Chapter 6 The arrow of time
11 minute read
Every fundamental law of motion at the microscopic scale shares a startling property. Newton's second law of motion, the Euler-Lagrange equations, Einstein's gravitational field equations, and Schrödinger's quantum wave equation all run equally well in both directions. If a camera films any microscopic collision of two atoms or any gravitational orbit of two stars, the reversed movie obeys the exact same laws of physics. Apart from a tiny weak-interaction asymmetry that is balanced when we also invert space and swap matter with antimatter, the microscopic laws of nature are completely reversible in time.
Furthermore, these microscopic laws strictly conserve information. In classical Hamiltonian mechanics, Joseph Liouville proved that the phase-space volume occupied by a cloud of trajectories remains constant forever as the system evolves. In quantum mechanics, unitary time evolution keeps the fine-grained von Neumann entropy strictly constant for all time.
Why, then, does the everyday world around us display such a relentless arrow of time? A porcelain teacup falls from a table and shatters into a hundred shards on the kitchen floor. Yet we never see shards on the floor gather heat from the tiles and leap back into an unbroken cup. A drop of ink disperses through a glass of water, stars pour radiant heat into the cold dark sky, and we remember yesterday while tomorrow remains unknown. Even more mysteriously, in a cosmos marching toward disorder, how do living cells, forests, and human minds build and maintain intricate internal order?
6.1Boltzmann and the statistical origin of entropy
In the 1870s, the Austrian physicist Ludwig Boltzmann solved the riddle of irreversible time. He drew a sharp distinction between the microscopic state of a system and its coarse-grained macroscopic appearance. A single liter of air inside a sealed room contains roughly 25 sextillion molecules bouncing off one another like tiny billiard balls.
A microstate specifies every single molecule's three position coordinates and three momentum components at a given instant. It is a single point in a multidimensional phase space. Because our instruments cannot track 25 sextillion molecules individually, we describe the gas through a macrostate defined by coarse-grained averages such as total energy, volume, pressure, and temperature. Billions upon billions of different microscopic arrangements of molecules look completely identical at the macroscopic level. Each macrostate therefore corresponds not to a single point in phase space, but to a region whose volume counts the number of indistinguishable microstates compatible with that macrostate.
Boltzmann realized that the thermodynamic entropy of a macrostate is simply a logarithmic measure of its microscopic multiplicity. That famous equation is now carved onto his tombstone in Vienna:
Why does nature use a logarithm here? Because if we bring two independent systems together, every microstate of the first system can pair with every microstate of the second. The total number of combined microstates multiplies, and the logarithm turns that product of possibilities into an additive sum of entropies.
To grasp the sheer combinatorial force of Boltzmann's formula, we can compare two macrostates of a box of gas molecules. In the first macrostate, all the molecules happen to be crowded into the left half of the box. In the second macrostate, the molecules are spread evenly between the left and right halves. If the box holds just four molecules, two raised to the fourth power is 16, so seeing all four molecules wander into the left half happens one time in 16. For a single mole of air, however, the evenly spread volume in phase space exceeds the half-empty volume by two raised to the power of 600 sextillion.
In phase space, the thermal equilibrium macrostate occupies virtually the entire energy surface. Low-entropy non-equilibrium states occupy unimaginably tiny specks of volume. A system does not move from low entropy to high entropy because a mysterious one-way force pushes it forward in time. Rather, if a system starts in a tiny low-entropy region, almost every trajectory leaving that region enters vastly larger phase-space volumes. Those volumes are simply overwhelmingly bigger. That statistical climb from rare macrostates to typical macrostates is the second law of thermodynamics, .
When a small system sits in thermal contact with a large heat reservoir at a fixed temperature, how likely is each microstate of a given energy? Maximizing the entropy at fixed average energy yields the canonical Boltzmann-Gibbs probability distribution, . Here the normalizing sum over all states is the partition function. From that single partition function, every equilibrium property of matter follows by simple differentiation. That includes internal energy, entropy, and the Helmholtz free energy available to perform useful work.
6.2Open systems and dissipative structures
If the second law demands that total entropy never decrease, how can a living bacterium, a redwood tree, or a human brain build molecular order and resist thermal decay?
