The Black Hole Convergence

The natural way to test a premise about the foundations of physics is not to argue for it philosophically and then wait. It is to follow the most developed traditions in the field to wherever they lead and ask whether they arrive at the same place. If entropy is the more fundamental quantity — if captures something that has always obscured — then the premise is either structural or it is not. A structural premise holds everywhere the relevant frameworks overlap, without exception and without residue. The test is to find that overlap and look at what it says.

The two traditions that bear most directly on this are mechanics and thermodynamics.

Mechanics begins with Newton: force, mass, acceleration, and time as the featureless river flowing uniformly beneath everything. The equations are clean and reversible, and is the axis on which all of it is plotted. Maxwell extends the framework into the electromagnetic field — waves propagating at a fixed velocity, whose constancy under all reference frames creates an irreconcilable tension with Newton's absolute time. Einstein resolves the tension by absorbing it: special relativity collapses the separation between space and time into a single four-dimensional manifold, and general relativity makes that manifold itself the dynamical object, curved by the presence of mass and energy, with gravity reconceived not as a force acting across distance but as the curvature that free bodies follow. At the extreme end of this development — where curvature is strong enough that no trajectory within the light cone leads away from the interior — are black holes. They are not postulated. They are solutions: exact solutions of the Einstein field equations, first derived by Schwarzschild in 1916, weeks after Einstein published the theory itself.

Thermodynamics arrives at the same object from a direction that shares no ancestry with mechanics. The problem of black body radiation — the spectrum of light emitted by a heated cavity, which classical physics predicted should diverge at high frequencies and does not — forced a rupture in the late nineteenth century that mechanics had no tools to repair. Planck resolved it by quantisation: energy emitted in discrete packets whose size depends on frequency. From that crack in the classical framework, quantum theory emerged. Alongside it, statistical mechanics — developed principally by Boltzmann — gave entropy a microscopic foundation. Entropy is a count: the logarithm of the number of microstates consistent with a macroscopic description. The Second Law is not imposed on the world from outside. It is the statistical consequence of disordered arrangements being overwhelmingly more numerous than ordered ones, so that any system evolving through accessible states will, with near certainty, move toward higher entropy. This tradition, tracing its lineage through Carnot, Clausius, Boltzmann, and Gibbs, eventually arrives at black holes — but from an entirely different direction. Bekenstein argued in 1972 that a black hole must carry entropy proportional to the area of its event horizon. The reasoning was thermodynamic: a black hole that absorbed matter would otherwise permit entropy to vanish behind the horizon, violating the Second Law. The entropy of the infalling matter must persist somewhere accessible to the thermodynamic accounting. It persists in the geometry. Hawking confirmed this in 1974 through a different route entirely: quantum field theory in the curved spacetime background near an event horizon requires thermal radiation, with a temperature proportional to the horizon's surface gravity. Black holes, the most extreme objects the mechanical tradition produces, are also thermodynamic objects — and they were found to be so by two independent approaches that share neither their mathematical vocabulary nor their founding assumptions.

Two traditions, two complete sets of tools, two lineages of reasoning developed largely in isolation from each other. Both arrive at the same object. The question is whether they describe it in compatible terms, or whether the overlap is superficial — a shared name for what are, beneath the language, two incompatible accounts.

The method I find most honest for testing this is the one that refuses to stop at a promising case. Take every distinct state a black hole can occupy — defined by every physical quantity a black hole can possess — and ask whether each of those quantities appears as a thermodynamic variable. If the premise is structural, the answer should be the same in every case. No exception is allowed. An exception would mean the correspondence is partial, and a partial correspondence is not a structural one.

The Schwarzschild black hole carries mass alone — no spin, no charge. The thermodynamic first law for this state is : a change in mass equals temperature times a change in entropy. Mass, the central mechanical quantity, is a thermodynamic variable. That is the Bekenstein-Hawking result.

Add spin. The Kerr black hole carries angular momentum in addition to mass. The first law becomes , where is the angular velocity of the event horizon. Angular momentum — a quantity whose conservation is a straightforward consequence of Newton's laws in the mechanical description — enters the thermodynamic first law as a thermodynamic variable, with the horizon's angular velocity as its conjugate potential. The mechanical and thermodynamic descriptions are still in contact.

Add electric charge. The Reissner-Nordström black hole carries charge . The first law becomes , where is the electrostatic potential at the horizon. Electromagnetism is now inside the thermodynamic description. Not appended to it. Not as an analogy drawn from the mechanical side. As a thermodynamic variable, on exactly the same footing as mass, with its own conjugate potential.

Combine all three. The Kerr-Newman black hole carries mass, spin, and electric charge — it is the most general solution of the coupled Einstein-Maxwell equations that can be characterised by a finite set of parameters. The first law is:

Every quantity the mechanical tradition uses to characterise the object appears as a thermodynamic variable. No mechanical degree of freedom is left unaccounted for in the thermodynamic description.

The test should not stop there. Extend the solution to include a cosmological constant — the vacuum energy density, the term Einstein introduced in 1917 and later called his greatest blunder, which has since returned in the context of dark energy and the observed accelerating expansion of the universe. The extended first law adds a pressure-volume term , where plays the role of thermodynamic pressure. Vacuum energy, the background energy density of empty space, enters as a thermodynamic variable.

Add magnetic charge. A dyon carries both electric and magnetic charge, each with its own potential. The first law accommodates both, adding a term alongside the electric term. No exception arises. The magnetic degree of freedom is a thermodynamic variable on the same footing as the electric one.

The universality is complete. Every physical quantity a black hole can possess — mass, spin, electric charge, magnetic charge, vacuum energy density — appears without exception as a thermodynamic variable in the first law. Nothing belonging to the mechanical description is absent from the thermodynamic one.

