The Missing Object: How the Dirichlet-to-Neumann Operator Closes the Spectral Action Programme

Solving the Yukawa coupling problem and the fermion mass problem in the Chamseddine–Connes noncommutative geometry Standard Model, the first spectral framework to determine all nine fermion masses from first principles.

The 28-Year Gap Is Closed

Since 1996, the Chamseddine–Connes spectral action has been one of the most mathematically elegant approaches to fundamental physics. It derives the entire bosonic Standard Model, Einstein gravity, Yang–Mills gauge fields, the Higgs mechanism from a single principle: the action is a spectral invariant of the Dirac operator on a product geometry.

But for 28 years, it has had the same open problem known variously as the Yukawa coupling problem, the fermion mass problem, or simply the undetermined internal Dirac operator. The bosonic sector is derived. The fermionic sector is not. The nine fermion masses, the mixing angles, the CP phase, the entire Yukawa sector go in by hand through an unexplained choice of the finite Dirac operator $D_F$. As Barrett noted in his 2006 analysis of the noncommutative Standard Model: the Connes–Chamseddine class of Dirac operators "requires an unexplained choice of $D_F$ which is rather special" and "there are no constraints on the parameters of the fermion mass matrix." Every paper in the spectral Standard Model programme since Connes' original work has either worked around this or acknowledged it as the central remaining gap. It has been the single most repeated criticism of the spectral action approach: "It beautifully derives gravity and gauge fields, but fermion masses are still free parameters."

That gap is now closed.

The missing object is the Dirichlet-to-Neumann operator $K_{\mathrm{OS}}$ (also known in the computational and Russian literature as the Poincaré–Steklov operator, and in medical imaging as the voltage-to-current map) the Osterwalder–Schrader boundary operator of the Standard Model spectral triple, constructed as the DN map of $D_A^2$ on a collar boundary. Coupled to the spectral action through a scale-by-scale capacity inequality, this single operator:

• Forces exactly three fermion generations, not two, not four, but three, from a double-squeeze mechanism
• Determines all nine charged fermion masses uniquely on a fixed carrier, via KKT saturation equations with no free parameters in the mass sector
• Predicts $\sin^2\theta_W(M_Z) = 0.2316$ (0.2% accuracy) from the gauge trace factors $(k_1, k_2, k_3) = (20, 12, 12)$
• Excludes the strong CP phase ($\theta = 0$, structurally, no axion required)
• Yields $\alpha^{-1} = 137$ through the E9** normalization bridge, with zero free parameters on the adopted $S^3$ carrier

The object itself is not exotic. It is the standard Dirichlet-to-Neumann map, one of the most studied operators in spectral geometry and inverse problems applied to the specific geometric setting of the Standard Model spectral triple with a collar boundary. It was always available. It just needed to be identified, constructed, and connected to the spectral action through the right coupling principle.

This is, to our knowledge, the first time a spectral framework has determined the fermion mass sector from first principles rather than inputting it by hand.

1. The Gap in the Spectral Action

The spectral action programme, also called the noncommutative geometry Standard Model, the Connes–Chamseddine model, or the spectral Standard Model, is built on a single principle due to Chamseddine and Connes (1997): the bosonic action of a physical theory is

where $D$ is the Dirac operator on a product geometry $M_4 \times F$, $\Lambda$ is a cutoff scale, and $f$ is a positive test function. The heat-kernel expansion gives:

where the Seeley–DeWitt coefficients $a_{2k}$ encode the geometric and gauge content. For the Standard Model spectral triple with algebra $\mathcal{A}_F = \mathbb{C} \oplus \mathbb{H} \oplus M_3(\mathbb{C})$, this expansion recovers:

• The Einstein–Hilbert action (from $a_2$)
• The Yang–Mills action for $\mathrm{SU}(3) \times \mathrm{SU}(2) \times \mathrm{U}(1)$ (from $a_4$)
• The Higgs potential with the correct quartic structure (from $a_4$)
• The minimal coupling of fermions to gauge fields (from the fermionic action $\langle \psi, D\psi \rangle$)

This is a remarkable achievement. The entire bosonic Standard Model Lagrangian emerges from the spectrum of a single operator.

