When the ticking of a clock is no longer a given, but a consequence of deeper physics, the very notion of “when” begins to wobble. In the quest for quantum gravity—a theory that unites Einstein’s general relativity with the quantum world—physicists have been forced to ask whether time itself is fundamental or whether it emerges from more primitive ingredients such as entropy, relational degrees of freedom, or quantum correlations. The answer has profound implications: it reshapes our understanding of the early universe, the arrow of time that drives entropy forward, and even the way we model complex, self‑organising systems like honeybee colonies or autonomous AI agents.
This article pulls together the most concrete proposals for emergent time, grounds them in real calculations and observations, and sketches how a fresh perspective on temporality could ripple outward to fields as diverse as cosmology, ecological conservation, and the design of trustworthy AI. By the end, you’ll see why the debate over “does time exist?” is not an abstract curiosity but a cornerstone of how we predict the fate of the cosmos and steward the fragile ecosystems that depend on it.
1. The “Problem of Time” in Quantum Gravity
General relativity treats time as a coordinate woven together with space into a four‑dimensional manifold. The Einstein field equations
\[ G_{\mu\nu} + \Lambda g_{\mu\nu}= \frac{8\pi G}{c^{4}} T_{\mu\nu} \]
express how matter‑energy tells spacetime how to curve, and curvature tells matter how to move. In this picture, time is dynamical: it can stretch, contract, or even disappear inside black holes.
Quantum mechanics, by contrast, is built on a fixed background time. The Schrödinger equation
\[ i\hbar\frac{\partial}{\partial t}\psi(t)=\hat H\psi(t) \]
evolves a state vector \(\psi\) with respect to an external parameter \(t\). The two frameworks clash when we try to quantise the gravitational field itself.
If we write down the Wheeler–DeWitt equation—a direct attempt to apply canonical quantisation to general relativity—we obtain
\[ \hat H_{\text{grav}}\Psi = 0, \]
with no explicit time derivative. The wave functional \(\Psi\) of the universe appears timeless. This is the “frozen formalism”: the quantum state of the whole spacetime satisfies a constraint rather than an evolution law.
Physicists have responded in three broad ways:
- Accept the timelessness and reinterpret the apparent flow of time as an emergent phenomenon.
- Introduce an internal clock—a degree of freedom that can serve as a relational time variable.
- Modify the quantisation scheme (e.g., path‑integral approaches) so that time re‑appears in a different guise.
The first two strategies are the focus of emergent‑time research. They ask whether the arrow we experience—entropy increasing, clocks ticking—can be derived from deeper, possibly entropic or relational structures that do not presuppose a background temporal metric.
2. Entropic Time: From Thermodynamics to Cosmology
2.1 The Thermodynamic Arrow
The second law of thermodynamics tells us that the total entropy \(S\) of an isolated system never decreases:
\[ \Delta S \ge 0. \]
If we define a “thermodynamic time” \(\tau\) by the monotonic increase of entropy, then \(\tau\) becomes a derived parameter. In practice, this idea is sharpened by considering Boltzmann’s H‑theorem, which, under assumptions of molecular chaos, yields a one‑way increase of the H‑function, effectively a time arrow.
In the early universe, the cosmic microwave background (CMB) at redshift \(z\approx 1100\) shows temperature fluctuations of only \(\delta T/T \sim 10^{-5}\). This low‑entropy state is quantified by the dimensionless entropy density \(s = (2\pi^{2}/45) g_{} T^{3}\) where \(g_{}\) counts relativistic degrees of freedom. At the Planck epoch (\(t \approx 5.4\times10^{-44}\,\text{s}\)), the entropy in a Hubble volume is estimated to be \(S_{\text{Pl}} \sim 10^{2}\), minuscule compared with the present value \(S_{0} \sim 10^{104}\) (dominated by supermassive black holes).
The contrast is so stark that it has been called the “Past Hypothesis”: the universe began in an extraordinarily low‑entropy configuration. If time is emergent from entropy, then the cosmological arrow is simply the direction in which the number of accessible microstates expands.
