Boundary conditions and Hilbert spaces in no-roll quantum cosmology

This paper constructs Hilbert spaces for minisuperspace quantum cosmology in the extreme slow-roll limit, demonstrating that fixing the potential energy yields a one-dimensional physical space favoring Vilenkin's tunnelling wavefunction, while allowing it to vary leads to an infinite-dimensional space where self-adjointness imposes boundary conditions that generally mix Hartle-Hawking and tunnelling wavefunctions, with a specific choice nearly recovering the Hartle-Hawking state.

The Gist

Imagine the entire Universe as a giant, expanding balloon. In the field of quantum cosmology, scientists try to figure out the "rules of the game" for how this balloon started and what its initial state looked like. This paper by Steffen Gielen tackles a specific, simplified version of this problem: a Universe that is perfectly round and smooth, filled with a special energy field that isn't moving or changing much (a "no-roll" limit).

Here is the breakdown of the paper's ideas using simple analogies:

1. The Two Ways to Look at the Universe

The author explores two different ways to set up the mathematical "game" for this Universe, which leads to two very different answers about how many possible states the Universe can be in.

Scenario A: The Fixed Recipe (Fixed Initial Conditions)
Imagine you are baking a cake, and you have already decided exactly how much sugar and flour to use. The recipe is fixed.

• The Paper's Claim: If you fix the energy of the Universe (like fixing the sugar amount), the math says there is essentially only one valid way the Universe can exist. It's like a movie that has only one single frame that is allowed to play.
• The Result: In this scenario, the "Hilbert space" (a fancy math term for the list of all possible states) is one-dimensional. There is no probability, no "maybe this, maybe that." The Universe just is in that one state. This aligns with some recent, radical ideas in physics suggesting a closed Universe has only one possible state.

Scenario B: The Open Menu (Arbitrary Initial Conditions)
Now, imagine you don't know how much sugar to use. You want to consider every possible amount of sugar, from a tiny pinch to a massive pile.

• The Paper's Claim: If you allow the energy to vary (like having an open menu), the math changes completely. Now, the Universe is like a piano with infinite keys. Each key represents a different energy level.
• The Result: This creates an infinite-dimensional Hilbert space. There are infinitely many possible states the Universe could be in, corresponding to different energy levels.

2. The Problem of the "Singularity" (The Zero Point)

In these equations, there is a point where the size of the Universe is zero ($a=0$). In classical physics, this is a "singularity"—a point where the math breaks down, like a black hole or the Big Bang moment.

To make the math work properly (specifically, to ensure the physics is "self-adjoint," which is a technical way of saying the laws of physics remain consistent and don't lose information), the author argues we must set a rule for what happens at this zero point.

• The Analogy: Think of a guitar string. To get a clear note, the string must be attached to the bridge in a specific way. If it's loose, the sound is garbage. If it's tied too tight, it snaps. You need a specific "boundary condition" to make the music work.
• The Paper's Discovery: The author finds that there isn't just one way to tie the string. There is a family of rules (a one-parameter family) you can use to attach the wavefunction at the zero point. This is a generalization of an old idea by physicist Bryce DeWitt, who suggested the wavefunction should just be zero at the start. The author shows that while DeWitt's rule is one option, there are many others that also work mathematically.

3. Choosing the "Right" Universe

Once you have these infinite possibilities and the rules for how they behave at the start, which one describes our Universe?

• The "No-Boundary" vs. "Tunnelling" Debate: For decades, physicists have debated between two famous candidates for the Universe's starting wavefunction:

  1. Hartle-Hawking (No-Boundary): Like a smooth, rounded hill that has no sharp edge at the start.
  2. Vilenkin (Tunnelling): Like a particle tunnelling through a wall to appear out of nothing.

• The Paper's Insight: The author shows that the mathematical rules required to keep the physics consistent (self-adjointness) force the Universe to be a mixture of these two candidates. You can't have just one; you need a blend.

• The "Almost" Winner: However, in the regime that looks like our real Universe (where the energy is positive but small), the author finds a specific rule that makes the mixture look almost exactly like the Hartle-Hawking (No-Boundary) wavefunction. The other part of the mixture is so tiny (exponentially suppressed) that it's practically invisible once the Universe gets big.

