14.  Fermion Content: Generations, Yukawa and Chirality

Every matter particle comes in three versions, identical in every way except mass, and nobody knows why three. This chapter argues the number is counted by the substrate’s own directional structure rather than being a brute fact to be accepted, and then takes on the harder question of where the wildly different masses come from.

14.1.  Fermionic representation content under $\Pi$

The Jordan–Wigner identification of §11.2 and the $G_{2}\to\mathrm{SU}(3)\times\mathrm{U}(1)$ decomposition of §13.5 together fix the projection $\Pi$ on the fermionic Hilbert space $\mathcal{H}_{D}^{-}$ explicitly. Combined with the local-Lorentz spinor bundle $\mathcal{S}=\mathcal{S}^{+}\oplus\mathcal{S}^{-}$ on $(M,g)$, the one-generation fermion content is fully determined.

The one-generation decomposition. The per-$Q_{k}$ $\mathrm{Cl}(0,2)$ chiral structure underlying this decomposition is verified in Computation 47: each of the three quaternionic sub-algebras $Q_{k}$ through $\hat{\tau}$ delivers a $4+4$ chiral split with $\omega_{k}^{2}=-I$, structurally identical across $k$. The first exterior power $\bigwedge^{1}\mathbb{C}^{7}$ of the substrate fibre, restricted to the chiral sectors of the spinor bundle and refined under $\mathrm{SU}(3)_{c}\times\mathrm{SU}(2)_{L}\times\mathrm{U}(1)_{Y}$, decomposes as

$$\bigwedge\nolimits^{\!1}\!\mathbb{C}^{7}\,\Big{|}_{\Pi}^{\text{1 gen}}\;=\;\begin{array}[]{l}(\mathbf{3},\mathbf{2})_{1/3}^{L}\;\oplus\;(\mathbf{3},\mathbf{1})_{4/3}^{R}\;\oplus\;(\mathbf{3},\mathbf{1})_{-2/3}^{R}\\[2.0pt] \;\oplus\;(\mathbf{1},\mathbf{2})_{-1}^{L}\;\oplus\;(\mathbf{1},\mathbf{1})_{-2}^{R}\;\oplus\;(\mathbf{1},\mathbf{1})_{0}^{R}.\end{array}$$ (196)

These are, in order, the left-handed quark doublet $Q_{L}$, the right-handed up-type singlet $u_{R}$, the right-handed down-type singlet $d_{R}$, the left-handed lepton doublet $L_{L}$, the right-handed charged lepton $e_{R}$, and the right-handed neutrino $\nu_{R}$.

Hypercharge consistency check. The hypercharges appearing as subscripts in equation (196) reproduce the assignments derived independently in equation (194) via Yukawa-invariance and anomaly cancellation: $Y(Q_{L})=+1/3$, $Y(u_{R})=+4/3$, $Y(d_{R})=-2/3$, $Y(L_{L})=-1$, $Y(e_{R})=-2$. The right-handed neutrino sits in the singlet $(\mathbf{1},\mathbf{1})_{0}$; its appearance is a structural prediction of PST consistent with observed non-zero neutrino masses.

Threefold replication. Nature exhibits three generations. Taking three copies of equation (196) gives the complete Standard Model matter content plus three right-handed neutrinos:

$$\mathcal{H}_{D}^{-}\bigg{|}_{\Pi}\;=\;\bigoplus_{k=1}^{3}\Bigl{[}Q_{L}\oplus u_{R}\oplus d_{R}\oplus L_{L}\oplus e_{R}\oplus\nu_{R}\Bigr{]}_{k}.$$ (197)

Total chiral state count: $3\times(15+1)=48$ Weyl fermions (the standard $45=15\times 3$ Standard Model [SM] chiral states plus three right-handed neutrinos). The three-generation count is placed in Furey 2014 [121]’s six-$\mathrm{SU}(3)$-triplet decomposition of $\mathrm{Cl}(0,6)$, whose six triplets, six singlets and their antiparticles are identified as three generations of quarks $+$ leptons in the standard octonionic-SM reading. The $\mathrm{SU}(2)_{L}$ doublet pairing is supplied by the Dixon synthesis (§19.5). A separate combinatorial structure on $V_{7}$ is exhibited alongside it: the three quaternionic sub-algebras of $\mathbb{O}$ containing $\hat{\tau}$ (§14.2). It is consistent with Furey’s identification, subject to the colour-versus-generation distinguishability caveat, which PST inherits rather than resolves (§13.5, Computation 175).

