The early universe can be close to thermal equilibrium in its matter and radiation while still presenting a profound question about gravitational order. Understanding that distinction prevents the low-entropy problem from becoming a contradiction manufactured by using one word for different descriptions.
Expansion need not generate radiation entropy
Consider thermal radiation evolving adiabatically in a comoving cell. Its entropy density is proportional to T³, while the cell's physical volume is proportional to a³. If T scales as 1/a, its total radiation entropy scales as a³T³ and remains constant.
Double a in this example. The volume increases eightfold and the temperature falls by one half. Entropy density falls to one eighth, exactly compensating for the larger volume. Expansion and cooling occur, but neither calculation establishes entropy production.
Introduce irreversible interactions or energy transfer and the answer can change. The point is to identify the process and the system whose entropy is being evaluated. A growing scale factor is not, by itself, a formula for an increasing entropy.
The earlier essay on open systems supplies the same discipline locally. Entropy can enter, leave or be produced inside a chosen subsystem. In cosmology, choosing a comoving cell, an observable region or a horizon-bounded region changes the accounting question.
Gravity changes which configurations matter
For a gas in a box with negligible self-gravity, a nearly uniform equilibrium state is familiar. A sufficiently large self-gravitating system has different possibilities: clustering, collapse and black-hole formation can open very different macroscopic configurations.
Consequently, describing an early plasma as almost thermal does not settle the entropy of its gravitational degrees of freedom. The special smoothness of the early geometry is part of the question. There is no single universally applicable local gravitational entropy density that lets us assign a simple number to every arbitrary spacetime.
This qualification matters when asking why the early state was special. An explanation must identify a state space, relevant constraints and a measure of typicality. Merely calling an initial condition “unlikely” leaves those ingredients unspecified.
A calculable gravitational benchmark
Black-hole thermodynamics supplies a more definite case. The Bekenstein–Hawking entropy is S = kBc³A/(4Għ), where A is horizon area. Bekenstein developed the area–entropy connection; Hawking's semiclassical radiation calculation fixes the familiar normalization through black-hole temperature.
For a Schwarzschild black hole, A = 16πG²M²/c⁴, so S = 4πkBGM²/(ħc). Doubling the mass multiplies this entropy by four. The quadratic scaling is a striking difference from an extensive entropy proportional to the amount of ordinary material.
Compare two widely separated nonrotating holes of mass M with one of mass 2M. Adding the two initial horizon entropies gives 2S(M); the final idealized value is 4S(M), twice that sum. This comparison neglects interactions and radiated energy. It is an algebraic benchmark, not an exact prediction of a physical merger.
If the final nonrotating mass were instead 1.9M, its entropy would be 3.61S(M), or 1.805 times the initial sum. Real mergers can leave spin and radiation, so the appropriate horizon area and exterior entropy must be included. The explicit assumptions make the simple calculation useful without pretending it is complete.
Why a cosmic arrow needs a boundary account
Local dissipative mechanisms explain how entropy can increase under specified conditions. They do not automatically explain why the universe supplies the gradients, records and special earlier state needed for those mechanisms to operate in a common temporal direction.
The SMM/DTT distinction between inner generation and completed history invites a question at this level: what restricts the allowed cosmic histories or their boundary conditions? An update rule that preserves records might support temporal asymmetry, but a cosmological account must also say which initial or boundary states are selected.
The book separates renewal and continuity from cosmological beginning conditions. Keeping that distinction permits two complementary investigations: how a history remains connected, and why this history has its observed large-scale organization.
A productive proposal would recover the adiabatic radiation limit, identify irreversible departures from it, and explain its gravitational entropy accounting. Those are concrete steps toward a cosmic arrow, rather than an inference from the existence of expansion alone.
Sources and further reading
- J. D. Bekenstein, Black Holes and Entropy, Physical Review D 7, 2333–2346 (1973).
- S. W. Hawking, Particle creation by black holes, Communications in Mathematical Physics 43, 199–220 (1975).
- Continue with the arrow of time and what an explanation of the arrow requires. The scaling and mass comparisons here are teaching calculations.

