Does Physics Need Causes?

Imagine you’re playing a game of pool. You line up your shot, strike the cue ball, and watch it roll across the table. When it hits the eight ball, the eight ball moves too. You’d probably say: the cue ball caused the eight ball to move.

That seems obvious. But now imagine a physicist describing the same scene. She might not say anything about causes at all. Instead, she’d write down equations—formulas that describe the balls’ positions, speeds, and directions at every moment. In her world, there are no “causes” and “effects.” There are just mathematical relationships.

This raises a strange question: do causes actually exist in the physical world, or are they something we humans add to our descriptions of the world? If physics—our most fundamental description of reality—doesn’t need the idea of causes, then maybe causes aren’t real at all.

For over a hundred years, philosophers have been arguing about this. The debate is still alive today.

The Attack on Causes

In 1912, the philosopher Bertrand Russell wrote a famous paper called “On the Notion of Cause.” His argument was blunt: the word “cause” is so vague and confused that it should be thrown out of physics entirely. When physicists write down equations, Russell said, they don’t talk about causes. They talk about what he called “functional dependencies”—mathematical relationships between measurements at different times. And those relationships work just as well in both directions. If Newton’s laws tell you what a pool ball will do after you hit it, they also tell you what must have happened before it ended up where it is. There’s nothing in the equations themselves that says “this is the cause and that is the effect.”

Russell wasn’t the first to think this. The physicist Ernst Mach had said similar things a few decades earlier. And modern philosophers have developed their arguments into five main challenges. Let’s look at the most important ones.

The Vagueness Challenge

Philosophers love precise definitions. But when they try to define “cause,” they run into trouble. In everyday life, we say things like “the short circuit caused the fire” or “your teasing caused her to cry.” But what exactly makes something a cause? Is the oxygen in the air a cause of the fire? If not, why not? Without oxygen, there’d be no fire.

Physicists’ equations, by contrast, are mathematically precise. They give you exact numbers. Causal talk, the argument goes, is just too fuzzy to belong in physics. When physicists use words like “causes” in their textbooks, they’re just speaking informally—it’s not part of the real content of the theory.

The Determinism Challenge

Causes are supposed to determine their effects. If you hit the cue ball in exactly the same way twice, the eight ball should move exactly the same way twice. That’s what “same cause, same effect” means.

But modern physics isn’t always deterministic. In quantum mechanics, the same initial conditions can lead to different outcomes. You can’t predict exactly what will happen—only the probabilities. If causes have to determine their effects, and quantum mechanics says effects aren’t always determined, then maybe there are no causes in the quantum world.

The Time-Asymmetry Challenge

This is probably the most powerful argument. We think causes come before effects. The cue ball hits the eight ball before the eight ball moves. But the fundamental equations of physics don’t care about time’s direction. They work the same forward and backward. If you filmed a pool shot and played it backwards, the physics would still look perfectly fine—it would just show the balls moving in reverse. The equations can’t tell you which direction is “cause” and which is “effect.”

So where does our sense that causes come before effects come from? It can’t come from the laws of physics themselves.

The Dominant Cause Challenge

When you say “the short circuit caused the fire,” you’re picking out one factor from an enormous web of conditions. You’re ignoring the presence of oxygen, the flammable materials, the lack of a working sprinkler system, and millions of other details. But from the perspective of physics, all of these are necessary. The laws of physics say: here’s the complete state of the system at one time, and here’s what follows. Nothing less will do.

So what makes the short circuit special? Why do we call it the cause and everything else “background conditions”? The answer seems to be that we humans decide, based on our interests and purposes. The physics itself doesn’t tell us. And if physics itself can’t distinguish causes from background conditions, then causes aren’t really part of physics.

Defending Causes

These are serious challenges. But many philosophers think they can be answered. One important response focuses on the fact that physics doesn’t just consist of laws. It also involves initial conditions—the specific state the universe happened to start in.

Here’s the key idea. Our universe seems to have started in an extremely special, low-entropy state. (Entropy is a measure of disorder. A tidy room has low entropy; a messy one has high entropy.) The universe began very orderly and has been becoming more disorderly ever since. This might not seem important, but philosophers like David Albert and Barry Loewer have argued that it makes all the difference.

Think about what happens when you drop an ice cube into warm water. The ice melts, and the water cools slightly. The physics of this process works the same backward in time—if you saw a video of ice forming from water, that would also obey the laws of physics. But that never actually happens. Why? Because it would require absurdly special initial conditions—the water molecules would all have to be moving in just the right way to produce ice. The universe didn’t start out that way.

This asymmetry between the past and the future—the universe’s past is special and orderly, while its future is not—is what gives us a direction for causation. When we say causes come before effects, we’re really relying on the fact that earlier states of the universe are more special and more organized than later ones.

Common Cause Reasoning

Consider a real example. In 2016, scientists detected gravitational waves from two colliding black holes. They didn’t directly observe the black holes. Instead, they detected signals at two detectors—one in Washington state and one in Louisiana. The signals were almost perfectly correlated: they arrived at the same time and had the same shape.

How did scientists know these signals came from a single event far away, rather than from two separate random events on Earth? They used what’s called common cause reasoning. The best explanation for two highly correlated events is that they share a common cause in their past—the black hole collision. The alternative would require an unbelievable coincidence: two independent sources just happening to produce identical signals at the same moment.

This kind of reasoning is everywhere in physics. It’s also everywhere in everyday life. If you and your friend both get the same rare disease at the same time, you look for something you both did—maybe you ate the same food. That’s common cause reasoning too.

