When Science Has "Fine Print": The Puzzle of Ceteris Paribus Laws

Imagine you’re playing catch in the backyard. You throw a ball, and your friend catches it. Now imagine a scientist trying to write a law that describes what happens when you throw a ball. They might write something like: “If you throw a ball, it will travel in a curved path and eventually hit the ground.”

But that’s not always true, is it? What if the wind catches it? What if your friend catches it mid-air? What if a bird snatches it? The simple law about thrown balls has exceptions—lots of them.

This creates a puzzle for scientists and philosophers. The most famous laws in physics—like Newton’s law that force equals mass times acceleration—seem to work everywhere, every time. But when you look at biology, psychology, or economics, the “laws” there are full of exceptions. Birds fly—except penguins, ostriches, and injured birds. People act in their own self-interest—except when they don’t. An increase in demand causes prices to rise—except when governments regulate prices, or when companies are giving things away, or when…

So here’s the question: can we call these things “laws” at all? Or are they something else? And if they are laws, what exactly are they saying?

What Makes a Law a Law?

Philosophers used to think that a real law of nature had to be three things: true, universal (no exceptions), and support counterfactuals—that’s a fancy way of saying it tells you what would happen in a situation that hasn’t actually occurred. “If you had thrown that ball harder, it would have traveled farther” is a counterfactual. The law about thrown balls supports that claim.

But the problem is that almost nothing outside of fundamental physics seems to meet all three requirements. Biologists, psychologists, and economists have genuine knowledge. They can predict things and explain things. But their generalizations have exceptions.

This leaves philosophers with three uncomfortable options:

Tough luck, special sciences. If the generalizations in biology and psychology aren’t universal, they aren’t real laws. And if you need laws to have real explanations, then those fields don’t give real explanations either. Most philosophers think this is too harsh.
Who needs laws anyway? Maybe explanation and prediction don’t require anything as strict as laws. The special sciences are doing fine; it’s the concept of “law” that needs to change.
These are laws—just with fine print. The generalizations in special sciences really are laws, but they come with an invisible clause: ceteris paribus, Latin for “other things being equal.” Birds can fly, ceteris paribus—assuming nothing interferes.

Most of the debate has been about option 3. And the big question is: what does that “ceteris paribus” actually mean?

The Big Problem: Falsity or Trivia

Imagine you say: “Ceteris paribus, birds fly.” What are you actually claiming?

If you mean “all birds fly,” that’s just false—penguins are birds, and they don’t fly. So the law would be false.

But if you mean “all birds fly, unless something interferes,” then what do you mean by “unless something interferes”? It seems like you’re saying: “All birds fly, unless they don’t.” That’s trivially true—it’s true no matter what. It’s like saying “either it will rain tomorrow, or it won’t.” That doesn’t tell you anything useful.

This is what philosophers call the dilemma of ceteris paribus laws. Either they’re false (which makes them useless for explaining things), or they’re trivially true (which also makes them useless for explaining things). Either way, they seem to fail.

So how do we rescue these generalizations?

Strategy 1: The Promise to Explain

One group of philosophers says: when scientists say “ceteris paribus,” they’re making a promise. The promise is: if the law fails in a particular case, we can explain why it failed by pointing to some specific interfering factor.

Think of it like a check. When someone writes you a check, they’re promising that the money is actually there. A ceteris paribus law works the same way. If the law says “birds fly” and someone finds a bird that doesn’t, you need to be able to say “oh, that bird has a broken wing” or “that bird is a penguin.” If you can’t explain the exception—if you just keep saying “ceteris paribus” without being able to say what went wrong—then the law was never any good.

This is an appealing idea, but it has a problem. Suppose someone says: “Ceteris paribus, all spherical objects conduct electricity.” If you find a spherical object that doesn’t conduct electricity, you can always explain it: “Oh, that one has a molecular structure that prevents it.” But wait—everything has some molecular structure that explains why it does or doesn’t conduct electricity. So this “law” turns out to be almost empty. You can claim it for anything, because you can always point to some explanation for why the exception happened.

The problem, in other words, is that the “promise” is too easy to keep.

Strategy 2: Staying Power

Another group of philosophers thinks the key feature of a law is its stability or invariance—how well it holds up when you imagine changing things.

Imagine you have a rubber band. Hooke’s law says that the force needed to stretch the band is proportional to how much you stretch it. This works pretty well—as long as you don’t stretch the band too far, or heat it up too much, or do anything weird to it.

But notice: even fundamental laws of physics have this character. Newton’s law of gravity works fine for apples and planets, but breaks down near black holes. The difference between fundamental laws and special science laws might be a matter of degree, not kind. Fundamental laws hold under a very wide range of counterfactual suppositions. Special science laws hold under a narrower range, but that doesn’t make them not-laws.

On this view, a law is anything that stays true under enough of the counterfactual situations that matter for a particular science. An economist doesn’t care whether the law of supply and demand holds for a planet where comets destroy the economy every Tuesday—that’s not the kind of situation economics is about. The law doesn’t need to hold everywhere and always; it just needs to hold where and when the science actually uses it.

This approach solves the dilemma nicely. The law isn’t false—it’s true for the range of cases that matter. And it isn’t trivially true, because there are plenty of imaginable situations where it would fail.

Strategy 3: Hidden Powers

A third approach goes back to the 19th-century philosopher John Stuart Mill. Mill said something clever: there’s no such thing as an exception to a law. When a balloon doesn’t fall to the ground, that’s not an exception to the law of gravity—it’s just that another law (the law of buoyancy) is also operating. Both laws are working all the time; they just happen to produce different results when combined.