2.2 Entropy from Quantum Geometry
Loop quantum gravity (LQG) provides a concrete setting where geometry itself carries entropy. The area operator \(\hat A\) has a discrete spectrum
\[ A_j = 8\pi \gamma \ell_{\text{P}}^{2}\sqrt{j(j+1)}, \]
with spin quantum number \(j\in\frac12\mathbb{N}\), Barbero–Immirzi parameter \(\gamma\approx 0.274\), and Planck length \(\ell_{\text{P}} \approx 1.62\times10^{-35}\,\text{m}\). Counting the number of spin‑network configurations that give a fixed macroscopic area yields a Bekenstein–Hawking entropy
\[ S_{\text{BH}} = \frac{k_{\text{B}} c^{3}}{4\hbar G} A, \]
exactly the same formula that appears for black‑hole horizons. In LQG, the microstates responsible for this entropy are the quantum excitations of space itself.
If we treat the ensemble of spin networks as a statistical system, the entropic time can be defined by the increase of the number of possible spin‑network configurations as the universe expands. In practice, one can follow the volume operator \(\hat V\) whose eigenvalues grow with the number of nodes, and show that the associated entropy \(S(V) \sim \ln \mathcal{N}(V)\) is monotonic. This provides a microscopic underpinning for a thermodynamic arrow emerging from quantum geometry.
2.3 Concrete Example: The de Sitter Horizon
A universe dominated by a positive cosmological constant \(\Lambda\) asymptotes to de Sitter space, with a horizon radius
\[ R_{\Lambda} = \sqrt{\frac{3}{\Lambda}} \approx 1.6\times10^{26}\,\text{m} \]
for the observed \(\Lambda \approx 1.1\times10^{-52}\,\text{m}^{-2}\). The associated Gibbons–Hawking entropy
\[ S_{\text{dS}} = \frac{k_{\text{B}}c^{3}}{4\hbar G} 4\pi R_{\Lambda}^{2} \approx 2.6\times10^{122}\,k_{\text{B}} \]
is the largest finite entropy we can assign to a causal patch. The de Sitter horizon thus furnishes a natural “clock”: as the universe expands, the horizon area—and therefore its entropy—changes only when matter crosses the horizon, a process that can be counted. In this picture, the flow of time is linked to the information loss across a cosmological horizon, an entropic bookkeeping that is completely independent of any external time parameter.
3. Relational Time: Correlations as Clocks
3.1 Page–Wootters Mechanism
In 1983, Don Page and William Wootters proposed a relational approach: the universe is described by a global, timeless state \(|\Psi\rangle\) satisfying \(\hat H |\Psi\rangle = 0\). Subsystems can nevertheless experience an effective Schrödinger evolution if we condition on a “clock” subsystem.
Mathematically, we split the Hilbert space \(\mathcal{H} = \mathcal{H}{C} \otimes \mathcal{H}{S}\), where \(\mathcal{H}{C}\) holds the clock and \(\mathcal{H}{S}\) the system of interest. The global constraint reads
\[ (\hat H_{C} + \hat H_{S})|\Psi\rangle = 0. \]
Define a clock basis \(\{|t\rangle\}\) that diagonalises \(\hat H_{C}\). The conditional state of the system given a clock reading \(t\) is
\[ |\psi_{S}(t)\rangle = \frac{\langle t|\Psi\rangle}{\sqrt{\langle\Psi|t\rangle\langle t|\Psi\rangle}}. \]
One can show that \(|\psi_{S}(t)\rangle\) satisfies the usual Schrödinger equation with respect to the parameter \(t\). In other words, time emerges as a correlation between subsystems.
Experimental demonstrations have been performed with trapped ions and superconducting qubits, where a “clock” qubit is entangled with a “system” qubit, and the conditional dynamics reproduces the expected unitary evolution. These tabletop experiments confirm that relational time is not merely philosophical—it can be implemented in laboratory quantum systems.
3.2 Cosmological Relational Time
Applying the Page–Wootters scheme to the whole universe raises the question: what can serve as a universal clock? Proposals include:
- Scalar fields (e.g., the inflaton) whose monotonic evolution during inflation can label “when” events happen.