But the Yukawa couplings are free. The internal Dirac operator $D_F$ on the finite space $F$ encodes the fermion mass matrices. Its entries are the Yukawa coupling constants. Nothing in the spectral action principle constrains these entries. They appear as free parameters in $D_F$, exactly as they appear as free parameters in the conventional Standard Model. As the Wikipedia article on Yukawa couplings states: "The Yukawa coupling for any given fermion in the Standard Model is an input to the theory. The ultimate reason for these couplings is not known: it would be something that a better, deeper theory should explain."

The spectral action programme reduces the problem from "why these gauge groups and this Lagrangian?" (which it answers) to "why these Yukawa couplings?" (which it does not). This is the fermion mass problem in noncommutative geometry. The question is what additional structure could close the gap without destroying the elegance of the spectral approach.

2. What Was Missing

The gap exists because the spectral action uses only the bulk spectral data of the Dirac operator, its eigenvalues, their multiplicities, and their heat-kernel asymptotics. This bulk data determines the gauge sector (through the representation content of $\mathcal{A}_F$) but not the Yukawa sector (which lives in the off-diagonal entries of $D_F$).

What was missing was a boundary channel, an independent spectral object that carries information about the Yukawa sector and provides constraints that the bulk spectral action cannot supply.

In retrospect, the clue was always present. In any well-posed elliptic boundary problem, the interior is constrained by its boundary data through a specific operator: the Dirichlet-to-Neumann map. The DN map is the operator that takes Dirichlet boundary data (the values of a function on the boundary) to Neumann boundary data (the normal derivative on the boundary). It encodes how the interior communicates with the exterior, how much information can pass through the boundary.

The DN map is one of the most studied objects in spectral geometry. It appears in:

• Calderón's inverse problem: can you determine the conductivity of a body from boundary voltage-to-current measurements?
• Electrical impedance tomography: medical imaging using boundary measurements
• Scattering theory: the S-matrix is a dressed version of the DN map
• AdS/CFT: the boundary-to-bulk propagator encodes holographic information

In the Euclidean quantum field theory setting, the DN map on a time-zero boundary is the Osterwalder–Schrader boundary operator, the positive operator whose positivity guarantees that the Euclidean theory can be analytically continued to a physical Lorentzian theory with a positive-definite Hilbert space.

The missing object was there all along. It just needed to be constructed explicitly for the Standard Model spectral triple and connected to the spectral action through the right coupling.

3. The Construction of $K_{\mathrm{OS}}$

The collar geometry

Consider the product geometry $M_4 \times F$ with a collar boundary. The collar is a tubular neighbourhood $\Sigma \times [0, \beta]$ where $\Sigma = S^3$ is the spatial boundary and $\beta$ is the collar thickness. The squared Dirac operator $D_A^2$ on the collar takes the form:

where $r$ is the collar coordinate and $D_\Sigma$ is the Dirac operator on the boundary.

The DN map

The Dirichlet-to-Neumann operator for $D_A^2$ on the collar boundary is defined as follows. Given Dirichlet data $\phi$ on $\Sigma$, solve $D_A^2 u = 0$ in the collar with $u|_\Sigma = \phi$. The DN map is:

For the Standard Model spectral triple on a collar of thickness $\beta$, the DN operator has eigenvalues:

where $m_\alpha$ are the nine charged fermion masses (the eigenvalues of the internal Dirac operator $D_F$) and the $\tanh$ factor encodes the collar geometry.

