No boundary density matrix in elliptic de Sitter dS/$\mathbb{Z}_2$

This article proposes that the Euclidean path integral on non-time-orientable elliptic de Sitter spacetime defines a no-boundary density matrix rather than a wave function, as demonstrated by the explicit calculation of entanglement entropies for free Dirac fermions, revealing a unique property in which the global Hilbert space is one-dimensional while the Hilbert spaces of individual observers remain nontrivial.

The Gist

The Big Picture: A Universe with a Twist

Imagine our universe as a huge, expanding balloon. In physics, we usually examine this balloon (called de Sitter space) as if it had a clear "front" and "back," a clear "past" and "future." One can move from the past to the future without ever being confused about which direction time flows.

However, this paper investigates a strange, twisted version of this universe. Imagine taking this balloon and gluing every single point on its surface to the point directly opposite it (the antipode). If you move forward in time, you suddenly find yourself on the other side of the universe, moving in reverse.

This creates a non-time-orientable universe. It is like a Möbius strip made of spacetime: if you travel far enough, you return to your starting point, but your clock runs backward. The authors call this elliptic de Sitter space.

The Problem: The "No-Boundary" Puzzle

In standard physics, when we want to describe the beginning of the universe (the "no-boundary" state), we use a mathematical tool called the path integral. Think of it like baking a cake:

• Standard Universe: You bake the cake, cut it in half, and look at one half. This half represents the "wave function" (a complete description of the universe's state). It is like a clear recipe for the whole cake.
• Twisted Universe (Elliptic): Since the universe is glued together in this strange Möbius-strip way, you cannot cleanly cut it in half. There is no "front" and "back" that can be separated. If you try to bake the cake using the standard recipe, chaos ensues. You cannot define a single "wave function" for the entire universe because the universe lacks a consistent time direction to define one.

The Solution: The "Density Matrix" Cake

The authors propose a clever detour. Since we cannot bake a single, perfect "wave function" cake for the entire twisted universe, let us stop trying to describe the whole thing all at once.

Instead, they propose that the mathematics actually describes a density matrix.

• The Analogy: Imagine you are in a room with a foggy window. You cannot see the entire garden outside (the global wave function), but you can recognize a specific patch of flowers through your window (the view of a local observer).
• The Claim: The mathematics in this twisted universe does not provide the recipe for the entire garden. Instead, it provides a statistical description of what a single observer sees. It is like a "blurry photo" of the universe that is perfectly valid for someone standing in one place, even if it makes no sense for the universe as a whole.

They call this the "No-Boundary Density Matrix." It is a way to describe the state of the universe without requiring a global "past" or "future" to exist first.

The Experiment: Calculating Entanglement

To prove that this idea works, the authors performed a complex calculation using a simplified model: a 2D universe filled with freely floating particles (fermions).

1. The Setup: They treated the twisted universe as a non-orientable surface (like a Klein bottle or a real projective plane).
2. The Calculation: They calculated something called entanglement entropy.
   • Simple Analogy: Imagine two friends, Alice and Bob, sharing a secret code. Entanglement entropy measures how much of this code is shared between them. If they share everything, the entropy is high. If they share nothing, it is low.
3. The Result: They found that for an observer looking at a small patch of this twisted universe, the "entanglement" behaves in a very specific, predictable way.
   • Key Insight: The larger the patch of the universe an observer considers, the more the "entanglement entropy" explodes (goes to infinity).
   • What This Means: This confirms that you cannot describe the entire twisted universe as a single, pure, perfect state. The "whole picture" is fundamentally broken or undefined, supporting their idea that we must instead use the "density matrix" (the local, blurry view).

The "One-State" Universe vs. the "Many-States" Observer

The paper ends with a fascinating paradox regarding the "size" of the universe's possibilities.

• The Global View: If you try to describe the entire twisted universe all at once, the mathematics says there is only one possible state. It is like a room with only one chair; there is no room for variation. The global Hilbert space (the list of all possible universes) is one-dimensional.
• The Local View: However, if you are a single observer living in this universe, you see a rich, complex world with infinite possibilities (a Fock space). You can have particles, energy, and motion.

The Conclusion: The universe as a whole is "empty" of variation due to its twisted geometry, but every single observer within it experiences a full, bustling reality. The "density matrix" is the mathematical bridge that allows us to describe this bustling local reality without being confused by the empty global reality.

Summary

This paper argues that in a universe where time repeats itself (elliptic de Sitter space), we cannot define a single, global "state of the universe." Instead, the mathematics naturally generates a statistical description (density matrix) that is valid for local observers. They proved this by calculating how "connected" different parts of such a universe are, showing that the global view is fundamentally undefined, while the local view is rich and complex.

