On Generalised Discrete Torsion

This paper generalizes Vafa's discrete torsion to a cohomological framework ($H^2_G(M; U(1))$) that assigns distinct local phases to singular loci in 2d gauged sigma models, thereby determining which specific smooth Calabi-Yau and $G_2$ geometries emerge from orbifold resolutions like $T^6/\mathbb{Z}_2^2$ and $T^7/\mathbb{Z}_2^3$.

The Gist

Imagine you are an architect designing a beautiful, complex building (a "Calabi-Yau" or "G2" manifold) for a futuristic city. However, your blueprint has some rough spots—sharp corners and jagged edges where the geometry breaks down. These are called singularities.

In the world of theoretical physics, specifically string theory, these buildings aren't just made of concrete; they are made of vibrating strings. To fix the rough spots in the blueprint, you have two main tools:

1. Resolution: You can "blow up" the corner, replacing the sharp point with a small, smooth bubble (like inflating a balloon at the tip of a cone).
2. Deformation: You can "squish" the corner, smoothing it out by changing the shape of the space around it (like melting ice to fill a crack).

For a long time, physicists thought that when you had a building with many of these rough spots, you had to make the same choice for all of them. If you decided to "blow up" one corner, you had to blow up every single corner to keep the building stable. This was like a rule: "All or nothing."

The New Discovery: "Generalised Discrete Torsion"

This paper, by Philip Boyle Smith and Yuji Tachikawa, introduces a new, more flexible tool called Generalised Discrete Torsion.

Think of Ordinary Discrete Torsion as a single, global switch on your building's control panel. If you flip it "On," it changes the physics of the whole building in one specific way. If you flip it "Off," it changes it in another. You can't choose differently for different rooms.

Generalised Discrete Torsion is like upgrading that control panel to a smart home system with individual switches for every single room. Now, you can theoretically decide to "blow up" the corner in the kitchen while "deforming" the corner in the bedroom.

The Catch: The "Neighborhood Watch" Rule

The authors discovered that while you can have different switches for different rooms, they aren't completely independent. There is a hidden "Neighborhood Watch" rule.

Imagine your building is a neighborhood. If two rough corners are right next to each other (they intersect), you can't treat them totally differently. If you try to blow up one and deform the other, the "wall" between them would tear, and the building would collapse.

The paper shows that:

• In the 6-dimensional case ($T^6/Z_2^2$): The rough spots are all connected in a giant web. Because they all touch each other, the "Neighborhood Watch" rule is strict. You are forced to make the same choice for the whole building. The new "smart switches" don't actually give you more freedom than the old "single switch" in this specific case.
• In the 7-dimensional case ($T^7/Z_2^3$): The rough spots are like islands in a sea; they don't touch each other. Here, you should be able to choose independently for each island. However, the authors found a surprise: The physics of the strings still links them together. Even though the islands are far apart, the "smart switches" are secretly wired to each other. You can't just pick any random combination of choices. You are limited to only 3 out of the 9 possible combinations that mathematicians thought were possible.

The Analogy of the "Magic Dice"

To understand why this happens, imagine you are rolling dice to decide the shape of each room.

• Mathematicians (Joyce) said: "You can roll a die for each of the 9 islands independently. You can get any combination of results."
• The Physicists (Smith & Tachikawa) said: "Wait, the dice are loaded. They are connected by invisible strings. If you roll a '6' for the first island, the dice for the other islands are forced to roll specific numbers. You can't get every combination; you can only get a few specific patterns."

Why Does This Matter?

This paper is a bridge between two worlds:

1. The World of Geometry: Where mathematicians build smooth shapes by hand, choosing exactly how to fix every singularity.
2. The World of String Theory: Where physicists describe these shapes using vibrating strings and quantum rules.

The authors realized that the "Generalised Discrete Torsion" (the smart switches) is the correct way to describe the string theory version of these shapes. However, they hit a puzzle: String theory seems to be more restrictive than pure geometry.