4. The Probability Puzzle

Finally, the paper asks: "If we have all these possible energy levels, which one is most likely?"

• The Path Integral Problem: When scientists try to calculate probabilities using "path integrals" (summing up all possible histories), they often run into trouble.
  • If they try to predict the No-Boundary state, the math suggests the Universe is most likely to have zero energy, which doesn't match our reality.
  • If they try to predict the Tunnelling state, the math suggests the Universe should have the maximum possible energy, which also doesn't match reality.
• The Conclusion: The author concludes that simply building a Hilbert space of these states doesn't magically solve the mystery of why our Universe has the specific energy it does. The math still struggles to pick a "winner" without adding extra, arbitrary rules (cutoffs) to stop the numbers from blowing up.

Summary

In short, this paper says:

1. If you fix the Universe's energy, there is only one possible state.
2. If you let the energy vary, there are infinitely many states.
3. To make the math work at the very beginning (the Big Bang), you must impose a specific boundary rule.
4. This rule naturally leads to a mix of the two most famous theories about the Universe's start, but one specific rule makes the "No-Boundary" theory look like the clear winner for our actual Universe.
5. Despite this progress, the paper admits that we still can't easily predict why the Universe has the specific energy it does just by looking at these mathematical rules.

Technical Summary

Technical Summary: Boundary conditions and Hilbert spaces in no-roll quantum cosmology

Problem Statement
The paper addresses the construction of physical Hilbert spaces for minisuperspace quantum cosmology, specifically for a closed Universe ($k=1$) in the "no-roll" limit where a scalar field $\phi$ is approximated as constant ($\phi = \phi_*$). In this regime, the scalar potential $V(\phi_*)$ acts as an integration constant analogous to the cosmological constant in unimodular gravity. The central tension lies in interpreting the Wheeler–DeWitt (WDW) equation:

1. If $V(\phi_*)$ is fixed, recent literature suggests the physical Hilbert space of closed universes is one-dimensional, rendering probabilistic statements about the scale factor $a$ meaningless.
2. If $V(\phi_*)$ is treated as a dynamical variable (an integration constant determined by initial conditions), the theory requires an infinite-dimensional Hilbert space of energy eigenstates.
   The paper investigates how to define a consistent inner product and boundary conditions in both scenarios, particularly focusing on the self-adjointness of the effective Hamiltonian and the resulting implications for the Hartle–Hawking (no-boundary) and Vilenkin (tunnelling) wavefunctions.

Methodology
The author employs a canonical quantization approach within the minisuperspace framework, utilizing the metric $ds^2 = -N^2(t)dt^2 + a^2(t)d\Omega^2$.

• Action and Hamiltonian: Starting from the Einstein–Hilbert action with a scalar field, the no-roll limit ($\dot{\phi} \approx 0$) reduces the system to pure gravity with a cosmological constant $\Lambda = V(\phi_*)$. The author derives the constrained Hamiltonian for fixed $V(\phi_*)$ and an unconstrained Hamiltonian $H_{UG}$ (analogous to unimodular gravity) when $V(\phi_*)$ is treated as an eigenvalue.
• Operator Ordering: The WDW equation is formulated using a specific operator ordering ($N=1/a$) that yields solutions in terms of Airy functions, $Ai(z)$ and $Bi(z)$. This ordering is chosen to connect with standard literature on no-boundary and tunnelling proposals.
• Hilbert Space Construction:
  • Case 1 (Fixed $V$): The author constructs the physical Hilbert space using the Klein–Gordon conserved current as an inner product.
  • Case 2 (Arbitrary $V$): The author treats $V(\phi_*)$ as the eigenvalue of an effective Hamiltonian $\hat{H}_S$. To define a unitary theory, $\hat{H}_S$ must be a self-adjoint operator. The author analyzes the deficiency indices of this operator to determine the necessary boundary conditions at the singularity $a=0$.
• Boundary Conditions: The requirement of self-adjointness leads to a one-parameter family of Robin-type boundary conditions at $a=0$, generalizing the DeWitt criterion ($\Psi(0)=0$).

