Orbifold completion of 3-categories

Imagine the universe of physics as a giant, multi-layered cake. In the simplest version of this cake (called a "closed" Topological Quantum Field Theory, or TQFT), the layers are smooth and uniform. But in the real world, and in more advanced physics theories, this cake has cracks, fillings, and different flavors mixed in. These are called defects.

This paper by Nils Carqueville and Lukas Müller is about building a massive, universal "instruction manual" (a mathematical structure called a 3-category) that can describe every possible way these defects can exist, interact, and be transformed in a 3-dimensional universe.

Here is the breakdown of their work using simple analogies:

1. The Problem: Too Many Rules, Too Many Shapes

Imagine you are trying to build a Lego castle. You have a set of basic bricks (the "bulk theories"). But you also have special pieces: walls (surface defects), pipes (line defects), and joints (point defects).

The old way: Physicists had to figure out the rules for how these special pieces fit together one by one. It was like trying to solve a puzzle where you only had a few pieces and had to guess the rest.
The new way: The authors created a "master recipe" called Orbifold Completion. This is a mathematical machine that takes your basic Lego set and automatically generates every possible valid way to add special pieces, ensuring they all fit together perfectly without breaking the laws of physics.

2. The Core Concept: The "Orbifold" Machine

Think of an "orbifold" not as a sci-fi portal, but as a universal translator for symmetry.

In the 2D world (flat surfaces), mathematicians already knew how to build this translator. It took a simple shape and showed you all the ways it could be folded or glued to create new, stable shapes.
This paper asks: "What does this translator look like in 3D?"
They built a 3D version of this machine. They call it $T_{orb}$ .
- Input: You feed it a "Gray category with duals" (a fancy math term for a 3D rulebook that already has some symmetry built-in).
- Output: It spits out a new, much richer rulebook ( $T_{orb}$ ) that contains all the possible defects and how they talk to each other.

3. The Ingredients: "Orbifold Data"

To make this machine work, they had to define exactly what a "valid defect" looks like in 3D. They call these Orbifold Data.

The Analogy: Imagine a 3D puzzle piece. For it to be a valid "orbifold" piece, it can't just be any shape. It has to satisfy specific "gluing rules" (mathematical equations) that ensure if you rotate it, flip it, or combine it with other pieces, the whole structure stays stable.
The authors wrote down these rules (shown as diagrams in the paper) which act like a quality control checklist. If a defect passes the checklist, it gets a seat at the table in the new rulebook.

4. The Big Discovery: The Machine is Self-Healing

One of the most surprising things they found is that this new machine is complete.

If you take your new, super-rich rulebook ( $T_{orb}$ ) and run it through the machine again, you don't get something new. You get the exact same thing back.
The Metaphor: It's like a mirror that, when you look into it, shows you a reflection of the mirror itself. It has reached a state of "perfection" where no new defects can be added that weren't already implied by the rules. They call this property idempotence (doing the same thing twice changes nothing).

5. Why This Matters: The "Universal State Sum"

The authors show how to use this machine to build State Sum Models.

The Analogy: Imagine you want to calculate the total "vibe" or energy of a complex 3D shape (like a knotted piece of string in space).
The Method: Instead of calculating the whole thing at once (which is impossible), you chop the shape into tiny triangles (a triangulation).
The Magic: Because the authors built their rulebook to be "triangulation invariant," it doesn't matter how you chop the shape. Whether you use big triangles or tiny ones, the final answer is the same.
They prove that by using their "Orbifold Completion," you can generate a universal 3D state sum model. This is a single mathematical formula that can describe:
- Standard 3D physics theories (like the Turaev-Viro model).
- Theories with "walls" and "pipes" (defects) running through them.
- Theories that connect different types of physics (Reshetikhin-Turaev theories).

6. The "Euler" Twist

The paper also mentions an "Euler completion."