What this does and does not predict. Equation (197) predicts the matter content and the hypercharges, both structural. It does not predict the Yukawa hierarchy or the Cabibbo–Kobayashi–Maskawa (CKM) matrix. At leading order the three generations in equation (197) are interchanged by the residual SU(3)-flavour symmetry of the $G_{2}$ structure (the symmetry that permutes the three complex directions of $\mathbf{3}_{\mathbb{C}}$). Because this symmetry treats the three generations identically, the leading PST Yukawa is degenerate ($y_{e}=y_{\mu}=y_{\tau}=y_{0}$ at $M_{*}$) and the leading PST CKM matrix is the identity. Observation falsifies both leading predictions: the observed lepton mass ratios at $M_{*}$ are $1:206:3477$ and the observed CKM mixing is non-trivial in every sector. The hierarchy and the mixing are therefore contingent in the precausal configuration $C$, not derivable from PST’s structural ingredients.

The underlying reason (the $S_{3}\subset S_{D}\times\mathbb{Z}_{2}$ invariance of the Bernoulli measure, the enumerated sub-leading mechanisms it rules out, and the resulting leading-order diagonal-equal-eigenvalue Yukawa and identity CKM) is established as a no-go theorem in §14.3; the observed values must arise from the precausal configuration $C$ itself.

SM-sector $\delta m_{h}^{2}$ cancellation: structural and coupling-weighted layers. The structural equal-weighting $N_{B}=N_{F}$ of $\mathcal{H}_{D}^{+}$ and $\mathcal{H}_{D}^{-}$ under the Bernoulli measure (§11.2) is a closed result on the substrate-side spectral action: $\mu$ assigns the same total weight to the bosonic and fermionic sub-Hilbert-spaces, so the spectral-action-level quadratic divergence in $m_{h}^{2}$ from the substrate DoF count vanishes structurally.

PST’s structural claim and the numerical SM Veltman bracket [94] $\mathcal{V}(M_{*}):=6\lambda+\tfrac{9}{4}g_{2}^{2}+\tfrac{3}{4}g_{1}^{2}-6y_{t}^{2}$ (the standard one-loop quadratic-sensitivity combination of equation (124), here evaluated with couplings run to $M_{*}$) are therefore not equivalent statements: the former is a substrate-side DoF-counting identity at the spectral-action level, the latter is an emergent-SM coupling-weighted one-loop combination at the matched scale $M_{*}\approx 1285\,\mathrm{GeV}$. Empirically $\mathcal{V}(M_{*})\approx-3.3$ at one loop. Equation (125) evaluates the same quantity to $\approx-3.5$ at electroweak-scale couplings; the difference is the running of the couplings between $v$ and $M_{*}$ under the renormalisation-group equations [RGE]. That numerical value reflects the SM-side renormalised couplings at $M_{*}$ (run from observed infrared [IR] values via SM RGE), not a violation of the structural identity. The substrate-side structural cancellation sits at the same layer as the partition-function-level Chamseddine–Connes correspondence of §12.5; the SM-side Veltman bracket numerical value is a downstream coupling-weighted quantity in the projected effective field theory (EFT), governed by SM RGE truncation in the same way that $\lambda_{\mathrm{SM}}(M_{*})\approx 0.0930$ is. The two layers are related by the projection $\Pi$ but not identified: the structural DoF-counting identity remains a closed substrate-side theorem, and the SM-side coupling-weighted bracket is a separate phenomenology evaluation that depends on the specific SM renormalised couplings at $M_{*}$.

