entirely on the assumed scale.

What would neutrinoless double beta decay prove?

Ordinary double beta decay — two neutrons converting to two protons, emitting two electrons and two antineutrinos — is a rare but observed process in certain nuclei. Its neutrinoless cousin emits the two electrons but no neutrinos: the two would-be antineutrinos annihilate each other, which is possible only if the neutrino is its own antiparticle, i.e. Majorana. Observing it would therefore prove two things at once: that lepton number is violated by two units, and that neutrinos have a Majorana mass — a cornerstone of the seesaw. The decay rate is proportional to the square of an effective Majorana mass, $m_{\beta\beta}=|\sum_i U_{ei}^2 m_i|$ , so a measurement would also constrain the absolute mass scale and the Majorana phases. Experiments like KamLAND-Zen and LEGEND are pushing the half-life bound past $10^{26}$ years in the hunt for it.

How does neutrino mass connect to the matter–antimatter asymmetry?

Through leptogenesis, one of the most attractive spin-offs of the seesaw. The heavy right-handed neutrinos $N$ are Majorana, so they have no conserved lepton number, and their Yukawa couplings can carry CP-violating phases. In the hot early universe these $N$ decay out of thermal equilibrium, and CP violation makes them produce slightly more leptons than antileptons — a net lepton asymmetry. Electroweak "sphaleron" processes, which are active at high temperature, then convert part of that lepton asymmetry into a baryon asymmetry. The upshot is that the same heavy Majorana neutrinos invoked to explain why the light neutrinos are so tiny can also explain why the universe is made of matter rather than nothing. The observed neutrino mass scale and the observed baryon asymmetry are, remarkably, compatible with a single high-scale seesaw — a connection that keeps leptogenesis a leading explanation of our existence.