The Analogy: Think of the Euler characteristic as a "counting number" for shapes (like how many corners and edges a shape has). Sometimes, the math works perfectly only if you add a tiny "correction factor" based on this count.
The authors show how to bake this correction factor directly into their machine, allowing it to handle even more complex scenarios, like those found in "Reshetikhin-Turaev" theories (which are used to study knots and quantum groups).

Summary

In plain English, this paper is a construction manual for the ultimate 3D Lego set.

They defined the rules for how 3D "defects" (special pieces) must behave to be stable.
They built a machine that automatically generates every possible stable configuration of these pieces.
They proved that once you build this set, you can't add anything new to it; it is mathematically "complete."
They showed that this set can be used to calculate physical properties of 3D shapes in a way that is robust and consistent, no matter how you look at them.

This work bridges the gap between abstract algebra (the rules of the game) and physical theories (the game itself), providing a unified framework to understand complex 3-dimensional quantum systems with defects.

Technical Summary: Orbifold Completion of 3-Categories

Problem Statement
Topological Quantum Field Theories (TQFTs) are traditionally defined as symmetric monoidal functors on bordism categories. While the theory of "generalised orbifolds" allows for the construction of new closed TQFTs from defect TQFTs by gauging specific defect labels, this construction has historically been limited to closed manifolds. The paper addresses the need to lift the orbifold construction from closed TQFTs to the richer domain of defect TQFTs, where manifolds are stratified and decorated with defects of various codimensions. Specifically, the authors aim to develop a general 3-categorical framework for "orbifold completion" that describes (generalised) orbifolds of 3-dimensional TQFTs and all their associated defects. This involves categorifying the 2-dimensional orbifold completion established in [CR] to the level of 3-categories, ensuring that the resulting structure admits adjoints for all 1- and 2-morphisms, a property essential for oriented theories.

Methodology
The authors work within the framework of "Gray categories with duals," a semistrict version of 3-categories with compatible adjoints for 1- and 2-morphisms, as defined in [BMS]. The methodology proceeds through the following steps:

Morita Categories of $E_1$ -Algebras: The authors first construct a 3-category $\mathcal{E}(\mathcal{T})$ for a given Gray category $\mathcal{T}$ . The objects of $\mathcal{E}(\mathcal{T})$ are unital $E_1$ -algebras in $\mathcal{T}$ . The 1-morphisms are bimodules over these algebras, and higher morphisms are bimodule maps. The composition of 1-morphisms is defined via relative tensor products, computed as 2-colimits (specifically, splitting of condensation 2-monads).
Definition of Orbifold Data: The authors define a "3-dimensional orbifold datum" as a specific type of $E_1$ -algebra in $\mathcal{T}$ satisfying a set of coherence conditions (labeled O1–O8 in Figure 4.1). These conditions categorify the relations of 2-dimensional $\Delta$ -separable symmetric Frobenius algebras and encode invariance under oriented 3-dimensional Pachner moves. Crucially, these conditions involve the identification of left and right adjoints via the pivotal structure of the Gray category.
Construction of $\mathcal{T}^{orb}$ : The orbifold completion $\mathcal{T}^{orb}$ is defined as a sub-3-category of $\mathcal{E}(\mathcal{T})$ . Its objects are the orbifold data defined above. Its 1-morphisms are bimodules satisfying "coloured" versions of the orbifold conditions (analogous to O1–O7), and its 2-morphisms satisfy further constraints (T1–T7) ensuring compatibility with the adjoint structure.
Proof of Duality: The core technical effort involves proving that $\mathcal{T}^{orb}$ admits adjoints for all 1- and 2-morphisms. This is achieved by explicitly constructing the adjoints using the graphical calculus of [BMS] and verifying the Zorro (snake) identities. The proof relies on the splitting of condensation 2-monads to compute relative products and the specific coherence axioms of the orbifold data.
Universal Property: The authors propose a universal property for orbifold completion in arbitrary dimension, characterizing it as the completion where all orbifold data split. They rigorously prove this property for the 2-dimensional case (recovering the result of [CR]) and outline its extension to dimension 3.

