Conditional Cosmological Recurrence in Finite Hilbert Spaces and Holographic Bounds Within Causal Patches

Abstract

A conditional framework of Conditional Cosmological Recurrence (CCR) is introduced, as follows: if a causal patch admits a finite operational Hilbert space dimension D (as motivated by holographic and entropy bounds), then unitary quantum dynamics guarantee almost-periodic evolution, leading to recurrences. The central contribution is the explicit formulation of a micro-to-macro bridge, as follows: (i) finite regions discretize field modes; (ii) gravitational bounds cap entropy and energy; and (iii) the number of accessible states is finite, yielding CCR. The analysis differentiates global microstate recurrences (with double-exponential timescales in $S_{max}$) from operationally relevant coarse-grained returns (exponential in subsystem entropy), with conservative timescale estimates. For predictivity in eternally inflating settings, a causal-diamond measure with xerographic typicality and a single no-Boltzmann-brain constraint is employed, thereby avoiding volume-weighting pathologies. The scope is explicitly conditional: if future quantum gravity demonstrates $D=\infty$ for causal patches, CCR is falsified.

1. Introduction

Cosmology often examines regions of the universe known as causal patches—the finite portion of spacetime that a single observer can, in principle, access. Within such a patch, gravity imposes strict limits on the amount of information that can be stored. These limits, expressed through holographic or entropy bounds [1,2,3,4,5], imply that only a limited number of physically distinct configurations may exist. This constraint gives rise to the phenomenon of recurrence [6]: if the state space is bounded and the system evolves according to unitary quantum mechanics, then it must eventually return arbitrarily close to an earlier configuration in the finite-D case. In practice, such recurrences occur on unimaginably long timescales, but the implication remains important: the universe may not evolve in endlessly novel ways but could instead recycle through states. This possibility is addressed in a strictly conditional sense: recurrence follows mathematically if holographic bounds imply a finite state budget for causal patches [7,8,9,10,11,12].

• Classical intuition (well-known example).

The phenomenon of recurrence is not new; it is a classical consequence of the Poincaré Recurrence Theorem (1890). A standard textbook illustration considers an isolated box containing a macroscopic object, such as an apple. Hamiltonian dynamics on a bounded phase space imply that—after an astronomically long time—the microstate of the system will return arbitrarily close to its initial configuration; in other words, the apple would eventually “reassemble.” Although not original to this study, this example serves as an intuition pump for the general recurrence phenomenon.

• From boxes to causal diamonds.

In a cosmological setting, the “box” is replaced by a causal patch: a finite region bounded by horizons, whose holographic entropy cap [13,14] implies a finite-dimensional Hilbert space. The CCR framework is then the direct quantum analogue of the recurrence principle in such bounded systems [15,16,17].

• Conventions.

Throughout this work, we employ natural units with $\hslash =c=1$, while keeping $k_B$ and G explicit to emphasize entropic and gravitational scales. The Planck length is ${\mathcal{l}}_P=\sqrt{\hslash G/c^3}\equiv \sqrt{G}$ in these units. In certain expressions (e.g., Lemma 2), we retain explicit factors of c and ℏ to facilitate comparison with the standard literature and to clarify the physical dimensions involved.

• Novel contributions.

While the core result—that finite Hilbert spaces imply quantum recurrence—is well-established, this work makes the following three complementary contributions:

1.

Explicit micro-to-macro bridge. A concrete microcanonical counting under gravitational caps (Proposition 1, Appendix C) demonstrates how infrared discretization and holographic bounds enforce bounded operational dimension, making finiteness tangible rather than heuristic.

2.

Conditional and falsifiable framing. The CCR theorem (Theorem 1) is stated in strictly conditional form: if causal patches admit finite D (assumption (A1)), recurrence follows rigorously; and if future quantum gravity shows $D=\infty$, CCR is falsified. This clean if–then structure emphasizes testability.

3.

Minimalistic measure prescription. A compact causal-diamond measure with xerographic typicality and a single no-Boltzmann-brain constraint (Section 4.1) avoids volume-weighting pathologies while remaining falsifiable.

2. Materials and Methods

2.1. Originality and Scope

The core idea that finite Hilbert spaces imply quantum recurrences is not new; it goes back to Bocchieri–Loinger (1957) and has been discussed in cosmological contexts (e.g., Dyson–Kleban–Susskind 2002). The contribution does not lie in proposing a new recurrence law, but in the following three complementary aspects:

1.

Explicit constructive micro-to-macro derivation. A concrete microcanonical counting argument under a gravitational energy cap (Proposition 1, Appendix C) is presented, showing explicitly how infrared discretization and holographic entropy bounds combine to enforce a bounded operational Hilbert space dimension. This makes the finiteness assumption tangible rather than heuristic.

2.

Conditional and falsifiable framing. The “Conditional Cosmological Recurrence” (CCR) theorem is stated in a strictly conditional form: if causal patches admit a finite Hilbert space dimension (assumption (A1)), then recurrence follows rigorously from unitarity. Conversely, if future quantum gravity demonstrates that $D=\infty$, the entire framework is falsified. This clean if–then structure is rarely made explicit in the prior literature and emphasizes testability.

3.

Minimalistic measure prescription. To address predictivity in eternally inflating settings, a compact causal-diamond measure is introduced, combined with xerographic typicality and a single no-Boltzmann-brain constraint. This avoids the volume-weighting pathologies highlighted in earlier works while keeping the prescription lean and falsifiable.

In short, the contribution is not the discovery of recurrence per se, but the systematic construction of a finite-state framework (micro-to-macro counting), its conditional and testable formulation (CCR theorem), and a clean measure proposal that, together, yield a more rigorous and falsifiable picture of recurrence in cosmological settings. Figure A1 in Appendix J summarizes the logical structure of this framework.

Key Point: The framework does not claim that the universe must recur. Rather, if gravitational entropy bounds apply to causal patches (yielding a finite Hilbert space dimension), then recurrence follows mathematically. The “if” is the physics question; the “then” is rigorous mathematics.

• Conditional Nature.