14.2.  The three-generation count from the $V_{7}$ directional content

The threefold replication of the Standard Model matter content, three generations of fermions with identical quantum numbers, has been the “famously unexplained input” of every noncommutative-geometry Standard-Model framework. In PST, with the $V_{7}$-valued modal order parameter of P3 and the substrate-determined direction $\hat{\tau}=T(D)/|T(D)|\in S^{6}\subset V_{7}$ of P2, the count $N_{\mathrm{gen}}=3$ is read structurally from two mutually consistent constructions: (a) Furey 2014 [121]’s six-$\mathrm{SU}(3)$-triplet decomposition of $\mathrm{Cl}(0,6)$ (the standard octonionic-SM identification of three generations of one-generation content), and (b) the three quaternionic sub-algebras of $\mathbb{O}$ containing $\hat{\tau}$ developed in this section. Both readings rely on the same octonionic structure; their structural compatibility with a colour-singlet generation label (i.e. the statement that a colour $\mathrm{SU}(3)$ rotation does not mix the three generations) is a known open caveat of the octonionic SM programme (Furey’s own Open Questions [121], p. 4). The Dixon synthesis (§19.5, Computation 48) does not resolve it: in the ambient $\mathbb{C}\otimes\mathbb{H}\otimes\mathbb{O}$ the three generation sectors are orthogonal as index sets, but orthogonality is not invariance, and a colour rotation carries one sector into another (Computation 106, Computation 175). On $\mathbb{O}$ alone they overlap on a generation-blind core (Computation 47), which reflects $\mathrm{SU}(3)_{\mathrm{c}}=\mathrm{Stab}_{G_{2}}(\hat{\tau})$ acting irreducibly on $\hat{\tau}^{\perp}\cong\mathbb{C}^{3}$.

14.2.1.  The structural chain

The direction $\hat{\tau}$ selected at modal sublimation is a unit vector in $V_{7}$. It plays four interlocking roles, all forced by the foundational postulates:

1. Symmetry breaking. $G_{2}$, the automorphism group of $V_{7}$ (and the symmetry group of the pre-sublimation potential, P3), is broken by $\hat{\tau}$ to its stabiliser, the subgroup of $G_{2}$ fixing the unit vector $\hat{\tau}$. This stabiliser is the colour gauge group $\mathrm{SU}(3)$: the standard isomorphism $G_{2}/\mathrm{SU}(3)\cong S^{6}$ means every point on $S^{6}$ has $\mathrm{SU}(3)$ stabiliser. Verified in Computation 49.

2. Complex structure on $\hat{\tau}^{\perp}$. The 6-dimensional subspace $\hat{\tau}^{\perp}\subset V_{7}$ orthogonal to $\hat{\tau}$ inherits a complex structure $J$ given by left-multiplication by $\hat{\tau}$ in the octonion product: $J(v)=\hat{\tau}\times v$ for $v\in\hat{\tau}^{\perp}$. Computation 49 verifies $J^{2}=-I$ on $\hat{\tau}^{\perp}$, so $\hat{\tau}^{\perp}\cong\mathbb{C}^{3}$ carries the $\mathrm{SU}(3)$ fundamental representation (the colour content of one fermion generation).

3. Three quaternionic sub-algebras. A classical structural fact about $\mathbb{O}$ (verified in Computation 46): exactly three quaternionic sub-algebras of $\mathbb{O}$ contain the complex sub-algebra $C_{\hat{\tau}}=\mathrm{span}\{1,\hat{\tau}\}$. For $\hat{\tau}=e_{1}$, these are

   $$Q_{1}=\{1,e_{1},e_{2},e_{3}\},\quad Q_{2}=\{1,e_{1},e_{4},e_{5}\},\quad Q_{3}=\{1,e_{1},e_{6},e_{7}\},$$ (198)

   one per Fano-plane line through $\hat{\tau}$. The three sub-algebras partition $\hat{\tau}^{\perp}$ (the 6-dim perpendicular space) into three disjoint 2-dim pairs. The count of $3$ is $G_{2}$-symmetric: it does not depend on which unit vector in $S^{6}$ is selected.