Technical Summary

Technical Summary: No-Boundary Density Matrix in Elliptic de Sitter Space dS/Z₂

Problem Statement
The article addresses the challenge of defining quantum states in elliptic de Sitter space (dS), denoted as $dS/Z_2$. While global de Sitter space admits a well-defined "No-Boundary" wavefunction prepared by a Euclidean path integral over a sphere ($S^{d+1}$), its antipodally identified counterpart, elliptic de Sitter space, corresponds in Euclidean signature to the real projective plane $RP^{d+1}$. Unlike the sphere, $RP^{d+1}$ lacks time-reversal symmetry that would separate the manifold into two disjoint hemispheres. Consequently, the Euclidean path integral over $RP^{d+1}$ cannot be interpreted as preparing a wavefunction $|\Psi\rangle$ on a spatial slice. This creates a gap in understanding the quantum state of the system and the nature of correlations for local observers in this non-time-orientable spacetime.

Methodology
The authors propose a field-theoretic framework to resolve this problem without resorting to dynamical gravity. The core methodology includes:

1. Proposal of a No-Boundary Density Matrix: Instead of a wavefunction, the authors propose that the Euclidean path integral over $RP^{d+1}$ defines a No-Boundary Density Matrix $\rho_A$ for a spatial subset $A$. This is constructed by cutting a slit along the subset $A$ at the equator ($\theta=0$) of the Euclidean manifold, imposing boundary conditions on both sides of the cut, and performing the path integral. This reflects the Ivo-Li-Maldacena (ILM) proposal for gravitational path integrals but is here applied to a fixed background geometry.
2. Replica Trick on Non-Orientable Surfaces: To characterize the spectrum of this density matrix, the authors calculate the Rényi entropies $S^{(n)}_A$ using the replica trick. This requires computing the partition function on an $n$-sheeted manifold $\Sigma_n$, constructed by gluing $n$ copies of $RP^2$ along the slit.
3. Analysis of Conformal Field Theory (CFT): The authors investigate the free Dirac fermion CFT in two dimensions ($d=1$) as an explicit example.
   • They utilize Bosonization to express the theory in terms of free bosons.
   • They identify the replica manifold as a non-orientable surface and use Crosscap States ($|C\rangle$) to represent the antipodal identification.
   • They derive the two-point functions of twist operators (which implement the replica boundary conditions) on non-orientable surfaces (Klein bottle $K^2$ and $RP^2$).
   • Crucially, they establish compatibility conditions between the sign choices of the twist operators (Dirichlet vs. Neumann boundary conditions) and the crosscap states ($|C_+\rangle$ vs. $|C_-\rangle$) to ensure valid correlation functions.

Main Contributions and Results

• Definition of the No-Boundary Density Matrix: The article formally defines the No-Boundary Density Matrix for elliptic de Sitter space and provides a concrete interpretation for the Euclidean path integral on $RP^{d+1}$, where a wavefunction interpretation fails.
• Analytic Calculation of Entropies: The authors analytically calculate the von Neumann and Rényi entropies for the free Dirac fermion CFT.
  • At $t=0$ (Equator): For a subsystem $A$ approaching the entire spatial slice, the entanglement entropy diverges. This indicates that the state on the entire spatial slice is not pure, consistent with the inability to define a global No-Boundary wavefunction.
  • Time Evolution: The authors examine the real-time evolution of entanglement entropy for two scenarios:
    1. Fixed Proper Length: The entropy remains constant and respects the symmetries of $dS_2/Z_2$.
    2. Co-expanding Subsystem: For a subsystem with fixed coordinate size, the entropy exhibits a phase transition when the subsystem crosses the "cosmological horizon" of a static observer. Beyond this point, the subsystem becomes timelike separated from itself, and the interpretation of the density matrix collapses (the argument of the entropy formula becomes negative).
• One-Dimensional Global Hilbert Space: In Section 5, the authors revisit canonical quantization in Lorentzian signature. They show that due to the lack of global time orientation, creation and annihilation operators are globally indistinguishable. Consequently, the global Hilbert space collapses to a one-dimensional space (containing only the vacuum). However, they demonstrate that for any local static observer (who possesses a well-defined time orientation within their static patch), a non-trivial Fock space can be constructed. This underscores an observer-dependent structure of the Hilbert space.
• Crosscap Quench Dynamics: As a byproduct, the authors apply their formalism to the "Crosscap Quench" in the free Dirac fermion CFT. They derive the time evolution of the entanglement entropy and show that the initial crosscap state behaves like an entangled antipodal pair state (EAP), initially exhibiting volume-law scaling, followed by oscillatory behavior with period $\pi$.

Significance and Claims
The article claims to provide a purely field-theoretic realization of the No-Boundary Density Matrix and offers a concrete model for how quantum information is encoded in non-time-orientable spacetimes without invoking dynamical gravity. The authors emphasize that although the global Hilbert space of elliptic de Sitter space is trivial (one-dimensional), local observers still perceive a rich quantum structure (non-trivial Fock spaces).

The work suggests that the phenomenon of a trivial global Hilbert space in $dS/Z_2$ mirrors parallel arguments in quantum gravity regarding closed universes, where non-perturbative effects lead to a one-dimensional Hilbert space, while observer-dependent algebras restore non-trivial dimensions. The authors position their work as a starting point for understanding entanglement structures on non-orientable surfaces and as a potential building block for future gravitational realizations of the No-Boundary Density Matrix, explicitly stating that a complete gravitational realization remains reserved for future work.