Operator properties

$K_{\mathrm{OS}}$ is:

• Positive: $K_{\mathrm{OS}} \geq 0$ on the dissipative subspace $H_{\mathrm{diss}} = \ker(K_{\mathrm{OS}})^\perp$
• Selfadjoint: as a classical elliptic pseudodifferential operator of order 1
• OS-reconstructible: its positivity is precisely the Osterwalder–Schrader reflection positivity condition needed for physical Hilbert space construction
• Spectrally determined: its eigenvalues are smooth functionals of the gauge connection and the fermion masses

The internal spectrum consists of nine eigenvalues (one per charged fermion) with sector multiplicities: quarks carry weight $w = 6$ (3 colours $\times$ 2 chiralities) and leptons carry weight $w = 2$ (1 colour $\times$ 2 chiralities), giving a total dissipative channel count $N_{\mathrm{diss}} = 84$.

4. What the Missing Object Does

The capacity inequality

$K_{\mathrm{OS}}$ enters the framework through the capacity inequality: a scale-by-scale constraint coupling the spectral content of the Dirac operator to the dissipative budget provided by $K_{\mathrm{OS}}$.

Define the $\Lambda$-budget:

where $\Lambda_T(K) = (1 - e^{-TK})/(TK)$ is the Fejér operator and $q(x) = -\log((1-e^{-x})/x)$ is the entropy kernel. This budget measures the total dissipative capacity available at proper-time scale $T$.

The capacity inequality states:

The Dirac-side spectral content cannot exceed the boundary dissipative budget at any scale. Flow cannot exceed Anchor.

What follows

Once $K_{\mathrm{OS}}$ is identified and the capacity inequality is imposed, the consequences are immediate and far-reaching:

Three generations forced. The capacity inequality, combined with CP capacity and the double-squeeze mechanism, proves that exactly $N = 3$ generations of fermions are admissible. Not 2 (excluded by CP capacity), not 4 or more (excluded by the double squeeze). This is an unconditional theorem, it does not depend on the carrier geometry.

Mass determination. At the proper-time scales where the capacity inequality saturates ($S_\Omega = D_T$), the KKT stationarity conditions produce the saturation equations that determine all nine fermion masses. On the adopted $S^3$ carrier with dominance margin $\Delta > 1$, the mass vector is globally unique, there are no spurious solutions.

Gauge trace factors. The representation content of $K_{\mathrm{OS}}$ restricted to the algebra $\mathcal{A}_F = \mathbb{C} \oplus \mathbb{H} \oplus M_3(\mathbb{C})$ determines the gauge trace factors $(k_1, k_2, k_3) = (20, 12, 12)$. The ratio $k_1/k_2 = 5/3$ is the structural input for the electroweak normalization, yielding the zero-parameter prediction $\sin^2\theta_W(M_Z) = 0.2316$ (0.2% accuracy).

Strong CP exclusion. The positivity $K_{\mathrm{OS}} > 0$ on $H_{\mathrm{diss}}$ is precisely the Osterwalder–Schrader reflection positivity condition. For $\theta \neq 0$, the Euclidean measure $e^{-S_{\mathrm{YM}} - i\theta Q_{\mathrm{top}}}$ is complex-valued and OS positivity is obstructed. Therefore $\theta = 0$ - structurally, not by fine-tuning, and without an axion.

The normalization bridge. The E9** equation connects the collar DN spectrum to the absolute gauge coupling:

where $B_{\mathrm{ren}}$ is the renormalized boundary functional, a convergent shell sum over $S^3$ eigenvalues weighted by the DN eigenvalues of $K_{\mathrm{OS}}$. With the collar width $\beta$ determined by the upstream Selector 9 chain, this yields $\alpha^{-1} = 137$ with zero free parameters on the adopted $S^3$ carrier.

5. Why It Was Missing

The DN operator was not hidden. It was not technically inaccessible. It was missing from the spectral action programme for a structural reason: the spectral action principle as formulated by Chamseddine and Connes is a bulk principle. It computes the action from the spectrum of $D^2$ on the interior. It does not incorporate boundary spectral data.

This is not a criticism of the Chamseddine–Connes programme, it's a precise diagnosis of where the programme stops. The spectral action gives you the Lagrangian. The DN operator gives you the constraints on the couplings. You need both.