This is not a list of analogies. When a quantity governs the dynamics of a system in one framework and appears as a thermodynamic variable in an independently developed framework, and when this correspondence holds without exception across every state the system can occupy, the correspondence is not formal. It is identifying something about what those quantities are. The two traditions arrived at black holes by different roads and found, when they compared their descriptions, that they had been writing the same equation in different notation all along.

The convergence has a shape. Bekenstein-Hawking equates the entropy of a black hole with the area of its horizon — in Planck units — a direct identification of a thermodynamic quantity with a geometric one. Information and geometry are the same object. Jacobson's 1995 derivation goes further: the Einstein field equations, the governing equations of spacetime curvature, follow from the Clausius relation applied to local Rindler horizons. General relativity is not imposed on thermodynamics. It emerges from it. The hierarchy this implies runs in the opposite direction from the one classical physics presents: entropy is the primitive quantity, gravity is what emerges from the entropic structure of spacetime, and Newtonian mechanics is the limit of that emergence in regimes of weak curvature and slowly varying entropy. The foundation was never . The foundation was always .

What remains is to ask what is a state of — and what happens to the equations of motion when you take that question seriously.

Entropy is not a free-floating quantity. It is always entropy of something — a system with an internal configuration, a probability distribution over the microstates that configuration can take. The Second Law is a statistical fact precisely because entropy is a property of a described object, not of description-free reality: it measures how many ways the microscopic details of a system can be arranged while leaving its macroscopic appearance unchanged. To place at the foundation is therefore not simply to change which symbol sits on the left side of the equations. It is to ask what kind of object carries entropic information in a way that is prior to both time and mechanics — what the entropic state actually is.

The structure of quantum mechanics offers the clearest analogy for what to look for. In quantum mechanics, the state of a system is described by a wave function — not itself an observable, but the object from which observables are extracted. The Hamiltonian operator acts on the wave function and returns the energy as an eigenvalue. The momentum operator returns the momentum. Every measurable quantity in quantum mechanics is an eigenvalue of some operator acting on the state; the state is the prior object that carries all of the system's available information before any particular measurement is made. What we observe is not the state itself. It is what an operator reads from the state.

If entropy and time are both emergent — if neither is written into the fundamental equations but both are properties that the universe exhibits rather than features it is built from — then the logical structure should be the same. There should be an entropic state that plays the role the wave function plays in quantum mechanics: not itself entropy, not itself time, but the object from which both are extracted by operators acting on it. The observables are the eigenvalues. The state is what those eigenvalues belong to.

The natural candidate is the density matrix . It is a more general object than the wave function: the wave function is a special case of it, describing a pure state with zero entropy, the idealized limit of a system about which everything is known. The density matrix describes the general case — a system in a statistical mixture of quantum states, with a full probability distribution over microstates, carrying the uncertainty that the wave function sets aside. Von Neumann entropy is defined directly on it: . Entropy is not appended to the density matrix from outside. It is read from it. The density matrix is the entropic state.

The operator that extracts a time-like flow from this state already exists in the mathematical formalism. It is the modular Hamiltonian, . Connes and Rovelli identified, in work on thermal time, that the modular Hamiltonian generates a one-parameter flow through the algebra of observables — a flow with precisely the properties we require a time parameter to have. The eigenvalues of encode the information content of individual microstates. Time, in this picture, is not a background parameter against which the system evolves. It is an emergent flow, generated by the operator that reads the entropic state. Thermal time is what modular flow looks like from the inside.

Page and Wootters reach the same structure from the direction of quantum gravity. The Wheeler-DeWitt equation — , the central equation of canonical quantum gravity — carries no . It describes the universe as a timeless constraint rather than a dynamical sequence. Time, in the Page-Wootters picture, is relational: it is encoded in the entanglement between a clock subsystem and the rest of the universe, readable from the correlations in the global density matrix without any external anywhere in the description. An observer experiences time because the quantum state of a clock is correlated with everything else. Reading the clock is reading the state.

There is a difficulty in standard quantum mechanics that this reframing navigates by changing the object rather than resolving the tension. Pauli showed that time cannot be a self-adjoint operator on a conventional Hilbert space — a self-adjoint time operator with the right commutation relations with the Hamiltonian would require the energy spectrum to be unbounded below, which it is not. Time therefore sits awkwardly in the standard framework: essential to the Schrödinger equation as the parameter on its left side, but not an observable in the same sense as energy or momentum, not extractable by an operator acting on the wave function. The tension is usually left unresolved.

In the entropic picture, time is not an operator on the conventional Hilbert space. It is an emergent eigenvalue of acting on the space of density matrices — a different class of object operating in a different space. The constraint Pauli identified applies to self-adjoint operators on bounded Hilbert spaces. The modular Hamiltonian is not that object. The objection does not reach it, not because it is wrong but because it is addressed to the wrong framework.

What the density matrix is to energy in quantum mechanics — the state from which the Hamiltonian extracts the observable — it is also to time. Not two separate functions. The same object, the same underlying state, with two different operators reading two different emergent properties from it. Entropy and time are both eigenvalues of operators acting on the entropic state. They are not the state's inputs. They are its outputs.

From here the question turns toward mechanics. If the density matrix is the entropic state and the modular Hamiltonian is the generator of time, then physics written in the language of is physics written in a derived language — a description of a structure that lives one level deeper. The question is what the equations of motion look like when you write them in that deeper language: when is the parameter, when the rate of entropy production is an explicit quantity rather than a background constant, and when Newtonian mechanics appears as a limiting case rather than a foundation.