In his 1943 Dublin lectures, published in 1944 as What Is Life?, Erwin Schrödinger addressed this question. Ilya Prigogine later placed the answer on a rigorous mathematical foundation through non-equilibrium thermodynamics. The classical statement that entropy always rises to a maximum applies only to an isolated system cut off from all exchanges of energy and matter with the outside world. A living organism, a planet, or a convecting fluid is not an isolated box; it is an open system through which energy and matter continuously flow.
For any open system, the small change in its own entropy during a time step splits into two parts, . The internal entropy production comes from irreversible friction, chemical reactions, and heat conduction inside the system. By the second law, this internal production can never be negative. The boundary flux term records the entropy carried across the border as the system exchanges heat and matter with its surroundings. Unlike internal production, the boundary flux can be either positive or negative.
Suppose an open system absorbs low-entropy free energy from its environment and exports a larger amount of degraded heat across its boundary. Then the net entropy of the system itself decreases or stays locked at a low-entropy steady state. The second law of thermodynamics is never violated, because the combined entropy of the system plus its environment still rises. As Schrödinger put it, a living organism stays alive by drawing negative entropy from its environment and heating the surrounding universe.
Earth's entire biosphere runs on this thermodynamic engine. Every second, the Earth intercepts high-grade optical photons from the Sun's 5,800-kelvin surface. It radiates an equal amount of energy back into space as 255-kelvin infrared photons. Because each solar photon carries roughly 23 times the energy of an infrared photon, the Earth radiates 23 infrared photons per solar photon. That 23-fold multiplication of photon number exports vast entropy into space. Green plants, ocean currents, and animals tap a tiny fraction of that free-energy waterfall to build the covalent bonds of life.
Prigogine discovered something even more remarkable when an open system is pushed far from thermodynamic equilibrium by a strong external gradient of temperature or chemical potential. Near equilibrium, small fluctuations simply damp out and the system settles into a uniform, featureless state. Beyond a critical threshold far from equilibrium, however, uniform states become unstable. The system spontaneously leaps into self-organized macroscopic patterns that Prigogine named dissipative structures.
When a shallow pan of oil is heated from below, the fluid above a critical temperature difference organizes into a honeycomb lattice of rolling convection cells. Each cell coordinates the motion of 10 sextillion molecules. Similarly, when chemical reagents mix in an open reactor, spiral chemical waves pulse like a beating heart. These self-organized structures emerge precisely because ordered coherent motion transports heat and dissipates the external free-energy gradient faster than disorganized random collisions ever could.
At the turn of the 21st century, Christopher Jarzynski and Gavin Crooks extended this non-equilibrium physics all the way down to the scale of individual molecules buffeted by thermal noise. Suppose a microscopic system—such as a single RNA hairpin or protein motor in warm water—is driven away from equilibrium along a forward microscopic path. Comparing the probability of observing that forward path to the probability of observing the exact time-reversed path when the protocol runs backward yields Crooks' fluctuation theorem:
Crooks' fluctuation theorem reveals a deep truth about the emergence of irreversibility and self-replication. A forward physical process becomes statistically irreversible when the forward trajectory is vastly more likely than its time-reverse. By Crooks' formula, that happens only when the system absorbs free energy and dissipates positive entropy into its surroundings. As the physicist Jeremy England showed in 2013, this sets a fundamental thermodynamic constraint on self-replication. Any molecular assembly that copies itself rapidly compared to its spontaneous decay rate must be an efficient engine for absorbing external work and dissipating heat.
6.3The past hypothesis and cosmic expansion
A subtle logical trap still lurks inside Boltzmann's statistical explanation of the second law. Boltzmann showed that if a system sits in a tiny low-entropy macrostate today, almost every trajectory leaving that state enters vastly larger phase-space volumes tomorrow. Yet the underlying microscopic laws of physics are completely symmetric between past and future. If we know only that a system sits in a low-entropy state right now, tracing its trajectories backward in time predicts that entropy was also higher in the past. Why was yesterday's entropy lower than today's, rather than higher?
The answer—traced by Richard Feynman, Roger Penrose, and the philosopher David Albert, who named it the past hypothesis—is cosmological. Statistical mechanics explains why entropy rises once a system starts in a low-entropy state. However, the global arrow of time depends on an extraordinary boundary condition at the beginning: the universe was born 13.8 billion years ago in an astonishingly low-entropy state.
At first glance, that claim sounds upside down. Radio telescopes mapping the cosmic microwave background reveal a primordial plasma that was nearly uniform across the sky, with temperature ripples of only one part in 100,000. Wait: in Section 6.1, wasn't a uniform gas spread evenly across a box the state of maximum entropy?
