Time goes in only one direction
Physicists say that entropy sets the direction of time. A broken teacup falls apart; it does not spring back together. Heat spreads and dissipates; it does not increase without a source of new heat. Unless the second law of thermodynamics is false, aggregate entropy in the universe can only increase, not decrease. Entropy advances inexorably—the force that points the arrow of time toward the future.
All things in spacetime therefore move only in one direction ultimately. That has been true since at least the Big Bang, the immediate result of which was a universe of very low entropy. Exploding from perhaps a microscopic singularity, the universe became almost instantaneously a massive macrocosm of ordered, compressed, very hot energy. That macrocosm has been expanding and expending that energy since, and there was ever only one direction for it to go—toward a state of lower heat and higher entropy.
But what makes time go at all?
My question is different from the one answered by entropy. I want to know not why time goes in one direction, but simply why time goes.
When you wind a pendulum clock, there is only one direction for the springs to go. They unwind; they do not wind up if left alone. They do not move if left alone, unless and until the pendulum moves. What swings the pendulum? What starts the clock?
If the universe could go in only one direction after the Big Bang, why did it go? Why didn’t the universe stand still without moving forward or backward? Why did the universe go in any direction?
Microcosmic quantum reality is not time as we know it
A physicist might say that the universe is a place of constant quantum activity, with so much inherent quantum evolution of systems or general quantum fluctuation that it is impossible for the universe ever to stand still. The universe must go; the quantum state of the universe demands it.
But the microcosmic quantum world may not follow the arrow of time or depend on entropy in the way prescribed in the macrocosmic world. A quantum wave function describes the probability amplitudes of different possibilities, all of which can remain in a state of superposition in the microcosm. Schrödinger’s cat can be both dead and alive. Without a single, unique state, it has no definite history of unique moments following one after the other. Instead, the cat exists in a state of constant possibility evolving into more possibilities.
It is not until the cat interacts with the macroscopic world that its fate is known—or that it has any fate at all. Without the macrocosm, the quantum world is little more than a soup of constant interaction and simultaneous possibility. And that quantum soup does not have a direct relationship with time or forward direction. At the quantum level there very likely is no time and no direction in time; there is only the quantum wave function.
Time requires more
The quantum microcosm underlies the grand sweep of history, but it is not history. History requires more than a quantum puzzle of entangled superpositions; it requires change and movement, actual unique events, a macrocosmic world evolving over billions of years. History and time assume a quantum wave function that results in more than superpositions with constantly changing probability amplitudes, but instead produces a stream of unique macrocosmic outcomes, a stream of history.
Time measures change and movement
Probability is a function of time, measuring the likelihood of change and movement. At its core the quantum wave function is a probability distribution that both calculates abstract possibilities in superposition and measures the likelihood of detecting an actual event[1] in a particular location and time. It effectively assumes the existence of time as a medium in which probabilities have the potential to resolve into outcomes.[2]
The quantum wave function exists in two phases. In the first phase, when a system is isolated, the wave function exists as superpositioned probability amplitudes evolving according to the deterministic rules of the Schrödinger equation. In the second phase, when a system is not isolated, when the wave function interacts with something distinct from itself—something in the macroscopic world—it evolves in a new and more random way. Instead of a microcosmic wave function of probability amplitudes, it becomes a set of actual probabilities that resolve into one unique outcome in the macrocosm. Essentially, the dice are thrown, resulting in a unique moment in time and history.[3]
We don’t know why or how interaction with the macroscopic world causes probability amplitudes to resolve into outcomes. We don’t know if quantum wave functions “collapse” to produce unique outcomes or if the universe splits into “many worlds” at every quantum intersection. We know only that the universe includes some form of interaction between the macrocosmic and quantum worlds that results in probabilities becoming outcomes—that produces a stream of outcomes.
Quantum wave functions as we know them interact with the macrocosm and resolve calculable sets of probabilities into unique outcomes. Each outcome influences another wave function and another set of probabilities, which resolves again into a new unique outcome. That is how the macrocosm functions today. That is how history progresses. That is how time works.
How did a quantum soup become a macrocosm of history and time?
Perhaps before the Big Bang there was only quantum fluctuation or isolated quantum evolution, with no history and no time. A world of possibilities and probabilities only, without direction, without outcomes, without resolution. There was no spacetime and no macrocosmic world, perhaps only a universal state of complete, unbroken entanglement.
In such a pre-Big Bang universe, there could have been no possible interaction between a macrocosmic world and the microcosmic quantum world. There was only one thoroughly entangled universe with no interactive mechanism for possibilities to become outcomes. Wave functions that interact with a macrocosm, driving the universe forward, resolving probabilities into outcomes—did not exist at all.
Yet that timeless quantum world eventually fluctuated into an explosive state that produced the Big Bang. The result was a macrocosmic world of massive, organized energy with very low entropy, a disentangled universe waiting for a new form of quantum interaction.
Did the Big Bang create both the arrow of time and the engine of time?
If interaction with the macroscopic world causes quantum wave functions to resolve into outcomes, did the state of low entropy that followed the Big Bang set the stage for that macroscopic interaction? Did the Big Bang create just enough separation and disentanglement in the universe to start the chain of collapses and unique outcomes that we know as history? Just enough of whatever it took to cause quantum wave functions to resolve into unique events, pushing all other possible outcomes into the realm of imagination or many worlds? Is the wave function the engine of time?
Does the wave function create the great illusion?
It may be reasonable to think of microcosmic quantum reality as a world of possibilities only—a world where all things are possible, a world that mathematically defines the probabilities of all outcomes. But a world without a mechanism for turning probabilities into outcomes.
It is the macrocosmic world that is about outcomes. The great illusion that is the macrocosmic world may be the mechanism by which the universe creates outcomes. Perhaps when quantum fluctuation caused the Big Bang and created spacetime, that new macrocosm became a Petri dish for a new form of quantum interaction that enabled probabilities to resolve into outcomes, starting the clock of history and birthing the great illusion that we live in today.

[1] Which may or may not be equivalent to the likelihood of an event occurring in a particular location and time.
[2] A probability and outcome assume movement and change. Something is probable and then something occurs. Does the concept of probability have meaning without the possibility of a future outcome? Perhaps. But is difficult to conceive of a world of probability that would not eventually, in some way, experience an outcome.
[3] Roger Penrose has suggested that this result, called quantum state-vector reduction, could be explained as a gravitational phenomenon in a yet-to-be-specified theory of quantum gravity. “My own point of view is that as soon as a ‘significant’ amount of space-time curvature is introduced, the rules of quantum linear superposition must fail. It is here that the complex-amplitude superpositions of potentially alternative states become replaced by probability-weighted actual alternatives—and one of the alternatives indeed actually takes place.” Penrose (1989), p. 475 (emphasis in original).