The specific reasons $K_{\mathrm{OS}}$ was not identified earlier:

The collar was not part of the standard setup. The Chamseddine–Connes spectral triple is formulated on a closed manifold $M_4 \times F$ without boundary. Introducing a collar boundary and asking "what is the DN map?" requires a different geometric starting point, a manifold with boundary, with the collar as a tubular neighbourhood. This is natural in the Euclidean QFT setting (where the time-zero boundary is fundamental to OS reconstruction) but was not part of the standard NCG formulation.

The capacity inequality is a new coupling. Even if you construct $K_{\mathrm{OS}}$, you need a principle telling you how it constrains the spectral action. The capacity inequality $S_\Omega \leq D_T$ is that principle. It says the Dirac spectral content is bounded by the DN dissipative budget. This is a new structural ingredient, it does not follow from the spectral action principle alone.

The connection to OS reconstruction was not exploited. The fact that the DN map is the OS boundary operator, and therefore that its positivity is equivalent to the existence of a physical quantum theory was known in the axiomatic QFT literature but had not been connected to the spectral action programme. Making this connection is what identifies $K_{\mathrm{OS}}$ as the unique candidate for the missing boundary operator.

6. The Uniqueness of $K_{\mathrm{OS}}$

The boundary operator is not just identified, it's proved unique.

$\Lambda$-budget uniqueness. The budget $D_T(K_{\mathrm{OS}})$ is the unique unbounded scale-dependent spectral invariant of the dissipative spectrum satisfying spectral invariance, Fejér compatibility, additivity, and correct UV asymptotics. Any admissible alternative $D_T'$ satisfies $|D_T'(K) - D_T(K)| \leq C$ for a constant $C$ independent of $T$.

OS forcing. Among positive selfadjoint operators on the boundary Hilbert space, $K_{\mathrm{OS}}$ is the unique operator that is simultaneously: (i) the DN map of the squared Dirac operator on the collar, (ii) OS-positive (guaranteeing reflection positivity), and (iii) spectrally compatible with the Standard Model fermion content.

The missing object is not one among many candidates. It is the unique operator satisfying the structural requirements.

7. Structural Boundaries

The identification of $K_{\mathrm{OS}}$ also reveals what the boundary operator cannot do. These are structural boundaries, proved limitations, not gaps.

Exact Yukawa ratios are not law-level outputs. On a fixed carrier, the nine fermion masses are uniquely determined by the KKT saturation equations. But the exact mass ratios depend on the geometric deficit $G_M(T)$ the carrier-specific spectral data, and are therefore carrier-specific, not universal. This is proved by independent no-go theorems: within the E1–E8 system (G₃-orbit freedom on the physical Yukawa quotient manifold) and at the Tier-0 lawhood level (across-the-board no-go under minimal lawhood axioms). The mass hierarchy is law-level; the exact ratios are not.

The strong coupling $\alpha_s$ is structurally inaccessible. The gauge trace factors $(k_1, k_2, k_3) = (20, 12, 12)$ determine $k_1/k_2$ (giving $\sin^2\theta_W$) but the strong coupling $g_3$ requires the absolute normalisation of $k_3$, which is blocked by two independent obstructions (Gauge Coupling Classification Theorem, GC-1/GC-2).

These boundaries are features, not bugs. A framework that honestly identifies its own limits is more credible than one that claims to predict everything.

8. Connection to Known Physics

$K_{\mathrm{OS}}$ connects the spectral action programme to several well-established areas of physics and mathematics:

Calderón's inverse problem. The question "can you hear the shape of a drum from its boundary measurements?" is the DN inverse problem. $K_{\mathrm{OS}}$ makes this question concrete for the Standard Model: the fermion masses are the interior data, and the DN eigenvalues $\kappa_\alpha = m_\alpha \tanh(m_\alpha \beta)$ are the boundary measurements. The capacity inequality then says that the interior spectral content (the "shape") is constrained by these boundary measurements.