4. Three Cl(0,2) chiral splittings. Each $Q_{k}$ generates a $\mathrm{Cl}(0,2)$ sub-Clifford structure on $\mathbb{R}^{8}$ via left multiplication by its two non-$\hat{\tau}$ generators. Computation 47 verifies each $Q_{k}$ delivers a $4+4$ chiral split with volume $\omega_{k}^{2}=-I$, structurally identical across the three $Q_{k}$. This matches exactly the observed pattern of three generations with identical Standard-Model quantum numbers.

14.2.2.  Three disjoint generation sectors

The chain-algebra space $\mathbb{C}\otimes\mathbb{H}\otimes\mathbb{O}\cong\mathbb{R}^{64}$, which carries the matter representation under PST’s Cl(0,6) chirality + Furey 2014 chain (§13.7), decomposes under the $\hat{\tau}$-selection as

$$\mathbb{C}\otimes\mathbb{H}\otimes\mathbb{O}\;=\;\underbrace{\mathbb{C}\otimes\mathbb{H}\otimes\mathrm{span}\{1,\hat{\tau}\}}_{\textbf{core},\;\dim=16}\;\oplus\;\bigoplus_{k=1}^{3}\underbrace{\mathbb{C}\otimes\mathbb{H}\otimes\mathrm{span}\{e_{a_{k}},e_{b_{k}}\}}_{G_{k},\;\dim=16}.$$ (199)

The decomposition is verified in Computation 48. The arithmetic is exact: $16+3\times 16=64$. The pairwise intersections $G_{i}\cap G_{j}=0$ for $i\neq j$; the three generation-specific sectors are mutually disjoint as real subspaces of $\mathbb{R}^{64}$. The 16-dimensional core carries the gauge structure (the algebra $A_{F}$ acting on its own minimal ideal); it is generation-blind. Each $G_{k}$ carries one generation’s matter content.

The per-generation Hilbert space is $\mathcal{H}_{k}=\mathrm{core}\oplus G_{k}$ of real dimension $32$. Its chiral half (under the extended chiral projector $f=(I+i_{\mathbb{C}}\otimes I_{\mathbb{H}}\otimes\omega_{6})/2$ of Computation 45) has dimension $16$, matching exactly the Standard-Model count of $15$ chiral fermions plus one right-handed neutrino per generation (§14.1). The chiral structure is identical across the three $\mathcal{H}_{k}$, recovering the observed pattern of three generations with identical quantum numbers.

14.2.3.  Colour and generation: two guises of one three

The number $3$ appears in PST in two guises, which this subsection shows to be one:

• Colour $N_{c}=3$: the rank of the $M_{3}(\mathbb{C})$ factor of $A_{F}$ (§13.5). It arises within a single generation, from the $\mathrm{SU}(3)\subset G_{2}$ stabiliser of $\hat{\tau}$.

• Generations $N_{\mathrm{gen}}=3$: the count of quaternionic sub-algebras of $\mathbb{O}$ containing the complex sub-algebra $C_{\hat{\tau}}$. It arises across generations, from the $\hat{\tau}$-direction selection in $V_{7}$.

Both counts emerge from the same $\hat{\tau}$-selection event, but they parameterise different structural data: colour is internal to each generation’s $\mathcal{H}_{k}$ via the $M_{3}(\mathbb{C})$ factor of $A_{F}$; generation $3$ counts the $\mathcal{H}_{k}$ themselves.

What the shared origin does and does not buy. That the two descend from one selection event makes them rigid rather than independent. Both are read off the single six-dimensional space $\hat{\tau}^{\perp}$: colour as the $\mathrm{SU}(3)$ acting on it as $\mathbb{C}^{3}$, the generation count as the three Fano lines through $\hat{\tau}$ that partition it. The two cannot be varied separately, and a theory that fixed each from its own source would have no reason to expect them to agree.