In the 7-dimensional case, the string theory description (the orbifold CFT) cannot produce all the smooth shapes that mathematicians can build. It misses some of the possibilities. This suggests that there might be other, more exotic ways to describe these shapes in string theory that we haven't discovered yet, or that some of the mathematical shapes simply cannot exist as consistent quantum string universes.

Summary in a Nutshell

• The Problem: How do we fix the jagged edges of quantum universes?
• The Old Idea: You have to fix them all the same way.
• The New Idea: You can fix them differently, but only if you follow a strict set of hidden rules.
• The Result: In some universes, the rules force you to act uniformly. In others, the rules allow some variety, but not as much as pure math suggests. The "quantum fabric" of the universe is more rigid than we thought.

This work refines our understanding of how the microscopic rules of quantum mechanics constrain the macroscopic shapes of the universe, showing that nature might be pickier about which smooth geometries it allows to exist than we previously believed.

Technical Summary

1. Problem Statement

The paper addresses a long-standing question in string theory and orbifold conformal field theory (CFT) regarding the consistency of generalised discrete torsion.

• Background: Ordinary discrete torsion, introduced by Vafa, is a phase factor $\alpha \in H^2(BG; U(1))$ associated with gauging a finite group $G$ in a 2D sigma model. It affects the spectrum and the geometry of the resulting orbifold $M/G$, specifically determining whether singularities are "resolved" (blown up) or "deformed."
• The Gap: In classical geometry (specifically for Calabi-Yau and $G_2$ manifolds constructed by Joyce), it is possible to resolve or deform singular loci independently at different locations. Gaberdiel and Kaste (2004) proposed that one could assign discrete torsion phases independently to each singular locus in the CFT to mimic this classical flexibility. However, the consistency of such independent assignments at higher genus (worldsheet topology) was unclear.
• The Core Question: Can one assign local discrete torsion phases to different singular loci of an orbifold $M/G$ independently while maintaining a consistent quantum theory? If so, what is the mathematical structure governing these phases, and what geometries do they produce?

2. Methodology

The authors propose that the correct mathematical framework for generalised discrete torsion is equivariant cohomology, specifically $H^2_G(M; U(1))$, rather than just group cohomology $H^2(BG; U(1))$.

• Theoretical Framework:

  • They model the gauged sigma model as a map $f: \Sigma \to (EG \times M)/G$, where $EG$ is the universal $G$-bundle.
  • The discrete torsion is defined by a cohomology class $\omega \in H^2_G(M; U(1))$.
  • They distinguish between the continuous part (related to flat $B$-fields) and the discrete deformation classes $[H^2_G(M; U(1))]$, which are finite abelian groups.
  • They utilize the Borel construction to relate the equivariant cohomology of the target space to the ordinary cohomology of a larger group $\hat{G}$ that acts on a covering space $\tilde{M}$. Specifically, for torus orbifolds $T^n/G$, they show $H^2_G(T^n; U(1)) \cong H^2(B\hat{G}; U(1))$, where $\hat{G}$ is an extension $1 \to \mathbb{Z}^n \to \hat{G} \to G \to 1$.

• Computational Strategy:

  • Local Analysis: They compute the pullback of the global torsion class $\omega$ to specific singular loci (fixed points or fixed tori). This local value determines the resolution/deformation choice at that specific locus.
  • Spectral Calculation: They explicitly compute the Hodge numbers ($h^{1,1}, h^{2,1}$) for Calabi-Yau orbifolds and Betti numbers ($b_2, b_3$) for $G_2$ orbifolds by analyzing the ground states of the gauged sigma model in the Ramond sector.
  • Case Studies: The paper focuses on two specific examples:
    1. $T^6/\mathbb{Z}_2^2$ (Calabi-Yau).
    2. $T^7/\mathbb{Z}_2^3$ ($G_2$ manifold).
  • Verification: They verify their results using two independent methods: explicit cell complex constructions and group extension theory (analyzing central extensions of the symmetry groups).