Cosmological Recurrence (CCR) is not a prediction of standard cosmology per se, but a logical consequence of the finite Hilbert space conjecture for de Sitter patches (A1). Consequently, the results are strictly conditional: falsifying assumption(A1 would directly falsify CCR. In this sense, the framework serves as a logical bridge between holographic bounds and recurrence phenomena, rather than as an independent dynamical model. If future developments in quantum gravity demonstrate that the Hilbert space of a causal patch is in fact infinite, the CCR framework would be invalidated in its entirety, and the recurrence problem would have to be reconsidered from a fundamentally different perspective.

• Separation of rigor and assumptions.

The mathematical content of this work (recurrence theorems, almost-periodicity) is rigorous, following from finite-dimensional Hilbert spaces and unitary dynamics. The physics content (finite Hilbert space dimension via holographic bounds) is an assumption, motivated by semiclassical gravity but not derived from first principles. Section 4.4 provides an explicit separation of these two levels.

2.2. From Electrons to a Finite Cosmic State Count: A Counting Argument

Lemma 1 (IR discretization). 

A quantum field in a finite spatial region of linear size R admits a discrete set of momentum modes $\mathbf{k}={\displaystyle \frac{\pi }{R}}(n_x,n_y,n_z)$ under standard boundary conditions.

Lemma 2 (UV/gravity cutoff). 

Let $E_{\mathrm{BH}}\left(R\right)={\displaystyle \frac{c^{4}R}{2G}}$ be the energy of a Schwarzschild black hole of radius R. Any state whose total energy localized within the region exceeds $E_{\mathrm{BH}}\left(R\right)$ collapses, and the Bekenstein–Hawking bound yields

$$S_{\max }\le {\displaystyle \frac{k_{B}A}{4{\mathcal{l}}_P^2}}={\displaystyle \frac{k_{B}4\pi R^2}{4{\mathcal{l}}_P^2}},{\mathcal{l}}_P^2={\displaystyle \frac{\hslash G}{c^3}}.$$

(1)

Proposition 1 (Finite-dimensional accessible Hilbert space). 

Consider the Fock space built from the discrete IR modes and impose the microcanonical cap ${\sum }_{\mathbf{k}}{n}_{\mathbf{k}}\hslash {\omega }_{\mathbf{k}}\le E_{\mathrm{BH}}\left(R\right)$. Then, the number of distinct Fock states below this cap is finite. Moreover, any operationally distinguishable state family within the region obeys $logD\le S_{\max }/k_B\lesssim A/\left(4{\mathcal{l}}_P^2\right)$, so the accessible Hilbert space dimension D is finite. Here, D denotes the effective Hilbert space dimension1.

(See Appendix C for an explicit microcanonical counting illustration under a gravitational cap).

Proof of Sketch. 

Finite volume ⇒ discrete k (Lemma 1). The energy cap forbids arbitrarily many high-frequency quanta (Lemma 2). Since ${\omega }_k\ge c\left|k\right|$ and the number of modes below any cutoff $\Lambda$ is finite, the integer solutions $\left\{n_k\right\}$ of $\sum n_k\hslash {\omega }_k\le E_{\mathrm{BH}}$ are finite. While derived here for free fields, the finiteness conclusion extends to weakly interacting theories: interactions modify the state space structure but cannot generate infinite independent states below the gravitational bound. Operational distinguishability is limited by the entropy bound, giving $logD\le S_{max}/k_B$. □

Physical intuition: A finite spatial region implies discrete field modes (analogous to a particle in a box). Gravitational collapse bounds the total energy—too much energy in a region would form a black hole. These two constraints together limit the total number of quantum states, implying a finite Hilbert space dimension D.

• Clarification: Two independent arguments.

This proposition employs the following two logically independent arguments:

1.

Microcanonical finiteness: Finite spatial volume ⇒ discrete modes (IR), and gravitational collapse threshold $E_{\mathrm{BH}}\left(R\right)$ ⇒ finite occupation (UV). Together, these imply the accessible Fock space is finite-dimensional.

2.

Holographic bound: The area-law entropy bound $S\le S_{max}\sim A/\left(4{\mathcal{l}}_P^2\right)$ (from black-hole thermodynamics and holographic arguments) provides an independent constraint on operational distinguishability: $logD\lesssim S_{max}/k_B$.

The microcanonical argument does not derive the area-law scaling—that is an additional input from holography. Both arguments point to finite D, and the area-law bound sets the operational scale.

• Clarification: Three notions of finiteness.

Before stating the formal assumptions, it is useful to distinguish the following three conceptually distinct notions of finiteness that appear in discussions of quantum gravity and cosmology:

1.

Kinematical finiteness: The spacetime itself is fundamentally discrete at the Planck scale, as in causal set theory [18] or loop quantum gravity. In such approaches, the continuum is an emergent, coarse-grained description, and the number of degrees of freedom in any finite region is strictly finite by construction.

2.

Operational finiteness: Even in a continuum quantum field theory, the number of operationally distinguishable states within a region is finite when measured to a fixed resolution $\delta$ in trace distance. This counting is made precise via metric entropy and packing arguments (see Appendix A), and yields an effective dimension $D_{\mathrm{eff}}\left(\delta \right)$ that is finite for any $\delta >0$, with $log{D}_{\mathrm{eff}}\left(\delta \right)\lesssim S_{max}/k_B+O(log(1/\delta ))$.

3.

Thermodynamic finiteness: Holographic entropy bounds impose an upper limit on the von Neumann entropy accessible within a causal patch: $S\le S_{max}\sim k_{B}A/\left(4{\mathcal{l}}_P^2\right)$. This constrains the dimension of the effective Hilbert space via $D\lesssim exp(S_{max}/k_B)$, which is finite (though extremely large) if A is finite.

Scope of this work. The CCR framework relies primarily on thermodynamic finiteness (item 3), motivated by holographic bounds, and employs operational finiteness (item 2) to make the notion of effective dimension D precise. We do not assume kinematical discreteness (item 1), though the conclusions remain compatible with such approaches. In what follows, assumption (A1) formalizes thermodynamic finiteness as the starting point for the recurrence theorem.