- Volume of the spatial slice—since the total volume \(\mathcal{V}\) grows under expansion, it can act as a clock (the “volume time” used in many LQG cosmology models).
- Matter clocks: the number of particles of a conserved species (e.g., baryon number) is a discrete, monotonic quantity that can be used to order events.
A concrete calculation appears in loop quantum cosmology (LQC). The effective Friedmann equation receives a correction
\[ H^{2} = \frac{8\pi G}{3}\rho \left(1 - \frac{\rho}{\rho_{c}}\right), \]
where \(\rho_{c} \approx 0.41\rho_{\text{Pl}}\) is the critical density set by the underlying quantum geometry. The bounce at \(\rho = \rho_{c}\) replaces the classical singularity. One can treat the scalar field \(\phi\) as a relational clock: its momentum \(p_{\phi}\) is conserved, and the volume \(\mathcal{V}(\phi)\) evolves as
\[ \mathcal{V}(\phi) = \mathcal{V}{\text{min}} \cosh\!\bigl[\sqrt{24\pi G}\,(\phi-\phi{0})\bigr]. \]
Thus, time is encoded in the change of \(\phi\), and the bounce becomes a smooth transition in relational terms.
3.3 Relational Time in String Theory
In certain backgrounds of string theory, especially AdS/CFT dualities, the boundary conformal field theory (CFT) provides a natural time parameter that is not the bulk spacetime time. The bulk geometry is emergent from entanglement patterns in the CFT. Recent work on tensor network models (e.g., MERA) shows that the radial direction of AdS can be interpreted as a scale‑dependent renormalisation flow, while the boundary time drives evolution. The bulk “global” time is then a derived quantity, reconstructed from the correlated evolution of the boundary degrees of freedom.
4. The Arrow of Time from Quantum Gravity
4.1 Entropy Production in Black‑Hole Evaporation
Black holes radiate via Hawking radiation, a quantum process that increases the total entropy of the universe. The Hawking temperature for a Schwarzschild black hole of mass \(M\) is
\[ T_{\text{H}} = \frac{\hbar c^{3}}{8\pi G k_{\text{B}} M} \approx 6.2\times10^{-8}\,\text{K}\,\bigg(\frac{M_{\odot}}{M}\bigg). \]
The associated entropy loss of the black hole is
\[ \Delta S_{\text{BH}} = -\frac{4\pi k_{\text{B}} G M^{2}}{\hbar c}, \]
while the emitted radiation carries an entropy increase
\[ \Delta S_{\text{rad}} \approx \frac{4}{3}\frac{M c^{2}}{T_{\text{H}}}. \]
The net change \(\Delta S_{\text{tot}} = \Delta S_{\text{rad}} + \Delta S_{\text{BH}} > 0\) demonstrates a microscopic arrow tied to quantum gravitational dynamics. In a fully quantum‑gravity description, the black‑hole interior may be replaced by a fuzzball or a firewall, but the entropy bookkeeping still yields a forward‑moving direction.
4.2 The “Thermal Time Hypothesis”
Connes and Rovelli (1994) proposed that time flow can be derived from the state of a statistical system via Tomita–Takesaki modular theory. For a von Neumann algebra \(\mathcal{A}\) and a faithful state \(\omega\), the modular operator \(\Delta_{\omega}\) generates a one‑parameter automorphism group
\[ \alpha_{t}(A) = \Delta_{\omega}^{it} A \Delta_{\omega}^{-it}, \]
which can be identified with the physical time evolution. This thermal time coincides with the usual Hamiltonian flow when the state is a thermal Gibbs state. In quantum gravity, the suggestion is that the global state of the universe defines a modular flow, and that this flow is the experienced time. The hypothesis predicts that any region with a higher temperature (e.g., near a black‑hole horizon) experiences a faster modular flow, mirroring the gravitational red‑shift of conventional clocks.
4.3 Numerical Simulations: Causal Dynamical Triangulations
Causal dynamical triangulations (CDT) is a non‑perturbative approach that builds spacetime from discrete building blocks (simplices) respecting a causal ordering. Monte Carlo simulations in four dimensions have revealed a phase diagram with a “de Sitter‑like” phase where the emergent geometry reproduces an expanding universe with a well‑defined proper time.