It is worth being exact about how much that is worth, because it invites an overclaim. It buys no observational signature, for two reasons this section has already supplied. First, the construction is $G_{2}$-equivariant: the count of $3$ does not depend on which unit vector of $S^{6}$ is selected, and a different $\hat{\tau}^{\prime}$ yields a $G_{2}$-rotated copy of the same $\Phi$ with the rotation an unphysical relabelling (§4.10). Nothing observable distinguishes one pinned direction from another, which is the same argument that empties the residual $Z_{2}$ choice of content. Second, the rigidity carries no free parameter: $N_{c}$ and $N_{\mathrm{gen}}$ are fixed by the Hurwitz and $G_{2}$ algebra before any physics enters, so there is no slack in which a correlation could be measured. What the shared origin delivers is a structural identity, not a prediction. Nor does it extend to a third structure: the chirality of $\mathrm{SU}(2)_{L}$ is oriented by the scalar $\varepsilon>0$ and not by $\hat{\tau}$ (§14.4), so it does not join this list.

The structural reading of $N_{\mathrm{gen}}=3$ via equation (199) is verified numerically in Computations 46–48, consistent with Furey 2014’s six-$\mathrm{SU}(3)$-triplet decomposition of $\mathrm{Cl}(0,6)$ (the standard octonionic Standard Model (SM) identification). The algebraic foundation underlying the chain-algebra space $\mathbb{C}\otimes\mathbb{H}\otimes\mathbb{O}$ is established in Computations 43 and 44 (Dixon-algebra foundation: the octonion chain algebra on $\mathbb{R}^{8}$ and its $\mathbb{C}\otimes\mathbb{H}\otimes\mathbb{O}$ extension carrying one full generation of SM matter content). The colour-vs-generation distinguishability of the three sectors on $\mathbb{O}$ alone is non-trivial: $\mathrm{Stab}_{G_{2}}(\hat{\tau})\cong\mathrm{SU}(3)$ acts on $\hat{\tau}^{\perp}\cong\mathbb{C}^{3}$ as the fundamental, so the three sub-algebras mix under colour $\mathrm{SU}(3)$, and that mixing is not undone by the ambient algebra. Passing to $\mathbb{C}\otimes\mathbb{H}\otimes\mathbb{O}$ makes the three sectors orthogonal as index sets (Computation 48), but orthogonal is not invariant: tensoring a colour-rotated line with factors colour does not touch leaves it colour-rotated. Computation 106 measures the leakage directly and finds it $O(1)$, with only the maximal torus $\mathrm{U}(1)^{2}\subset\mathrm{SU}(3)_{c}$ preserving all three lines. There is a dimension obstruction too: independent colour and generation factors would need $\mathbf{3}\otimes\mathbf{3}$ inside the octonionic slot, real dimension at least $18$, while $\hat{\tau}^{\perp}$ has real dimension $6$. Colour and generation are not two threes here; they are the same three, which is the octonionic programme’s own open question rather than something PST closes. Inside this theory the question is settled in the negative: $\hat{\tau}^{\perp}$ is the irreducible $\mathbf{3}$, so no colour-invariant generation splitting exists in it at all (Computation 175, Appendix C.4); what PST inherits is where the programme puts generations if not there, and that is not carried as a structural item of this book. Two refinements would sharpen the reading: (i) demonstrate that each $G_{k}$ carries the explicit $\mathrm{SU}(3)\times\mathrm{SU}(2)\times\mathrm{U}(1)$ representation content matching the one-generation decomposition of equation (196); (ii) establish consistency of the $\hat{\tau}$-determination $\hat{\tau}=T(D)/|T(D)|$ with the Mosco-limit projection $\Phi:S\to G$. Both are concrete operator-algebra tasks of the same character as the section-round-trip and Dirac-aware-lift closures of §18.3.