2.3. Assumptions and Definitions

Consider a causal patch, defined as the region causally connected to an observer, with horizon area A, adopting the following:

(A1) 

Finite information bound: $S_{max}\le k_{B}A/4{\mathcal{l}}_P^2$, implying $dim\mathcal{H}=D<\infty$ (motivated by Bekenstein–Hawking/Gibbons–Hawking bounds and recent discussions of finite Hilbert space in de Sitter [11,19,20]).

(A2) 

Unitary dynamics: $|\psi \left(t\right)\rangle =e^{-iHt}|\psi \left(0\right)\rangle$, with self-adjoint Hamiltonian H.

(A3) 

Sector mixing (optional): Dynamics are not confined to a null-measure subset within the relevant superselection sector.

(A4) 

Finite-resolution observers: Macrostates identified up to a trace-distance tolerance $\epsilon >0$ on local density matrices.

Remark 1 (Spectral assumptions). 

The recurrence theorem relies on the following properties of the Hamiltonian spectrum, which follow from (A1) and (A2):

1. 

Discrete spectrum:  Finite-dimensional Hilbert space ⇒ purely discrete spectrum ${\left\{E_j\right\}}_{j=1}^D$.

2. 

Non-degeneracy:  For generic Hamiltonians, eigenvalues are distinct. Degeneracies, if present, do not obstruct recurrence [6].

3. 

Absence of continuous spectrum:  No accumulation points in $\mathrm{spec}\left(H\right)$ below the gravitational cap $E_{\mathrm{BH}}\left(R\right)$.

4. 

Diophantine gaps:  Quantitative lower bounds on recurrence times rely on the energy differences $\{E_j-E_k\}$, avoiding low-order rational relations (Diophantine conditions). For generic quantum systems, the spectrum is expected to be linearly independent over the rationals. Under such conditions, rigorous theorems on simultaneous Diophantine approximation [21,22] ensure that the return time scales at least exponentially with the dimension D.

2.4. CCR Theorem and Coarse-Grained Recurrence

Theorem 1 (Conditional cosmological recurrence (CCR)). 

Under (A1) and (A2), for any initial pure state $\left|\psi \right(0)\rangle$ and any $\epsilon >0$, there exists T, such that

$$F\left(T\right):={\left|\langle \psi \left(0\right)|\psi \left(T\right)\rangle \right|}^2>1-\epsilon .$$

(2)

Moreover, recurrence times obey the lower-bound behavior in Remark 2.

Proof of Sketch. 

A finite D implies a discrete spectrum and almost–periodic evolution (Bocchieri–Loinger [6]). Lower bounds on recurrence times follow from Diophantine properties of spectral gaps, namely the degree to which the energy differences $\{E_j-E_k\}$ avoid low-order rational relations. Quantitative control relies on simultaneous Diophantine approximation on tori (Kronecker-type theorems) and metric results in number theory; see Cassels [21] and Schmidt [22]. See also Remark 2. □

Bocchieri–Loinger theorem2

• Recurrence time (careful statement).

The term “recurrence time” is employed here in the usual almost-periodicity sense. In a model-independent manner, generic recurrence times grow at least exponentially with the Hilbert space dimension D:

$$T_{\mathrm{rec}}\gtrsim t_{\mathrm{micro}}exp\left(cD\right),$$

for some $c=O\left(1\right)$ that depends on the Diophantine properties of spectral gaps and on the chosen tolerance. While rigorous number-theoretic bounds exist (see Remark 1), the specific prefactors are model-dependent heuristic estimates. Since $D\lesssim exp(S_{\max }/k_B)$ by the holographic/entropic cap, a conservative expectation is3

$$T_{\mathrm{rec}}\gtrsim exp\left\{\alpha exp(S_{\max }/k_B)\right\},$$

i.e., double-exponential in the entropy (up to polynomial prefactors on microscopic scales). For illustration, even a toy system with dimension $D={10}^3$ already yields a recurrence time of order $T_{\mathrm{rec}}\sim e^{{10}^3}\approx {10}^{434}$ steps, while for cosmological entropies ($D\sim {10}^{120}$), the scale becomes incomprehensibly large ($T_{\mathrm{rec}}\sim e^{{10}^{120}}$).

Remark 2 (Global vs. local/coarse-grained returns). 

While the heuristic estimate $T_{\mathrm{rec}}\sim exp\left(D\right)$ is standard in the physics literature, rigorous lower bounds for almost-periodic functions dictate that for a generic Hamiltonian satisfying Diophantine spectral conditions, the return time $T_{ϵ}$ for tolerance ϵ obeys $T_{ϵ}\ge C\left(ϵ\right)exp\left(cD\right)$ [21]. Since $D\sim e^{S_{\max }}$, this confirms the double-exponential scaling, as follows:

$$T_{\mathrm{rec}}^{\left(\mathrm{global}\right)}\gtrsim exp\left\{\alpha exp\left(\frac{S_{max}}{k_B}\right)\right\}.$$

such global microstate recurrences occur on timescales vastly exceeding any operational significance and are, therefore, physically meaningless beyond their role as mathematical consequences of finiteness. By contrast, coarse-grained or subsystem recurrences scale only single-exponentially with the entropy of the subsystem,

$$T_{\mathrm{rec}}^{\left(A\right)}\left(\epsilon \right)\sim exp\left\{c{S}_A\right\},$$

which, while still extremely large, are vastly more frequent than exact microstate returns. Scrambling times in fast scramblers scale as $t_{\mathrm{scr}}\sim (\beta /2\pi )lnS$ [23,24,25]. Only the latter two regimes carry potential operational relevance (see

Table 1. Representative timescales (schematic). S denotes entropy in units of $k_B$.

Regime	Timescale	Operational?
Global microstate	${\displaystyle T_{\mathrm{rec}}^{\left(\mathrm{global}\right)}\sim exp\left\{\alpha e^{S_{\max }}\right\}}$	No
Coarse-grained A	${\displaystyle T_{\mathrm{rec}}^{\left(A\right)}\left(\epsilon \right)\sim exp\left\{c{S}_A\right\}}$	Typically No
Scrambling	${\displaystyle t_{\mathrm{scr}}\sim (\beta /2\pi )lnS}$	Yes (toy models)

).