Key Contributions and Results

Theorem 1.1: The primary result establishes that for a Gray category with duals $\mathcal{T}$ (satisfying mild finiteness conditions on colimits), the constructed 3-category $\mathcal{T}^{orb}$ admits adjoints for all 1- and 2-morphisms. This confirms that the orbifold completion preserves the necessary duality structures for oriented TQFTs.
Universal Property (Proposition 4.17): The paper proves that the 2-dimensional orbifold completion $\mathcal{B}^{orb}$ satisfies a universal property: it is the unique pivotal 2-category where every orbifold datum splits, and any pivotal functor from $\mathcal{B}$ to a category where orbifold data split factors essentially uniquely through $\mathcal{B}^{orb}$ . The authors argue (without full proof for $n=3$ ) that $\mathcal{T}^{orb}$ satisfies an analogous universal property.
State Sum Models and Spherical Fusion Categories (Theorem 5.18): The authors apply their theory to the 2-category of separable symmetric Frobenius algebras ($ssFrob$). They establish a pivotal equivalence between $ssFrob$ and the 2-category of semisimple Calabi–Yau categories ($CYcats$). Consequently, they prove that the 3-category of spherical fusion categories ($sFus$), as defined in [Sc2], is a subcategory of the orbifold completion of the Euler completion of the delooping of $ssFrob$. This provides a new proof that spherical fusion categories are 3-dualisable in the 3-category of algebras in 2-vector spaces.
Reshetikhin–Turaev Defects (Theorem 5.21): The paper demonstrates that the defect TQFTs constructed in [KMRS] for Reshetikhin–Turaev theories based on modular fusion categories arise naturally as a subcategory of the orbifold completion of the 3-category of commutative $\Delta$ -separable Frobenius algebras. Specifically, the "Frobenius algebras over pairs of algebras" introduced in [KMRS] are identified as 1-morphisms in the orbifold completion.
Defect TQFT Construction (Theorem 6.4): The authors show that the orbifold construction lifts a defect TQFT $Z$ to a new defect TQFT $Z^{orb}$ . This new theory is defined on bordisms decorated with the orbifold data of the original theory. The evaluation of $Z^{orb}$ is performed by choosing a stratified triangulation, labeling the dual stratification with orbifold data, and taking a colimit. This construction recovers known state sum models (like Turaev–Viro–Barrett–Westbury) as special cases where the stratification is trivial.

Significance and Claims
The paper claims to provide a general, algebraic theory of 3-dimensional orbifold completion that unifies the description of defects in TQFTs. By constructing $\mathcal{T}^{orb}$ , the authors enable the systematic generation of new defect TQFTs from existing ones, extending the scope of orbifold constructions beyond closed manifolds.

The authors explicitly state that their work is a categorification of the 2-dimensional results found in [CR]. They argue that the "oriented" orbifold completion is richer than "framed" condensation completion (as discussed in [GJF]), particularly regarding the existence of homotopy fixed-point structures for $SO(n)$ actions.

The paper is modest regarding the full 3-dimensional universal property. While they propose a definition and prove it for $n=2$ , they acknowledge that rigorously constructing the 4-category of 3-categories with compatible adjoints required to prove the universal property for $n=3$ is a formidable task left for future work. Similarly, while they expect $(\mathcal{T}^{orb})^{orb} \cong \mathcal{T}^{orb}$ , they do not provide a proof for the 3-dimensional case, leaving it as a conjecture based on the 2-dimensional result.

The applications presented are strictly algebraic and theoretical, focusing on the structural properties of the resulting 3-categories and their relation to existing literature on state sum models and Reshetikhin–Turaev theories. The authors note that their construction of 4-dimensional state sum models from first principles, which relies on these 3-categorical results, will be detailed in a separate joint work [CMM].