Electrical impedance tomography. In medical imaging, the DN map (also called the voltage-to-current map or the Poincaré–Steklov operator) encodes how internal conductivity affects boundary measurements. The mathematical structure is identical: an elliptic interior operator, a boundary operator encoding the interior-to-boundary information transfer, and an inverse problem relating the two.

Holography. In the AdS/CFT correspondence, the boundary operator encodes how bulk information is accessible at the boundary. $K_{\mathrm{OS}}$ plays a structurally analogous role: it encodes how the bulk Dirac spectral content (the "interior" gauge and Yukawa structure) constrains and is constrained by the boundary dissipative budget.

OS reconstruction. In axiomatic quantum field theory, the Osterwalder–Schrader programme reconstructs a physical Lorentzian theory from Euclidean data by requiring reflection positivity. The positivity of $K_{\mathrm{OS}}$ is precisely this condition. The identification of the DN map as the OS boundary operator makes the reconstruction programme concrete and computable for the Standard Model.

9. Summary

The spectral action programme of Chamseddine and Connes derives the bosonic Standard Model from the spectrum of a Dirac operator. It leaves the Yukawa couplings, the fermion masses as free parameters.

The missing object is the Dirichlet-to-Neumann operator $K_{\mathrm{OS}}$, constructed as the DN map of the squared Dirac operator on a collar boundary of the product geometry $M_4 \times F$. It is the Osterwalder–Schrader boundary operator of the Standard Model spectral triple.

Through the capacity inequality $S_\Omega \leq D_T(K_{\mathrm{OS}})$, this single operator:

• Forces exactly three fermion generations
• Determines all nine fermion masses on a fixed carrier
• Provides the gauge trace factors supporting $\sin^2\theta_W = 0.2316$
• Excludes the strong CP phase ($\theta = 0$, no axion)
• Supplies the normalization bridge yielding $\alpha^{-1} = 137$

The object was not exotic or speculative. It was the standard Dirichlet-to-Neumann map, one of the most studied operators in spectral geometry applied to the specific geometric setting of the Standard Model spectral triple with a collar boundary. It was always available. It just needed to be identified, constructed, and connected to the spectral action through the capacity inequality.

The Yukawa sector is no longer free. The missing object closes the programme.

Author: Jeremy Rodgers Framework: The Everything Equation Status: March 2026 Technical paper: The Missing Object: $K_{\mathrm{OS}}$ as the OS Boundary Operator of the SM Spectral Triple - see the papers section for the full construction, uniqueness proofs, and spectral computation.

© 2026 Jeremy Rodgers. All rights reserved. Content released under CC BY-NC-ND 4.0 unless otherwise stated.

Related historical papers

• The Coupled Dirac–Λ Dynamical System: Unified Operator Equations for a Capacity-Constrained Spectral Action Framework →
• Generation Forcing in the Capacity-Constrained Dirac–Lambda Framework: The Capacity Box Construction →
• KOS as the Osterwalder–Schrader Boundary Operator of the Standard Model Spectral Triple →
• The Fermion Mass Prediction Problem in the Coupled Dirac–Lambda Framework: Balance Equations and Multi-Scale KKT Forcing →
• Structural Closure of the Coupled Dirac–Lambda Framework: Global Mass Determination and Scheme Rigidity →
• Geometric Fixed-Point Existence and Spectral Rigidity in the Coupled Dirac–Lambda System →
• Determinant-Closed Unification from a Capacity-Constrained Dirac–Λ System: A Record-Admissible Forcing of the Standard Model →
• Determinant-Constrained Forcing of the Standard Model from a Capacity-Coupled Dirac–Λ System →
• Structural Vanishing of the Strong CP Phase in the Coupled Dirac–Λ Framework: A Record-Admissibility and Positivity-Based Resolution →
• No-Go Theorem for Exact Yukawa Prediction in the Capacity-Coupled Dirac–Lambda Framework →