Proposition 2 (Coarse-grained/local recurrences). 

Let A be a finite subsystem. Under (A1) and (A2) and standard typicality/ETH hypotheses [16,17,26,27,28,29], for any $\epsilon >0$, there exist arbitrarily large times T such that ${\parallel {\rho }_A\left(T\right)-{\rho }_A\left(0\right)\parallel }_1<\epsilon$. Consequently, macroscopic properties recur much more frequently than exact microstates.

3. Results

Toy Models and Metrics (Summary)

To complement the theoretical results and provide concrete orders of magnitude, two finite-dimensional quantum “toy models” are considered as surrogates for causal patches with a bounded Hilbert space dimension. The goals are as follows: (i) to exhibit global (microstate) almost-recurrences via the fidelity $F\left(t\right)={\left|\langle \psi \left(0\right)|\psi \left(t\right)\rangle \right|}^2$, (ii) to quantify local/coarse-grained returns via the trace distance on a subsystem A, and (iii) to extract a practical scrambling proxy from the entanglement–growth curve, with probability-based visualizations highlighting the contrast between global and coarse-grained returns.

• Models.

(M1) Finite-D spin chain. Transverse-field Ising model with an additional $ZZ$ term to avoid fine-tuned revivals:

$$H=J\sum _{i=1}^N{Z}_i{Z}_{i+1}+h\sum _{i=1}^N{X}_i+g\sum _{i=1}^N{Z}_i,$$

with periodic boundary conditions and parameters $J=1.0,h=1.05,g=0.7$. System sizes $N=10,14$ are reported in the main text (up to $N=16$ in Appendix H).

(M2) Discretized free scalar field (1D lattice). Quadratic Hamiltonian on L sites:

$$H=\frac{1}{2}\sum _{j=1}^L\left(p_j^2+m^2{q}_j^2+\kappa {(q_{j+1}-q_j)}^2\right),$$

with periodic boundary conditions and mode frequencies ${\omega }_k^2=m^2+4\kappa {sin}^2(\pi k/L)$. The parameters used are $m=0.8,\kappa =1.0$, and $L=8,12$.

• Metrics and thresholds.

Global recurrence time $T_{\mathrm{rec}}^{\left(\epsilon \right)}$ is the first passage of the fidelity above a tolerance,

$$T_{\mathrm{rec}}^{\left(\epsilon \right)}=min\{t:F\left(t\right)\ge 1-\epsilon \},\epsilon \in \{{10}^{-2},{10}^{-3}\}.$$

Local/coarse-grained recurrence for a fixed subsystem A (half-chain) is the first passage of the trace distance below a tolerance,

$$T_{\mathrm{rec}}^{\left(A\right)}\left({\epsilon }_A\right)=min\{t:\parallel {\rho }_A\left(t\right)-{\rho }_A\left(0\right){\parallel }_1<{\epsilon }_A\},{\epsilon }_A=0.1.$$

As a scrambling proxy, the time for the von Neumann entanglement entropy $S_{\mathrm{vN}}\left(t\right)$ of A to reach $90\%$ of its long-time plateau is monitored.

• Key observations (from numerical runs).

As shown in Figure 1, the global fidelity $F\left(t\right)$ shows rare full recurrences, while the local proxy signal returns much more frequently. The corresponding timescales are summarized in

Table 2. Summary of recurrence and scrambling times for the toy finite-dimensional model. $T_{\mathrm{rec}}^{\left({10}^{-2}\right)}$: first global return time with fidelity threshold $\epsilon ={10}^{-2}$. $T_{\mathrm{rec}}^{\left({10}^{-3}\right)}$: same with $\epsilon ={10}^{-3}$. $T_{\mathrm{rec}}^{\left(A\right)}({\epsilon }_A=0.1)$: first local return time for the proxy signal with threshold ${\epsilon }_A=0.1$. $t_{\mathrm{scr}}$: scrambling time, defined as the earliest t where the proxy signal reaches $90\%$ of its late-time plateau. Toy model values from 1000 realizations; see Appendix H.

Model	${\mathit{T}}_{\mathbf{rec}}^{\left({10}^{-2}\right)}$	${\mathit{T}}_{\mathbf{rec}}^{\left({10}^{-3}\right)}$	${\mathit{T}}_{\mathbf{rec}}^{\left(\mathit{A}\right)}({\mathit{\epsilon }}_{\mathit{A}}=0.1)$	${\mathit{t}}_{\mathbf{scr}}$
Random Finite-D	45.6	n/a *	12.3	5.7

* n/a: not achieved within the simulation time window ($t_{max}=100$).

. For (M1), the global $T_{\mathrm{rec}}^{\left(\epsilon \right)}$ grows rapidly with N (consistent with exponential-in-$D=2^N$ behavior), while local returns occur much more frequently, and the scrambling proxy scales approximately linearly in N. For (M2), on-site autocorrelations display clear near-revivals due to the finite mode set; global near-recurrence is significantly delayed as L increases, whereas coarse proxies recur on shorter times.

To make this contrast more explicit, Figure 2, Figure 3 and Figure 4 show the cumulative recurrence probability $P\left(T\right)=1-e^{-T/\tau }$ for both global and coarse-grained definitions. At fixed finite D (Figure 2), global probabilities rise only on double-exponential timescales, while coarse-grained returns approach unity quickly. The 3D landscapes (Figure 3 and Figure 4) illustrate how global recurrences become vanishingly rare as D grows, whereas coarse-grained probabilities remain substantial.

• Interpretation.

Although these toy models are far simpler than the cosmological setting, they illustrate the same underlying mechanism: once the number of microstates is finite, recurrences are inevitable on sufficiently long timescales. The probability plots reinforce this message: global recurrences are double-exponentially suppressed in $S_{max}$ and are, therefore, of negligible physical relevance, whereas coarse-grained recurrences occur frequently on exponential timescales in subsystem entropy. This operational distinction is crucial for CCR: it is only the coarse-grained returns that may have physical significance, while global microstate recurrences remain far beyond any observational horizon.

Full details of the toy model construction, parameter choices, and numerical procedures are given in Appendix H.