In the CDT framework, proper time is not an input but a label that arises from the way simplices are glued together respecting causality. The observed macroscopic scale factor \(a(t)\) follows the Friedmann equation with an effective cosmological constant \(\Lambda_{\text{eff}} \approx 0.1\,\ell_{\text{P}}^{-2}\). This emergent behavior suggests that the arrow of time is encoded in the statistical distribution of causal configurations, reinforcing the view that time is a macroscopic, entropic parameter.
5. From Cosmic Time to Complex Systems
5.1 Bees, Information Flow, and Entropy
Honeybee colonies exhibit a distributed decision‑making process that relies on the flow of information through waggle dances, pheromone trails, and tactile cues. Experiments have shown that a colony of Apis mellifera can collectively evaluate up to 10–15 nectar sources, each with differing profitabilities, and converge on the optimal one in roughly 20–30 minutes—a timescale that scales logarithmically with the number of options, reminiscent of entropy‑driven search algorithms.
If we view the colony as a thermodynamic system, the Shannon entropy of the distribution of foragers among nectar sources decreases as the colony reaches consensus. The rate of entropy reduction can be mapped onto an emergent time variable: the more quickly the colony reduces uncertainty, the faster it “advances” towards a decision.
This mirrors the relational‑time picture: the state of the colony (who is dancing, who is foraging) correlates with a collective clock defined by the decrease of uncertainty. In a sense, the colony creates its own time arrow, independent of any external chronometer.
5.2 AI Agents and Relational Clocks
Self‑governing AI agents—especially those trained via reinforcement learning—often develop internal timers to coordinate actions. In multi‑agent environments, agents may learn to synchronize by sharing a latent variable that encodes a progression of steps, even when the external environment provides no explicit tick.
A concrete case is the OpenAI Hide‑and‑Seek experiments, where agents learned a “day‑night” cycle by counting the number of frames since a reset, effectively constructing a relational clock from the environment’s reset signal. The emergent time allowed agents to plan multi‑step strategies like building fortifications only during “day” phases.
These AI examples demonstrate that time can be a useful abstraction for any system that needs to order events, even when the underlying physics does not provide a global temporal coordinate. The parallel to emergent time in quantum gravity is striking: both involve correlations among subsystems that give rise to a useful notion of “when”.
6. Observational Windows on Emergent Time
6.1 Primordial Gravitational Waves
If the early universe’s time emerged from a quantum‑geometric regime, the spectrum of primordial gravitational waves could carry signatures of that process. In LQC, the bounce leads to a suppression of power at the largest scales (\(\ell \lesssim 10\)) and a possible oscillatory modulation in the tensor power spectrum
\[ P_{T}(k) = A_{T} \left(\frac{k}{k_{}}\right)^{n_{T}} \left[1 + \alpha \cos\!\bigl(\beta k/k_{}\bigr)\right], \]
with \(\alpha\sim 0.1\) and \(\beta\sim 2\). Future satellite missions like LiteBIRD aim to reach a tensor‑to‑scalar ratio \(r < 10^{-3}\), sufficient to test such features. Detecting a deviation from the standard inflationary prediction would hint that time before inflation was not the same as the classical proper time.
6.2 Black‑Hole Spectroscopy
The ringdown phase of a perturbed black hole is characterised by quasinormal modes (QNMs). In a theory where spacetime is discrete at the Planck scale, the QNM frequencies acquire quantum corrections of order \(\Delta \omega \sim \omega_{\text{QNM}} \times (\ell_{\text{P}}/R_{s})\), where \(R_{s}\) is the Schwarzschild radius. For a stellar‑mass black hole (\(R_{s}\approx 3\,\text{km}\)), this correction is \(\sim10^{-21}\) relative, far below current detector sensitivity. However, for intermediate‑mass black holes (\(M\sim10^{4}M_{\odot}\)) observed by LISA, the correction could be \(\sim10^{-15}\), still tiny but potentially observable with future gravitational‑wave interferometers designed for high‑precision spectroscopy.