Key Contributions and Results

1. Fixed Initial Conditions (One- or Two-Dimensional Hilbert Space):

   • When $V(\phi_*)$ is fixed, the WDW equation admits two independent solutions. Using the Klein–Gordon inner product, only Vilenkin's tunnelling wavefunction (a specific complex combination of Airy functions) has a positive norm; the other solution has negative norm and is discarded. This yields a one-dimensional physical Hilbert space, consistent with recent claims that closed universes possess a single physical state.
   • Alternatively, if a positive-definite inner product is chosen (e.g., via group averaging), a two-dimensional Hilbert space emerges, allowing for superpositions of expanding and contracting de Sitter solutions, including the real Hartle–Hawking state.

2. Arbitrary Initial Conditions (Infinite-Dimensional Hilbert Space):

   • When $V(\phi_*)$ is allowed to vary, the system is described by an unconstrained Hamiltonian $\hat{H}_S$. The physical Hilbert space becomes infinite-dimensional, spanned by eigenstates corresponding to different values of $V(\phi_*)$.
   • Self-Adjointness and Boundary Conditions: For $\hat{H}_S$ to be self-adjoint, wavefunctions must satisfy a boundary condition at $a=0$ of the form $\Psi(0) = \gamma \lim_{a\to 0} \frac{1}{a}\Psi'(a)$. This condition is required for unitarity in the unimodular time conjugate to the potential energy, rather than for singularity resolution.
   • Selection of Wavefunctions: The boundary condition does not strictly select either the Hartle–Hawking or Vilenkin wavefunction; it generally requires a mixture of both. However, in the physically relevant regime ($0 < V(\phi_*) \ll 1$ in Planck units), there exists a specific choice of the parameter $\gamma$ (specifically $\gamma \approx 1/12\pi^2$) that makes the physical wavefunctions "almost" identical to the Hartle–Hawking no-boundary wavefunction. The deviation involves terms proportional to the Airy function $Bi(z)$, which are exponentially suppressed near $a=0$ and negligible for macroscopic scales.

3. Probabilistic Interpretation and Path Integrals:

   • The paper analyzes the probability distribution of $V(\phi_*)$ derived from the constructed Hilbert space.
   • Attempting to reproduce the standard no-boundary result (where the probability is weighted by $\exp(12\pi^2/V)$) leads to a non-normalizable state with severe infrared (IR) divergences as $V \to 0$. This suggests that the traditional path-integral interpretation of the no-boundary wavefunction as a probability distribution for $V$ is not predictive without an arbitrary IR cutoff.
   • Similarly, the tunnelling wavefunction (weighted by $\exp(-12\pi^2/V)$) leads to a distribution peaked at the ultraviolet (UV) cutoff, predicting a large cosmological constant inconsistent with observation.
   • The author concludes that the rigorous Hilbert space construction does not resolve the ambiguity in selecting initial conditions via path integrals; the natural basis states behave effectively as plane waves at large $a$.

Significance and Claims
The paper claims to provide a novel angle on constructing physical Hilbert spaces in quantum cosmology based strictly on the principles of quantum mechanics (unitarity and self-adjointness) without invoking a specific theory of quantum gravity.

• It reconciles the one-dimensional Hilbert space claim for closed universes with the traditional minisuperspace view by distinguishing between fixed and arbitrary initial conditions.
• It demonstrates that the requirement of a self-adjoint Hamiltonian naturally leads to a boundary condition at the singularity that generalizes the DeWitt criterion.
• It offers a new criterion for the preference of the Hartle–Hawking wavefunction: it is not strictly selected, but it is the "almost" unique solution in the low-energy regime for a specific self-adjoint extension, with corrections that are exponentially small.
• The work highlights that the standard probabilistic interpretations of the no-boundary and tunnelling proposals face significant challenges regarding normalizability and cutoff dependence when viewed through the lens of a rigorous Hilbert space construction.

The author remains modest, noting that these results are specific to the minisuperspace approximation and that the presence of local degrees of freedom in full gravity would significantly complicate the construction, though the embedding into unimodular gravity suggests a generalizable framework for defining unitarity.