• Cosmological Consequences

Although strictly conditional, the CCR framework has direct implications for key problems in cosmology. Even though the rigorous recurrence times of a full causal patch are double-exponential and, thus, operationally irrelevant, the conditional structure itself carries genuine cosmological significance:

• Predictivity in eternal inflation. A finite Hilbert space dimension implies a finite outcome space, resolving the measure ambiguities of volume-weighted approaches. Probabilities are then defined over a compact state space, yielding a falsifiable statistical framework.
• Boltzmann-brain constraint. In conventional measures, observers dominated by Boltzmann brains undermine predictivity. In the CCR framework, imposing the single condition ${\Gamma }_{\mathrm{decay}}\gg {\Gamma }_{\mathrm{BB}}$ ensures that ordinary observers prevail, thereby avoiding the paradox.
• Observational falsifiability. If future developments in quantum gravity or cosmology demonstrate unbounded entropy growth in causal patches, the CCR framework would be falsified in its entirety. Conversely, any evidence for metastable vacuum decay would be consistent with the no-BB prescription.
• Conceptual bridge. The CCR program provides a bridge between holographic entropy bounds and macroscopic cosmology. In this sense, “recurrence” is not merely a mathematical curiosity but a structural constraint with direct implications for cosmological predictivity.

In short, the CCR framework translates a mathematical recurrence theorem into a conditional, testable perspective on cosmology: finite information bounds entail finite Hilbert spaces, which, in turn, entail recurrence; the consistency of predictions requires suppressing Boltzmann brains. This endows recurrence with a concrete role in addressing central puzzles of cosmology. While the above discussion applies to a single causal patch, the same logic can be extended to multiverse settings, as discussed in the following subsection.

• Beyond a Single Patch: Multiverse Considerations

Although the preceding discussion focused on a single causal patch, the logic of CCR extends naturally to multiverse scenarios. In eternal inflation, the global spacetime can be viewed as an ensemble of quasi-independent causal patches, each bounded by its own horizon and subject to holographic entropy limits. If each such patch admits a finite Hilbert space, then the CCR theorem applies patch by patch. In this sense, recurrence is not peculiar to “our universe”, but a structural feature of any holographically bounded region.

This suggests a modular picture of the multiverse: a collection of finite-state systems, each evolving unitarily with conditional recurrences. Whether correlations between patches can be defined consistently remains an open problem, but the CCR framework provides a baseline for treating each region as a finite information-bearing system.

4. Discussion

4.1. Measure and Predictivity

• Why ambiguity arises.

In eternally inflating or infinitely extended scenarios, everything that can happen tends to happen infinitely many times. Raw frequencies become $\infty /\infty$, so probabilities are undefined without a regularization/cutoff (a measure). Different cutoffs induce different weights—hence, the measure problem.

• Prescription.

The approach adopted here employs a causal-diamond (local) measure combined with xerographic typicality over ordinary observers, and a single canonical no-Boltzmann-brain constraint (Figure 5), as follows:

$${\Gamma }_{\mathrm{decay}}\gg {\Gamma }_{\mathrm{BB}}\sim H^4{e}^{-S_{\mathrm{BB}}},S_{\mathrm{BB}}\sim {\displaystyle \frac{E_{\mathrm{BB}}}{k_B{T}_{\mathrm{dS}}}},T_{\mathrm{dS}}={\displaystyle \frac{H}{2\pi }}$$

(3)

equivalently ${\tau }_{\mathrm{decay}}\ll {\tau }_{\mathrm{BB}}$. This avoids volume biases and Boltzmann-brain domination in eternal-de Sitter scenarios and restores finite, stable probabilities [15,30].

• Xerographic typicality (definition).

Within a given causal diamond, let ${\mathcal{O}}_{\mathrm{ord}}$ denote the set of ordinary observation instances (excludes BBs). Xerographic typicality assumes that our data $\left(\mathcal{D}\right)$ are sampled uniformly at random from ${\mathcal{O}}_{\mathrm{ord}}$; i.e., the prior weight of an observer is proportional to the count of ordinary observation instances realized in the diamond, with BBs assigned zero weight by the ${\Gamma }_{\mathrm{decay}}\gg {\Gamma }_{\mathrm{BB}}$ constraint.

• Implementation.

In eternal-inflation models, evolve rate equations for vacuum populations $p_i$ with transition rates ${\kappa }_{ij}$ and apply local weighting within a causal diamond. In cyclic settings, count per-cycle events within the diamond under $\epsilon$-coarse-graining. See the worked two-vacuum toy model below.

• Worked example: two-vacuum toy model (no-BB in action)

Consider the following two metastable vacua: 1 (higher H) and 2 (lower H), with transition rates ${\kappa }_{12}$ and ${\kappa }_{21}$ (including decays to terminals as effective leakage). The population fractions obey

$${\displaystyle \frac{d{p}_1}{dt}}=-{\kappa }_{12}{p}_1+{\kappa }_{21}{p}_2,{\displaystyle \frac{d{p}_2}{dt}}=-{\kappa }_{21}{p}_2+{\kappa }_{12}{p}_1,p_1+p_2=1.$$

(4)

Within a causal diamond, weight events by local occurrence rates. Boltzmann brains in vacuum i arise at rate per four-volume ${\Gamma }_{\mathrm{BB},i}\sim H_i^4{e}^{-S_{\mathrm{BB},i}}$. The canonical constraint demands ${\kappa }_{i\to \mathrm{term}}\equiv {\Gamma }_{\mathrm{decay},i}\gg {\Gamma }_{\mathrm{BB},i}$ in each long-lived de Sitter vacuum. For example, if

$$H_1=10H_2,S_{\mathrm{BB},1}={10}^{40},S_{\mathrm{BB},2}={10}^{42},{\kappa }_{12}={10}^{-100}{H}_1,{\kappa }_{21}={10}^{-200}{H}_2,$$

then ${\Gamma }_{\mathrm{BB},i}$ are double-exponentially suppressed, and even minute vacuum-decay rates satisfy ${\Gamma }_{\mathrm{decay},i}\gg {\Gamma }_{\mathrm{BB},i}$. Diamond-weighted observer counts are dominated by ordinary observers prior to decay, avoiding BB dominance while yielding finite, stable probabilities.