On this view, laws describe tendencies or capacities, not actual behavior. The law of gravity says that heavy objects tend to fall—they have a capacity to fall, which might be masked or overridden by other forces. When you see a balloon floating, you’re not seeing a failure of gravity; you’re seeing gravity at work plus air pressure at work.

This is appealing because it means laws can be strictly true—they just describe capacities rather than outcomes. The law “all heavy bodies tend to fall” is true of every heavy body, even balloons. The balloon still has the tendency; it’s just that another tendency (to float) is stronger.

But this approach has its own problem. If laws describe tendencies, how do you know what tendency something has? You can’t just look at what it does, because what it does depends on multiple tendencies. And if you can’t independently identify the tendency, the explanation becomes circular: “The law is true because objects have this tendency, and we know about the tendency from the law.”

Strategy 4: What’s Normal

A final approach says that ceteris paribus laws describe what happens under normal conditions. “Birds fly” means “under normal conditions, birds fly.” A penguin is abnormal—it’s an exception to what’s normal for birds.

“Normal” here doesn’t mean “morally good.” It means something like “statistically typical” or “what we usually expect.” Normal birds have working wings and can fly. Normal humans don’t spontaneously explode. Normal economies follow the law of supply and demand.

This works especially well for the “life sciences”—biology, psychology, social sciences—because living things have evolved to be a certain way. Evolution doesn’t produce perfect systems; it produces systems that work most of the time in normal environments. A bird that can’t fly is a bird that will probably die before reproducing, so evolution weeds out the non-fliers. That’s why “normal” birds fly—not because it’s a universal law, but because the ones that couldn’t fly didn’t survive.

The tricky part is defining “normal” without cheating. “Normal conditions” can’t just mean “conditions where the law holds”—that would be circular. But if you define “normal” statistically, you run into problems: what if most actual birds have had their wings clipped by scientists? Are those birds “abnormal”? They’re certainly statistically common. So “normal” must mean something more like “the conditions the system evolved to function in,” which is a much more complicated idea.

So What’s the Answer?

Nobody has fully solved this puzzle. Different philosophers favor different approaches, and each has strengths and weaknesses. What’s interesting is that this isn’t just an abstract philosophical game. Real science depends on these “fine print” laws. When a doctor says “this medicine will cure your infection, ceteris paribus,” they mean it will work assuming nothing unusual is going on with your body. When an economist says “raising the minimum wage will reduce employment, ceteris paribus,” they’re making a claim about what usually happens, not what always happens.

The challenge is: how do we talk about what usually happens with the same precision and confidence we use for what always happens? That’s what philosophers are still arguing about.

Key Terms

Term	What it does in this debate
Ceteris paribus	Latin for “other things being equal”—the fine print that says a law holds only when nothing interferes
Universal law	A law with no exceptions, like fundamental physics is supposed to have
Counterfactual	A claim about what would happen in a situation that didn’t actually occur
Completer	An explanation for why an exception happened, which “completes” the law by accounting for the interference
Stability/Invariance	A property of laws: how well they hold up when you imagine changing circumstances
Tendency/Capacity	A hidden power or disposition that an object has, which may not always be visible on the surface
Normic law	A law that describes what normally happens, rather than what always happens

Key People

John Stuart Mill (19th-century British philosopher) — Argued that what look like exceptions are actually just other laws working at the same time; laws describe tendencies
Jerry Fodor (20th-century American philosopher) — Argued that special sciences like psychology have real laws, just with ceteris paribus clauses that could be “completed” by explanations from lower-level sciences
James Woodward (contemporary American philosopher) — Developed the “invariance” approach: a law is anything that stays true under enough interventions that matter for a given science
Marc Lange (contemporary American philosopher) — Argued that laws are defined by their “stability” under counterfactual suppositions; special science laws are stable for a narrower range of suppositions
Nancy Cartwright (contemporary British philosopher) — Argued that even physics laws are really ceteris paribus and that understanding them requires talking about capacities and tendencies

Things to Think About

Think of a rule you follow at school—“if you raise your hand, the teacher will call on you.” This has exceptions (the teacher might not see you, might be busy, etc.). Does this make it not a real rule? What would it take to turn this into a strict, universal rule?
The “normal conditions” approach says laws describe what happens normally. But whose normal? A bird in a zoo with clipped wings is normal for that zoo, but abnormal for wild birds. If a law says “birds fly,” which perspective matters?
If all laws turn out to be ceteris paribus—even the laws of physics—does that mean nothing is ever really certain? Or does it mean that science works fine without certainty?
A friend says to you: “I always win at this game, ceteris paribus.” You play, and they lose. They say: “That didn’t count because I was distracted.” Are they allowed to keep adding excuses? When does a ceteris paribus clause stop being useful fine print and become just a way of never being wrong?

Where This Shows Up

Everyday arguments — People constantly say things like “normally, that works” or “except when…” The ceteris paribus problem is the problem of how to make those statements meaningful
Medicine — Drug trials test whether a medicine works “on average” or “for most people.” Side effects and individual differences are the “disturbing factors” in ceteris paribus medical laws
Sports statistics — “This player scores 40% of the time” is a statistical generalization with exceptions; it doesn’t tell you what will happen on any particular shot
Artificial intelligence — AI systems learn from patterns that have exceptions; how to handle exceptions is one of the hardest problems in machine learning
Law and rules — Laws in the legal sense also have ceteris paribus-like clauses: “No vehicles in the park” seems clear until someone asks about ambulances or skateboards or baby strollers