Crucially, this reasoning relies on an assumption: that initial conditions are random and uncorrelated, while final conditions are not. In other words, it’s easier to have two effects of a single cause than to have two independent causes that just happen to line up. This assumption is not built into the laws of physics themselves. It comes from the fact that the universe started in that special, low-entropy state. But once we add that assumption, we can see why causal reasoning has a place in physics.

A Tool, Not a Feature of Reality?

Notice what’s happening here. The philosophers defending causes aren’t saying that causes are part of the fundamental fabric of the universe. They’re saying that causal reasoning is a useful tool we use to understand the world, and that this tool has a place in physics as well as in everyday life.

This brings us to an important distinction. Philosophers distinguish between three different projects when they talk about causation:

The metaphysical project asks: Are causes real? Do they exist in the world independently of us? This is where the Russellian challenges hit hardest. If causes aren’t in the fundamental equations, maybe they’re not fundamental features of reality.

The descriptive project asks: What do we mean when we say “cause”? What’s the concept we’re actually using?

The functional project asks: What does causal thinking do for us? What purposes does it serve? This is where the defense becomes strongest.

From a functional perspective, the question isn’t whether causes are “really real.” It’s whether causal concepts help us understand, predict, and control the world. And they clearly do. The scientists who detected gravitational waves used causal reasoning to infer the black hole collision. Engineers use causal reasoning to design bridges and computers. You use causal reasoning every time you decide to study for a test because studying causes better grades.

The neo-Russellians might be right that causal language doesn’t appear in the fundamental equations of physics. But that doesn’t mean causal reasoning plays no role in physics at all. Physics involves more than equations—it involves experiments, data analysis, and explanations. And all of those involve causal reasoning.

The Quantum Question

Quantum mechanics makes this even stranger. In certain experiments, subatomic particles that are far apart seem to be connected in ways that can’t be explained by any ordinary cause-and-effect story. Two particles can be “entangled” so that measuring one instantly determines the state of the other, even if they’re light-years apart.

Some physicists and philosophers think this means we need to give up on causation entirely in the quantum realm. Others are working to develop new kinds of causal models that could handle quantum weirdness. This is an active area of research.

So Where Are We?

The debate about causation in physics isn’t settled. Here’s what most philosophers now agree on:

The fundamental equations of physics don’t talk about causes. They describe mathematical relationships.
But physics as practiced—as a human activity involving experiments, explanations, and predictions—uses causal reasoning all the time.
The direction of causation (cause before effect) seems to come from a fact about our universe’s history, not from the laws of physics themselves.
Whether “causes” are real features of the world or just useful tools is still an open question.

Maybe causation is like the lines of latitude and longitude on a map. They’re not actually drawn on the Earth’s surface. But they’re incredibly useful for navigating. And nobody thinks we should stop using maps just because the lines aren’t “real.”

Or maybe causation is like heat—something we used to think was a fundamental feature of the world, but that turned out to be something else entirely. Only time (and more philosophy) will tell.

Appendices

Key Terms

Term	What it does in this debate
Cause	The thing we think produces an effect; the main concept under debate
Effect	What we think a cause produces
Determinism	The idea that the same causes always produce the same effects
Entropy	A measure of disorder; the universe started with low entropy and has been increasing it
Common cause	A single event that explains why two other events are correlated
Initial conditions	The specific state the universe (or a system) started in
Functional project	Asking what causal concepts do for us, rather than whether they’re “real”
Time-reversal symmetry	When the laws of physics work the same backward and forward in time
Intervention	Changing one part of a system to see what happens; a key part of causal reasoning

Key People

Bertrand Russell (1872–1970): A British philosopher who argued that the concept of cause is so vague it should be thrown out of philosophy and physics.
Ernst Mach (1838–1916): An Austrian physicist and philosopher who argued that causal language is too imprecise for the mathematically exact science of physics.
David Albert (born 1954): A contemporary philosopher of physics who argues that the special initial state of the universe (the “Past Hypothesis”) is what gives us a direction for causation.
Barry Loewer (born 1945): A philosopher who works with Albert on connecting the universe’s initial state to the asymmetry of causation.
James Woodward (born 1944): A philosopher who developed an “interventionist” account of causation, focusing on what causal concepts do for us (the functional project).

Things to Think About

If the laws of physics work the same forward and backward, why do we experience time as moving in only one direction? Is the “arrow of time” the same thing as the “arrow of causation”?
Imagine you’re designing a robot that needs to learn about the world through sensors. Would the robot need the concept of cause to make sense of its data, or could it get by with just patterns and correlations?
The article says that common cause reasoning relies on the assumption that initial conditions are random and uncorrelated. But what if our universe didn’t start that way? Could we still make sense of causation?
Suppose scientists discovered that everything in the universe is actually determined (no randomness at all). Would that make causation clearer or more mysterious?

Where This Shows Up

Artificial intelligence: Machine learning systems often need to distinguish correlation from causation to make good predictions. This is literally the same problem philosophers are debating.
Everyday arguments: When you say “you made me angry,” you’re making a causal claim. But is it really that simple? The debate about causes in physics connects to how we think about responsibility and blame.
Science education: Physics textbooks constantly use causal language (“the force causes the object to accelerate”), even though the underlying equations don’t mention causes. Teachers and textbook writers face this tension every day.
Medical research: When scientists test whether a drug works, they’re trying to figure out whether it causes improvement. The philosophical debate about causation has practical implications for how we design experiments and interpret data.