• Comparison with alternative measures.
• Pragmatic choice.

We emphasize that the causal-diamond measure combined with the no-BB constraint is a pragmatic choice among several proposed measures, not a derivation from first principles. While it avoids known pathologies (volume-weighting divergences, Boltzmann-brain domination), it remains an assumption whose ultimate justification must await a complete theory of quantum cosmology. We now compare this choice with alternatives.

Several measure proposals have been advanced to address the cosmological measure problem, each with distinct advantages and challenges. Scale-factor time cutoffs [31] regulate the ensemble by fixing a universal time coordinate, but suffer from gauge ambiguities and observer-dependence in spatially infinite scenarios. Light-cone time and fat geodesic prescriptions [32] weight by proper time along worldlines, but introduce ambiguities in defining typical observers and can exhibit instabilities under small perturbations of initial conditions. Causal patch measures [13] (related to but distinct from the causal-diamond prescription adopted here) count events within observer-accessible regions, avoiding some volume-weighting pathologies but leaving open questions about how to define the reference observer ensemble.

The causal-diamond measure employed in this work shares the local-weighting philosophy of causal patch approaches but differs in the following two key respects: (i) it explicitly restricts to a finite diamond bounded by past and future horizons, ensuring a compact integration domain; and (ii) it couples this local weighting with xerographic typicality—the assumption that our observations are drawn uniformly from the set of ordinary observation instances within the diamond, excluding Boltzmann brains via the single canonical constraint ${\Gamma }_{\mathrm{decay}}\gg {\Gamma }_{\mathrm{BB}}$. This combination avoids the “Youngness Paradox” (wherein most observers appear arbitrarily early) and the Boltzmann brain dominance that plague global cutoffs, while remaining falsifiable: evidence of vacuum stability (${\Gamma }_{\mathrm{decay}}\to 0$) would directly contradict the prescription. Known criticisms of local measures—such as sensitivity to the choice of fiducial observer and potential dependence on anthropic selection effects—apply here as well, but are mitigated by the operational focus on ordinary observers (those arising from standard thermalization rather than quantum fluctuations). A full resolution of the measure problem remains open, but the prescription adopted here satisfies the minimal consistency requirement: finite, stable probabilities that remain robust under variations in vacuum decay rates within the range ${\Gamma }_{\mathrm{decay}}\gg {\Gamma }_{\mathrm{BB}}$.

• Necessity and robustness of the no-BB constraint.

The condition ${\Gamma }_{\mathrm{decay}}\gg {\Gamma }_{\mathrm{BB}}$ is sufficient to suppress Boltzmann brain domination, but is it also necessary? In principle, one might imagine alternative mechanisms that render BBs unobservable—for example, selection effects that suppress their measure by other means, or modified entropy bounds that raise $S_{\mathrm{BB}}$ to unattainably large values. However, within the standard framework of metastable de Sitter vacua and semiclassical nucleation, no such mechanism is known. If ${\Gamma }_{\mathrm{decay}}\lesssim {\Gamma }_{\mathrm{BB}}$, the vacuum persists long enough for thermal fluctuations to dominate the observer count, leading directly to the Boltzmann brain paradox. In this sense, the condition is effectively necessary within the scope of known physics, though not a strict logical requirement.

Sensitivity analysis. How robust are the predictions of variations in the threshold? For concreteness, consider a de Sitter vacuum with Hubble parameter $H\sim {10}^{-33}$ eV (close to the observed cosmological constant scale). The Boltzmann brain production rate scales as ${\Gamma }_{\mathrm{BB}}\sim H^{4}exp(-S_{\mathrm{BB}})$, where $S_{\mathrm{BB}}\sim E_{\mathrm{BB}}/\left(k_B{T}_{\mathrm{dS}}\right)$ with $T_{\mathrm{dS}}=H/\left(2\pi k_B\right)\sim {10}^{-30}$ K. For a human-scale brain ($E_{\mathrm{BB}}\sim {10}^{10}$ J $\sim {10}^{29}$ eV), this yields $S_{\mathrm{BB}}\sim {10}^{60}$; hence, ${\Gamma }_{\mathrm{BB}}\sim {10}^{-132}exp(-{10}^{60})\sim {10}^{-{10}^{60}}$ per Hubble volume per Hubble time—utterly negligible. By contrast, plausible vacuum decay rates in string landscape scenarios range from ${\Gamma }_{\mathrm{decay}}\sim {10}^{-600}$ (extremely stable) to ${10}^{-100}$ (marginally metastable). Even the most stable estimates satisfy ${\Gamma }_{\mathrm{decay}}\gg {\Gamma }_{\mathrm{BB}}$ by many orders of magnitude, demonstrating that the constraint is satisfied with an enormous safety margin for any realistic vacuum. Only in the unphysical limit of exact de Sitter stability (${\Gamma }_{\mathrm{decay}}=0$) does the condition fail—a scenario already disfavored by swampland conjectures [9,10]. Thus, the no-BB prescription is both necessary (within known frameworks) and robust (insensitive to order-of-magnitude variations in decay rates).

A schematic comparison of the causal-diamond approach with global volume-weighted measures is shown in Figure 5.

To address the robustness of the measure quantitatively, as requested, we consider the following:

Sensitivity Analysis: Is the condition ${\Gamma }_{\mathrm{decay}}\gg {\Gamma }_{\mathrm{BB}}$ fine-tuned? Consider a typical landscape vacuum with $H\sim {10}^{-33}$ eV. The Boltzmann brain rate is suppressed by the entropic factor ${\Gamma }_{\mathrm{BB}}\sim exp(-{10}^{60})$. In contrast, vacuum decay rates ${\Gamma }_{\mathrm{decay}}$, even for extremely stable vacua, rarely drop below $exp(-{10}^{500})$. This implies a safety margin of hundreds of orders of magnitude in the exponent. The prediction is robust against massive variations in parameters, failing only in the unphysical limit ${\Gamma }_{\mathrm{decay}}\to 0$ (ruled out by Swampland conjectures).