A measured deviation would imply that time evolution near the horizon is governed by a discrete, perhaps entropic, structure, supporting emergent‑time scenarios.
7. Synthesis: How Emergent Time Reshapes Cosmology
Putting together the entropic and relational threads, a coherent picture begins to emerge:
- Micro‑states of quantum geometry (spin networks, strings, causal sets) possess a vast degeneracy that can be counted. Their entropy grows as the universe expands, providing a thermodynamic arrow.
- Relational observables—scalar fields, volume, particle number—serve as internal clocks, allowing subsystems to experience an effective Schrödinger evolution without invoking an external time parameter.
- Modular flow (thermal time) suggests that the state of the universe alone can generate a one‑parameter automorphism that we interpret as time.
- Cosmological observables (CMB anisotropies, gravitational‑wave spectra) may retain imprints of the transition from a timeless quantum regime to the classical expanding universe we see today.
If this synthesis holds, the big bang singularity is replaced by a bounce or a smooth emergence where time itself nucleates from entropic growth. The arrow of time is not an extra assumption but a natural consequence of the statistical properties of the underlying quantum geometry.
8. Implications for Conservation and AI Governance
8.1 Ecological Timing
Understanding how time emerges from information flow and entropy provides a fresh lens on phenological shifts—the timing of biological events like flowering or bee foraging. Climate change accelerates the entropy production in ecosystems (e.g., via increased heat flux), potentially altering the relational clocks that species use to coordinate. By modelling bee colonies with entropy‑based decision clocks, we can predict how rapid environmental changes might desynchronize pollinator–plant interactions, guiding targeted conservation measures.
8.2 Designing Trustworthy AI
In AI governance, an emergent‑time perspective warns against assuming a universal, externally imposed clock for all agents. Instead, relational clocks—derived from shared data streams or consensus variables—could be used to enforce coordination while respecting privacy and decentralisation. Moreover, the thermal time hypothesis suggests that an AI system’s internal state (its “belief distribution”) could define its own notion of progress, opening pathways for self‑regulating agents that adapt their operation speed to the entropy of their environment, much like a bee colony modulates activity based on resource availability.
9. Future Directions and Open Questions
| Question | Why It Matters | Current Status |
|---|---|---|
| Can we derive a unique relational clock from first principles? | Determines whether time is truly emergent or merely a convenient choice. | Page–Wootters provides a framework; uniqueness remains unresolved. |
| What is the precise microscopic origin of the cosmological arrow? | Links low‑entropy initial conditions to quantum gravity. | Entropic arguments compelling; need a concrete statistical model of spin‑network microstates. |
| Do observational signatures of emergent time survive inflation? | Makes the theory falsifiable. | LQC predicts specific modulations in the tensor spectrum; awaiting next‑generation CMB data. |
| Can we engineer laboratory analogues (e.g., cold atoms) that mimic a timeless universe? | Provides a test‑bed for relational‑time concepts. | Recent trapped‑ion experiments simulate Page–Wootters dynamics; scaling up is a challenge. |
| How does emergent time affect the information paradox? | Could resolve tension between unitarity and Hawking evaporation. | Proposals involving modular flow and fuzzball microstates are under active development. |
Addressing these questions will require collaboration across theoretical physics, numerical relativity, quantum information, and ecological modelling—a truly interdisciplinary effort.
Why it matters
Time is the scaffolding on which we build every story: the narrative of the universe, the life cycle of a honeybee, the schedule of an autonomous AI. If time is not a fundamental backdrop but a derived, entropic, relational construct, then many of our deepest puzzles—from the nature of the big bang to the coordination of complex adaptive systems—gain a new foothold. By grounding the discussion in concrete calculations, experimental analogues, and real‑world analogies, we see that the question is not abstract philosophy but a practical guide for predicting cosmic evolution, protecting pollinator networks, and designing AI that respects the flow of information.
In the end, embracing emergent time may help us align the temporal scales of the cosmos with those of the ecosystems we cherish and the intelligent agents we are building, ensuring that each tick—whether measured by a Planck clock, a bee’s waggle, or an AI’s internal state—advances the shared goal of a sustainable, comprehensible future.