3. Key Contributions

• Formalisation of Generalised Discrete Torsion: The authors rigorously define generalised discrete torsion as an element of $H^2_G(M; U(1))$. They demonstrate that this formulation naturally incorporates the "local" phases suggested by Gaberdiel and Kaste but imposes consistency constraints derived from the global topology of the orbifold.
• Resolution of the Independence Paradox: They prove that while local torsion phases can differ, they are not completely independent. The global cohomology class $\omega$ imposes linear constraints on the local phases.
• Reconciliation with Classical Geometry:
  • For $T^6/\mathbb{Z}_2^2$, they show that the singular $T^2$ loci intersect. To maintain smoothness at intersections, the resolution choice must be uniform across all loci. This recovers the standard result where only ordinary discrete torsion (global choice) is allowed, contradicting the idea of fully independent local choices in this specific geometry.
  • For $T^7/\mathbb{Z}_2^3$, the singular $T^3$ loci do not intersect. Classically, one expects to resolve them independently. However, the authors find that the worldsheet CFT with generalised torsion cannot realize all classical resolutions. Specifically, the local phases are constrained such that only an even number of loci can be "deformed" (or resolved) in a specific way.

4. Key Results

Case A: $T^6/\mathbb{Z}_2^2$ (Calabi-Yau)

• Cohomology: The group of deformation classes is $H^2_{\mathbb{Z}_2^2}(T^6; U(1)) \cong U(1)^3 \times \mathbb{Z}_2^{19}$.
• Local Constraints: The 19 discrete phases control the resolution of 48 singular $T^2$ loci.
• Finding: The condition for the geometry to be fully desingularized (smooth) requires that the local torsion phases vanish on the intersections. Since the $T^2$ loci intersect, the phases must be identical everywhere.
• Outcome: The theory only realizes the two standard Calabi-Yau manifolds with $(h^{1,1}, h^{2,1}) = (51, 3)$ or $(3, 51)$. It cannot realize the intermediate geometries with mixed resolution/deformation proposed by Gaberdiel and Kaste.

Case B: $T^7/\mathbb{Z}_2^3$ ($G_2$ Manifold)

• Cohomology: The group of deformation classes is $H^2_{\mathbb{Z}_2^3}(T^7; U(1)) \cong \mathbb{Z}_2^{21}$.
• Local Constraints: There are 21 discrete phases controlling the resolution of 48 singular $T^3$ loci.
• Finding: Unlike the $T^6$ case, the singular $T^3$ loci do not intersect. However, the global cohomology class $\omega$ imposes a constraint: the sum of the local torsion phases (specifically the "deformation" choice) over the orbits of $\gamma$-fixed loci must be even.
• Outcome:
  • The classical geometry allows for $0 \le \ell \le 8$ independent resolutions (where $\ell$ is the number of deformed loci).
  • The orbifold CFT with generalised torsion only realizes 3 out of the 9 possible Betti numbers (specifically $\ell = 0, 4, 8$).
  • The CFT cannot access the "odd" resolution configurations ($\ell = 1, 2, 3, 5, 6, 7$).

5. Significance and Implications

• Limit of Orbifold CFTs: The paper concludes that the worldsheet description of certain smooth $G_2$ and Calabi-Yau manifolds (those with "mixed" resolution choices) cannot be captured by a simple orbifold CFT with generalised discrete torsion valued in $H^2_G(M; U(1))$.
• Consistency vs. Geometry: There is a fundamental tension between the consistency of the 2D quantum field theory (which enforces global constraints on local phases) and the flexibility of classical differential geometry (which allows independent local choices).
• Refinement of Discrete Torsion: The work clarifies that "generalised discrete torsion" is not a free parameter for every singularity but is a global topological invariant of the gauged sigma model.
• Future Directions: The results suggest that to describe the full moduli space of these geometries (including the "odd" resolutions), one may need to go beyond standard orbifold CFTs, potentially involving non-geometric phases, different gauging procedures, or more complex topological terms not captured by $H^2_G(M; U(1))$.

In summary, Smith and Tachikawa provide a rigorous cohomological framework for discrete torsion, demonstrating that while it allows for some local variation, global consistency constraints prevent the realization of all classically possible smooth geometries within the standard orbifold CFT formalism.