Comparative Summary: Unlike global measures (scale factor) that suffer from gauge ambiguities, or light-cone measures that depend on initial conditions, the causal-diamond measure is local and compact. By coupling it with the no-BB constraint, we resolve the “Youngness Paradox” without determining the global geometry of the multiverse.

4.2. Information/Computation Bounds (CS Viewpoint)

A finite Hilbert space dimension also places limits on the information-theoretic and computational capacity of a causal patch. This provides a complementary “computer science” viewpoint on the consequences of bounded D.

• Bit budget.

The maximum reliably distinguishable information inside a causal patch is set by the Hilbert space dimension,

$${log}_{2}D\lesssim {\displaystyle \frac{S_{max}}{ln2}}\le {\displaystyle \frac{\Lambda }{4{\mathcal{l}}_P^{2}ln2}}\mathrm{bits},$$

where $S_{max}$ is the maximal entropy and $\Lambda$ is the de Sitter horizon area. This inequality expresses the intuitive statement that a causal patch behaves as a finite-capacity information channel.

• Operation rate.

Given mean energy E, the Margolus–Levitin bound and Landauer’s principle imply that logical operations within the patch are constrained by

$$N_{\mathrm{ops}}\lesssim {\displaystyle \frac{2E}{\pi \hslash }},E_{\mathrm{erase}}\ge k_{B}Tln2.$$

Together with collapse thresholds, these results define a feasible computation envelope per patch, consistent with Lloyd’s “ultimate physical limits to computation” [33,34].

• Implication.

These bounds highlight that the CCR framework is not only about abstract recurrence times: finite Hilbert space dimension directly limits both the storage capacity and the computational rate of any causal patch. In this sense, recurrence, predictivity, and information processing are facets of the same underlying finiteness constraint.

4.3. Indirect Observational Probes

• Compact spatial topology: Cosmic Microwave Background (CMB) “matched circles” and low-ℓ anomalies; current Planck searches find no matched circles above angular radii of $\gtrsim {10}^{°}$ and set strong lower bounds on the size of a fundamental domain [35,36].
• Bubble collisions (eternal inflation): disk-like imprints in CMB temperature/polarization; non-detection constrains the landscape of eternal-inflation scenarios consistent with CCR.
• Cyclic/CCC-like memory: specific non-Gaussian residuals; localized hot/cold spots. Existing Planck analyses report no significant evidence for the concentric-ring patterns proposed in CCC and place upper bounds on such features [37,38]. Prospects improve with CMB-S4 [39].
• Dark-energy metastability: either small deviations $w\left(z\right)\ne -1$ or lower bounds on ${\Gamma }_{\mathrm{decay}}$ consistent with the no-BB prescription, providing a key falsifiability condition.
• Primordial GWs: spectra incompatible with simple slow-roll but compatible with cyclic/pre-inflationary scenarios, potentially hinting at finite-dimensional pre-inflationary dynamics.

• Final remark.

No direct observational constraints currently exist on the Hilbert space dimension D or on $S_{max}$ of a causal patch. Existing limits are only indirect—for example, from the absence of non-trivial cosmic topology in the CMB, or from lower bounds on vacuum decay rates. These motivate the conditional framing adopted here, but do not yet provide a direct empirical handle on finiteness.

Future probes of the fundamental information content of spacetime could nonetheless offer indirect tests: for instance, precise CMB measurements might reveal signs of a finite mode spectrum, or holographic approaches might connect D to observable entropy production in reheating. The framework is, therefore, empirically vulnerable: any observational evidence of unbounded entropy or state-space growth would falsify CCR outright.

• Falsifiability and direct tests.

Direct observational tests of the finite Hilbert space dimension D are not currently possible with existing or near-future technology. However, the CCR framework is falsifiable in principle: any evidence of unbounded entropy growth within a causal patch would falsify the core assumption (A1).

Concrete falsification scenarios include the following:

• Detection of ever-growing horizon entropy in accelerating expansion (requiring sustained entropy production beyond $S_{max}$).
• Observation of cosmological dynamics inconsistent with any finite state budget (e.g., unbounded particle production rates).
• Demonstration from completed quantum gravity that causal patches admit $D=\infty$.

Current tests remain necessarily indirect (CMB topology, vacuum decay bounds), but the framework’s conditional structure ensures clear empirical stakes.

4.4. Rigor Addendum: Theorem-Level vs. Assumptions

The CCR framework combines mathematically rigorous ingredients with physics-level assumptions. It is essential to separate the two, as follows:

• Theorem-level results (rigorous).

  (i)

  For finite-dimensional $\mathcal{H}$ and unitary evolution, quantum almost-periodicity and recurrences follow directly (Bocchieri–Loinger).

  (ii)

  Coarse-grained/local recurrences follow from almost-periodicity together with continuity of the partial trace in trace norm.

  (iii)

  Canonical typicality results (e.g., Goldstein–Lebowitz–Tumulka–Zanghì; Popescu–Short–Winter) establish the operational relevance of coarse-grained returns.

• Physics-level assumptions.
  • The finiteness of the operational state budget (A1) is not a theorem but an assumption, motivated by Bekenstein–Hawking/Gibbons–Hawking entropy bounds and supported in Loop Quantum Gravity (LQG) and holographic contexts.
  • The microcanonical counting with a gravitational energy cap developed herein provides a physically controlled argument for free/weakly coupled fields.
  • ETH/ergodicity statements are used in their standard “genericity” sense.

• Summary.

If (A1) fails, the CCR framework is falsified in its entirety. If (A1) holds, then the CCR theorem follows rigorously. This clear conditional structure ensures that the distinction between mathematics and physics is explicit.

4.5. Limitations and Open Issues

4.5.1. Breakdown Scenarios

The Conditional Recurrence Theorem relies critically on assumption (A1), namely that the Hilbert space dimension $dim{\mathcal{H}}_{\mathrm{patch}}$ of a causal patch is finite. If (A1) is violated, CCR’s finite-D premise fails. Below, we summarize the main scenarios where the assumption breaks down.

• Infinite volume limits.