14.3.  Yukawa and CKM: the structural-scope theorem

The two-layer distinction. PST separates its outputs into two kinds. The structural layer is whatever follows from the postulates P1–P3 alone, and is therefore the same in every instantiated reality: the gauge group, the three-generation count, chirality, spin-statistics. The contingent layer is whatever depends on which particular precausal configuration $C$ actually sublimated, encoded in its asymmetric-tension distribution $T(C)$. The contingent layer plays the role that initial conditions play elsewhere in physics: real content of the world, but not derivable from the laws (§20.7 gives the quantitative comparison with the Standard Model’s free-parameter count). The observed Yukawa hierarchy (charged-lepton masses in the ratio $1:206:3477$ at $M_{*}$) and the non-trivial Cabibbo–Kobayashi–Maskawa (CKM) mixing matrix belong to the contingent layer. That placement is proved, not assumed: this section states the theorem, established in Computation 1, and closes the escape routes.

The no-go theorem that forces the placement (substrate level). The Bernoulli product measure $\mu=\bigotimes_{a}\mathrm{Bern}(1/2)$ on $\mathcal{P}(D)$ is invariant under the full permutation group $S_{D}\times\mathbb{Z}_{2}$ acting on the $D$ substrate bits, and generation permutation $S_{3}\subset S_{D}$ (permuting the bit-triples assigned to the three generations) is a subgroup of it. Any observable computed from the structural ingredients $(D,\delta,\mu)$ alone must therefore be $S_{3}$-symmetric across generations. Computation 1 verifies this numerically for the leading proposed generation-asymmetric mechanisms: the Yukawa overlap integral, the multi-instanton Atiyah–Singer index [178] per representation, and the Jordan–Wigner expectation in the even-parity sector are all $S_{3}$-invariant. The consequence: at leading structural order the three Yukawa eigenvalues per sector are equal and the CKM matrix is the identity, so the observed hierarchy and mixing cannot come from the structural layer. The only remaining source is $T(C)$.

Why “by construction”. This is not a failed derivation reclassified as scope: the symmetry of the structural ingredients makes generation asymmetry unreachable from them in principle, so the mass spectrum was never in scope for P1–P3. Two closures secure the placement. First, at the substrate level, the proposals for generating the hierarchy from structure are examined and shown not to break $S_{3}$ invariance. Three are considered: the $V_{7}$-direction constraint (Computation 76), the dynamical per-$Q_{k}$ directed modal threshold (Computation 77), and structural-$\beta$ candidates in the Boltzmann [183] mechanism. The last of these runs through the structural candidates of Computation 78, sharpened in Computation 79 to the closed form $\beta=28+1/4=28.25$ above $\dim\mathrm{SO}(8)=28$, with the Spin-related $p_{1}$ closure in Computation 80. Each candidate inherits an $S_{3}$-invariant Bernoulli-measure trace and cannot lift the leading-order generation degeneracy on its own. Second, the spectral-triple extension: the theorem survives the Connes–Chamseddine framework via inner-fluctuation analysis. Since $A_{F}$ acts on $H_{F}$ as $a^{(1)}\otimes I_{\mathrm{gen}}$ with no generation labels, an inner fluctuation $\omega=\sum a_{i}\,[D,b_{i}]$ inherits the generation structure of $D_{F}$ unchanged; it cannot lift a degeneracy absent from $D_{F}$ nor remove one present. The generation pattern is locked in $D_{F}$, which is a contingent input, a direct corollary of the bimodule structure (Computation 1, section 7 of the script).

Two supporting checks. Quantitatively, the observed flavour parameters number about twenty for Dirac neutrinos: nine Yukawa eigenvalues, four CKM parameters, and seven Pontecorvo–Maki–Nakagawa–Sakata [PMNS]. They consume about $31\%$ of the $2^{D}=64$ real degrees of freedom available in $T(C)$ at $D=6$ substrate bits (Computation 1, section 8 of the script). PST is therefore neither over-determined nor under-equipped to encode the observed flavour pattern: there is ample structural room for the contingency, with a definite count of how much is consumed. Empirically, Computation 138 checks the measured CKM matrix against the form the theorem requires, identity plus a small $T(C)$-contingent perturbation: the deviation is Cabibbo-sized, every off-diagonal entry is small, and the magnitudes are unitary to sub-percent. A genuinely order-one off-diagonal CKM matrix would have refuted the placement; the observed matrix conforms to it.

