Axial Torsion as the Default: Where It Lives and What It Does

Axial torsion is the engine that distinguishes chirality, drives overlap physics, and deforms geometry from the inside.

What happens when geometry isn’t just curved, but twisted? In conventional setups, torsion is either zero or perturbative—something you switch on when convenient. In GU, torsion is built-in. It’s not a feature; it’s the default.

Why axial? Because the totally antisymmetric part of the torsion tensor—the part you can wedge into a 3-form—is the only one that survives all the project’s constraints: gauge covariance, pullback consistency, and compatibility with chiral fermions.

More specifically, the augmented torsion variable

$$ T := \xi - \varepsilon^{-1} d_{A_0} \varepsilon $$

lives in the translation module $N = \Omega^1(Y, \mathfrak{ad})$ and enters directly into the dynamics via the action:

$$ I_1(\omega) = \int_Y \langle T, *_Y (\bullet_\varepsilon F_B - \kappa_1 T)\rangle $$

In this post, we spell out where this torsion lives in the decomposition $TY|_X \simeq TX \oplus N_\iota$, how it acts, and why axial torsion is the only one that does the job without cheating.

Definitions / Notation used

  • $\omega = (\varepsilon, \xi)$: transport variable (gauge + translation)
  • $A_0$: fixed background connection on $Y$
  • $T := \xi - \varepsilon^{-1} d_{A_0} \varepsilon$: augmented torsion
  • $F_B$: curvature of $B_\omega := A_0 \cdot \varepsilon$
  • $\Upsilon_\omega := \bullet_\varepsilon(F_B) - \kappa_1 T$: dynamic variable for curvature-minus-torsion
  • $I_1(\omega)$: action $\int_Y \langle T, *_Y \Upsilon_\omega\rangle$
  • $TY|_X \simeq TX \oplus N_\iota$: tangent/normal decomposition
  • $TX$: tangential bundle (directions along $X$)
  • $N_\iota$: normal bundle (directions orthogonal to $X$)
  • Axial torsion: totally antisymmetric part, expressible as a 3-form

The Technical Heart: Where Torsion Lives

Let’s decompose torsion into its components under the $TX \oplus N_\iota$ split. A 1-form $T$ with values in $\mathfrak{ad}(P_H)$ can have three kinds of components:

  1. Tangential: $T_\mu dx^\mu$ — observable directions
  2. Mixed: $T_\mu^a dx^\mu \otimes n^a$ — hybrid
  3. Normal: $T^a n^a$ — internal-only

But torsion enters dynamics not linearly but quadratically, via wedge products and $*_Y$ duals. That means only combinations that result in gauge-invariant 3-forms matter. The only such 3-form that is:

  • gauge-covariant
  • nontrivial under pullback
  • consistent with spinor coupling

is the totally antisymmetric axial torsion:

$$ T = T_{abc} n^a \wedge n^b \wedge n^c $$

This is a 3-form fully in the normal bundle—nothing tangential survives. It is “purely internal” but has real effects: through its coupling to curvature (via the Shiab operator $\bullet_\varepsilon$) and through fermion overlap, it governs both chirality selection and mass-like effects.

Why axial? Because it’s the only torsion component that:

  • is manifestly antisymmetric
  • wedges into $\Theta_E$ consistently
  • survives under $\mathrm{Spin}(7,7)$ projection
  • doesn’t violate pullback rules

One Diagram in Words

Imagine the 10D normal space $N_\iota$ at each point of $X$ as a flexible internal frame. Axial torsion is a twisting of this frame that wraps three normal directions at once. It doesn’t affect $X$ directly—but it skews how fields pull back, how chiral modes survive, and how overlaps induce mass scales.

Assumptions vs Consequences

Assumptions (Ansatz):

  • Torsion is present by default in the connection.
  • Only the totally antisymmetric (axial) component is turned on.
  • Torsion is valued in $\mathfrak{ad}(P_H)$, and transforms covariantly.
  • Dynamics are written in terms of augmented torsion $T = \xi - \varepsilon^{-1} d_{A_0} \varepsilon$.

Consequences (Operational):

  • Only internal (normal-normal-normal) components of torsion survive in the action.
  • Tangential or mixed torsion components vanish or violate the pullback structure.
  • The action couples $T$ to curvature via Shiab: no naive Ricci-like traces.
  • $T$ acts as the geometric selector for chiral fermion modes on $X$.

Why This Matters

  • Fermion chirality is not inserted—it is selected by the geometry of axial torsion.
  • Mass scales from overlap depend on torsion-localized Hermite modes.
  • Gauge invariance of the action $I_1(\omega)$ hinges on using $T$, not raw connection differences.
  • Cosmological terms inherit $\sigma$-weighted suppression from the wedge structure of $T$.

Key Takeaway

In this program, axial torsion isn’t a tweak—it’s the backbone of chirality, mass generation, and internal geometry. It lives entirely in the normal bundle and defines the gravitational interaction structure from the inside out.

Technical Takeaway

  • $T = \xi - \varepsilon^{-1} d_{A_0} \varepsilon$
  • Axial torsion: $T_{abc} n^a \wedge n^b \wedge n^c \in \Omega^3(Y, \mathfrak{ad})$
  • Only totally antisymmetric components enter the action $I_1(\omega)$

Loading comments…