In continuum QFT on a spatial region of volume V, the number of modes below a UV cutoff $\Lambda$ scales as

$$M(\Lambda )\sim {\displaystyle \frac{V}{{\left(2\pi \right)}^3}}{\displaystyle \frac{4\pi }{3}}{\Lambda }^3.$$

For non-compact or infinite-volume backgrounds ($V\to \infty$), one obtains $M(\Lambda )\to \infty$ even at finite $\Lambda$. Hence, the Fock space dimension is unbounded. Without a gravitational entropy cap, there is no finite operational state budget and almost-periodicity theorems for finite D do not apply.

• Unbounded entropy growth.

In open FRW or certain eternal-inflation scenarios, coarse-grained entropy can increase without bound (e.g., ever-growing horizons or sustained particle production). If

$$S_{\mathrm{req}}\left(t\right)\to \infty \mathrm{as}t\to \infty ,$$

then the effective dimension $D_{\mathrm{eff}}\left(t\right)\sim exp\{S_{\mathrm{req}}\left(t\right)/k_B\}$ diverges with time, so there is no time-independent finite-D Hilbert space on which to invoke CCR.

• Continuous spectra/non-compact holography.

In AdS/CFT with non-compact spatial slices (or in flat-space holography), the dual theory typically has a continuous spectrum in the relevant sector. Continuous spectral measures preclude the kind of quasi-periodic phase realignment guaranteed on a compact torus of phases. Thus, unitarity alone does not ensure recurrence.

• Non-unitary evolution.

Objective-collapse or non-Hamiltonian cosmological models break unitarity. Since the CCR theorem relies on unitary evolution, any such modification of quantum dynamics invalidates recurrence statements derived from it.

• Consequence. In all the above scenarios, the key CCR hypothesis (finite operational D) fails. Global quantum recurrences are not mathematically guaranteed; at best, one may observe model-dependent quasi-recurrences or ergodic features, but no universal bound on return times.

4.5.2. Caveats and Consequences

Even if (A1) fails, while (A2–A4) still hold, one recovers the conventional picture of unbounded state space, in which the notion of recurrence becomes physically irrelevant for observers. Specific caveats include the following:

• Unitarity fails at cosmological scales (as in certain objective-collapse or non-Hamiltonian models),
• The causal patch framework may not be the correct coarse-graining of cosmology,
• Non-unitary models: Objective-collapse dynamics would invalidate recurrence,
• Sector locking/non-ergodicity: Horizons and conserved charges may confine dynamics; recurrence then holds only within sectors,
• Measure ambiguity: Different measures yield different predictions; the prescription in Section 4.1 satisfies basic sanity checks but remains an assumption. The no-BB constraint is stated once, canonically, in Section 4.1,
• Timescales: Global recurrences occur on timescales at least double-exponential in $S_{\max }$ (Remark 2); only coarse-grained returns may have operational relevance,
• Falsifiability: The finite-dimensional Hilbert space conjecture (A1) could be falsified if any observation demonstrates unbounded entropy growth within a causal patch.

4.5.3. Compatibility with Holography and Swampland

Despite these limitations, CCR is compatible with several frameworks, as follows:

• de Sitter holography: Entropic bounds in FRW/de Sitter are compelling but not rigorously derived from a completed quantum gravity,
• Compatibility: CCR is consistent with Poincaré recurrence in finite systems and with holographic formulations; in de Sitter cosmology, it aligns with proposals of finite $dim\mathcal{H}\sim e^{S_{\mathrm{dS}}}$ [7,8] while differing from eternal-inflation pictures with infinite state spaces,
• Swampland constraints: If fully stable de Sitter vacua are absent or highly constrained [9,10], a non-negligible decay rate is expected, which supports the no-BB requirement and enhances predictivity.

• Closing remark.

CCR is not a universal law but a conditional framework. Its validity hinges on the finiteness of Hilbert space per causal patch; its falsification would follow from any observational or theoretical evidence for unbounded entropy growth, while its consistency would strengthen the case for holographic entropy bounds in cosmology.

4.6. Connections to Broader Frameworks

• String theory and holography.

In AdS/CFT, a CFT on a compact spatial manifold has a discrete spectrum and finite thermal entropy for finite energy, so (coarse-grained) recurrences follow on exponentially long timescales. Black-hole entropy scales with area, matching the holographic spirit of assumption (A1).

While a complete de Sitter holography is not yet established, proposals such as dS/CFT and static-patch finiteness [40,41,42,43,44] motivate treating $S_{\max }\sim A/4{\mathcal{l}}_P^2$ as an operational entropy cap. Under such an assumption, the CCR statement acts as a direct, causal-patch analogue, as follows:

$$\mathrm{finite}\mathrm{operational}D+\mathrm{unitarity}\Rightarrow \left(\mathrm{coarse}\text{-}\mathrm{grained}\right)\mathrm{quantum}\mathrm{recurrences}.$$

This operational perspective is distinct from a UV-complete duality but resonates with Swampland considerations that disfavor fully stable de Sitter vacua. In this sense, CCR can be viewed as a low-energy, information-theoretic corollary of holographic entropy bounds, independent of whether a complete dS holography is ultimately realized.

5. Conclusions

The Conditional Cosmological Recurrence (CCR) framework developed here establishes a clear micro-to-macro bridge from holographic entropy bounds to finite Hilbert space dimensionality, and, hence, to quantum recurrence. Its structure is strictly conditional: if causal patches admit a finite operational Hilbert space dimension, recurrence follows rigorously from unitarity; conversely, if $D=\infty$, the framework is falsified in its entirety.

This conditional framing allows CCR to serve as a conceptual bridge between microscopic entropy bounds and macroscopic cosmology. Although global recurrences occur on timescales that are double-exponential and operationally irrelevant, the framework yields the following falsifiable implications: predictivity in eternal inflation, suppression of Boltzmann brains, and observational vulnerability to evidence of unbounded entropy growth.

Looking ahead, several avenues appear promising, as follows: (i) refinement of toy models to probe coarse-grained recurrence scaling, (ii) exploration of connections with holographic dualities and swampland constraints, and (iii) indirect observational probes such as CMB topology tests [43], primordial gravitational waves, or vacuum decay signatures. Together, these directions may clarify whether CCR captures a structural feature of quantum cosmology or must ultimately be replaced by a different organizing principle.