The honest scope. PST does not derive contingent parameter values; it derives the structural form within which those parameters live. In good company, this matches the noncommutative-geometric Standard Model of Connes–Chamseddine, in which the Yukawa eigenvalues are inputs to $D_{F}$ rather than outputs.

14.4.  Chirality of $\mathrm{SU}(2)_{L}$ as a structural consequence

The chirality of the weak gauge group (that $\mathrm{SU}(2)_{L}$ acts on left-handed doublets only) is a structural consequence of PST’s substrate combined with the directed modal threshold of Chapter 4. This is the chiral-bimodule support property of the Connes–Chamseddine finite triple, established for PST in Computations 7, 8, and 11.

The substrate’s three-sphere factor $S^{3}$ in $M=\mathbb{R}\times S^{3}$ carries an isometry algebra $\mathrm{so}(4)=\mathrm{su}(2)_{L}\oplus\mathrm{su}(2)_{R}$, equivalently the algebra of quaternion left and right multiplications on $\mathbb{H}$. This furnishes two candidate $\mathrm{su}(2)$ representations on the internal Hilbert space $H_{F}$, one acting on each chirality block of $\gamma_{F}$. Both have clean block support and the two are exact parity conjugates of each other ($P\cdot H_{A}\cdot P^{-1}=H_{B}$ verified numerically; Computation 11). Together with Computation 7’s verification that the order-one condition selects the Yukawa form given the chiral bimodule, the chirality reduces to a single physical question: which of the two parity-conjugate $\mathrm{su}(2)$ embeddings is realised in nature?

The answer is the parity-violation mechanism PST already commits to in Chapter 4: the directed modal threshold $\varepsilon>0$ orients the $S^{3}$ volume form, fixing one factor as the physical $\mathrm{SU}(2)_{L}$. The same mechanism that produces parity violation selects the chiral bimodule. Chirality of $\mathrm{SU}(2)_{L}$ is therefore structural, conditional only on the threshold-direction parity argument PST commits to elsewhere.

Which quantity does the orienting. The orientation just used is supplied by $\varepsilon=|T(C)|-\tau>0$, which is a scalar: an oriented inequality recording that tension exceeds the threshold from below. It is not supplied by the substrate direction $\hat{\tau}=T(D)/|T(D)|$ of P2, a $V_{7}$ vector fixed by $D$ alone, which is the quantity that fixes colour and the generation count (§14.2). The two are different objects and the distinction is not cosmetic: whether the vector and the scalar orientation are the same datum is exactly the consistency of the $\hat{\tau}$-determination with the Mosco-limit projection $\Phi$, which §14.2 does not establish. Until it is established, chirality is not licensed as a third structure descending from $\hat{\tau}$, and §14.2 draws the consequence.

Handedness convention. The threshold orientation predicts the existence of a chirality, not the conventional “left” vs. “right” label: $L\leftrightarrow R$ is a relabelling under spatial parity, and the assignment that the chiral factor coincides with the SM’s $\mathrm{SU}(2)_{L}$ is a convention fixed by matching to the observed weak-interaction handedness. The mechanism predicts chirality; the empirical sign of the threshold direction picks the handedness label.

14.5.  Summary

This chapter settles what the fermion sector carries structurally and what it leaves to the configuration. The representation content of one generation under $\Pi$ follows from the canonical anticommutation relation (CAR) identification and the internal algebra (§14.1). The three-generation count is read from the directional content of $V_{7}$: $\hat{\tau}^{\perp}$ decomposes into three quaternionic sub-algebras through $\hat{\tau}$, and colour and generation are one three, not two (§14.2, Computation 175). The Yukawa couplings and the CKM matrix lie in the configuration’s content by the structural-scope theorem (§14.3): substrate-level $S_{3}$ invariance forces $S_{3}$-symmetric Yukawa eigenvalues at leading order, so the observed hierarchy is $T(C)$-contingent rather than structural. The chirality of $\mathrm{SU}(2)_{L}$ is a structural consequence, the empirical sign of the threshold direction picking the handedness label (§14.4).