For an ordinary gas in a small box where gravity is negligible, uniformity is indeed maximum entropy. When gravity dominates across cosmic scales, however, the exact opposite is true. Because gravity is universally attractive and never screens itself, a smooth, uniform sea of matter spread evenly across space is like a trillion pencils balanced on their tips. Gravity pulls matter together into dense clumps—stars, galaxies, and ultimately black holes. While thermal pressure wants gas to spread out uniformly, gravitational entropy grows when matter clumps together.
Using the Bekenstein-Hawking black hole entropy formula from Chapter 4, Roger Penrose tallied the entropy budget of our observable universe. At the time of the cosmic microwave background, the thermal entropy of that smooth hot radiation and gas was roughly 10 to the 88th power in Boltzmann units. Today, after gravity has clumped matter into galaxies housing supermassive black holes at their centers, the entropy of our observable universe has risen to 10 to the 104th power. If all the mass inside our observable horizon collapsed into a single gigantic black hole, its Bekenstein-Hawking entropy would reach 10 to the 123rd power.
Our universe began with smooth spacetime geometry and vanishing gravitational clumping, giving it a vast reservoir of gravitational free energy. Every star that condenses out of interstellar gas, ignites nuclear fusion, and pours bright sunlight into cold space is cashing a check from that primordial gravitational bank account.
How does the spacetime stage itself evolve as this cosmic drama unfolds? On scales larger than 300 million light-years, observations confirm the cosmological principle: the universe is spatially homogeneous (the same everywhere) and isotropic (the same in every direction). Applying this symmetry to Einstein's field equations yields the Friedmann-Lemaître-Robertson-Walker geometry. In this geometry, physical distances between galaxies stretch in proportion to a dimensionless cosmic scale factor governed by the first Friedmann equation:
Here the ratio on the left is the squared Hubble expansion rate, first measured observationally by Georges Lemaître in 1927 and Edwin Hubble in 1929. On the right sit the three competing drivers of cosmic evolution: the total equivalent mass density of matter and radiation, the spatial curvature constant, and the cosmological constant. For any cosmic fluid with a constant ratio between pressure and energy density, local energy conservation dictates that its density dilutes with expansion as .
This single scaling law choreographs the entire four-act history of our expanding universe. In the opening inflationary epoch during the first fraction of a second, proposed by Alan Guth, Andrei Linde, and Alexei Starobinsky, space was dominated by a primordial inflaton field. That vacuum energy drove an exponential expansion that stretched the infant cosmos by at least 26 orders of magnitude in the blink of an eye. Inflation ironed out any initial spatial curvature and magnified microscopic quantum fluctuations into the tiny density seeds from which all galaxies later grew.
When the inflaton decayed into a blazing soup of relativistic particles, the cosmos entered the radiation-dominated era, lasting until about 50,000 years after the Big Bang. As space expands, photon number density falls with volume as the inverse cube of the scale factor, while each photon's wavelength stretches with the expansion. Radiation energy density therefore falls as the inverse fourth power of the scale factor. During the first three minutes, as the cosmic furnace cooled through a billion kelvins, protons and neutrons fused into the primordial light elements: roughly 75% hydrogen and 25% helium.
Because radiation dilutes with the fourth power of the scale factor while slow-moving matter dilutes with the third power, matter overtook radiation around 50,000 years, inaugurating the matter-dominated era. At 380,000 years, the expanding plasma cooled to 3,000 kelvins, allowing free electrons to bind with nuclei into neutral hydrogen and helium atoms. Suddenly transparent, the universe released the cosmic microwave background light, while cold dark matter and ordinary gas fell into primordial gravitational wells to ignite the first stars.
Finally, as matter diluted across expanding space, empty space's unchanging dark energy began accelerating cosmic expansion about six billion years ago and overtook matter density 3.5 billion years ago. Because vacuum energy has negative effective pressure, it acts in Einstein's equations as a cosmic repulsion, driving the universe toward an exponential future. Yet during the multibillion-year window of the matter-dominated era, inside the nuclear furnaces of massive stars, primordial hydrogen and helium fused into heavier nuclei. Those stellar furnaces forged carbon, nitrogen, oxygen, phosphorus, and iron—the raw atomic alphabet from which chemistry and living matter would arise.