What is a paradox? Famous paradoxes and what they teach

A paradox is an argument that starts from premises that seem true, moves by steps that seem valid, and arrives at a conclusion that seems false, or contradicts itself. Since true premises and valid steps can't lead to a falsehood, something has to give: a premise is wrong, a step is flawed, or the conclusion is true and intuition is what's mistaken. Working out which is how paradoxes have pushed mathematics and logic forward.

In short

• A paradox is a valid-looking argument from believable premises to an unbelievable conclusion. Resolving it means finding what gives.
• Some paradoxes are surprising truths (the birthday paradox), some hide an error (Zeno's race), and some expose a contradiction in ideas we rely on (the liar).
• The hardest ones test whole fields: Zeno's are answered by the mathematics of infinite sums, and Russell's forced a rebuilding of set theory.

What counts as a paradox?

In everyday speech, "paradox" often just means irony or an odd contrast: "it's a paradox that the busiest people get the most done." In logic and philosophy it means something sharper. A paradox is a piece of reasoning in which each part looks acceptable and the whole looks impossible. The rules of logic say that valid steps from true premises can't reach a false conclusion, so a paradox is a sign that something we accepted shouldn't have been.

The philosopher W. V. O. Quine, writing in 1962, sorted paradoxes by what turns out to give, as the overview of paradoxes describes:

• A veridical paradox has a conclusion that seems absurd but is true. The mistake is in our intuition.
• A falsidical paradox has a conclusion that seems false and is false, because the argument hides an error.
• An antinomy reaches a contradiction by reasoning we normally trust, and so shows that a basic idea needs repair.

Paradoxes differ from questions settled by evidence. No experiment resolves one; the work is in the reasoning, which is why so many are posed as thought experiments.

What are the famous paradoxes, and are they solved?

Paradox	The argument in brief	Kind	Resolved?
The birthday paradox	In a room of just 23 people, two probably share a birthday	Veridical	Yes: it's true, and the arithmetic shows why
Achilles and the tortoise (Zeno, 5th century BCE)	Achilles must reach where the tortoise was, by which time it has moved on, forever, so he never catches it	Falsidical	Mostly: infinitely many stages can add up to a finite distance and time
The arrow (Zeno)	At each instant a flying arrow is still, and time is made of instants, so it never moves	Falsidical	Mostly: speed is defined over intervals, not single instants; some details still argued
The liar	"This sentence is false." If it's true, it's false; if it's false, it's true	Antinomy	No agreed solution
Russell's paradox (1901)	The set of all sets that don't contain themselves contains itself if and only if it doesn't	Antinomy	In mathematics, yes, by limiting which sets may be formed

Some famous puzzles called paradoxes are different again. The Fermi paradox is a clash between an estimate and an observation, not a logical contradiction. The ship of Theseus is a puzzle about what makes something the same thing over time.

Why do 23 people probably share a birthday?

The birthday paradox is veridical: the conclusion sounds wrong and is right. Assume 365 equally likely birthdays, ignoring 29 February and twins. The trick is to work out the chance that no one shares, then subtract from 1.

1. The first person can have any birthday: 365 ÷ 365.
2. The second must avoid the first person's day: 364 ÷ 365.
3. The third must avoid two days: 363 ÷ 365.
4. Continue to the 23rd person, who must avoid 22 days: 343 ÷ 365.
5. Multiply them all: (365 × 364 × … × 343) ÷ 365²³ ≈ 0.4927. That is the chance that all 23 birthdays differ.
6. Subtract from 1: 1 − 0.4927 ≈ 0.5073, a 50.7% chance that at least two people share a birthday.

People in the room	Chance at least two share a birthday
10	11.7%
23	50.7%
30	70.6%
50	97.0%
70	99.9%

Intuition fails because it pictures the wrong question. Most people imagine someone matching their birthday, and for that the chance with 22 others is under 6%; you would need 253 other people to reach 50%. But a match can be between any two people, and 23 people make 23 × 22 ÷ 2 = 253 pairs. The surprise comes from counting people when what matters is pairs. Real birthdays aren't spread perfectly evenly, which only raises the chance of a match. The same arithmetic of probability lies behind the Monty Hall problem, another veridical paradox.

How can Achilles catch the tortoise?

Zeno of Elea, a Greek philosopher born around 490 BCE, set out several paradoxes of motion, preserved mainly by Aristotle. Give the tortoise a 100-meter head start and let Achilles run 10 times as fast, at 10 meters a second. By the time he covers the 100 meters, in 10 seconds, the tortoise has gone 10 meters further. He covers those in 1 second; the tortoise has gone 1 meter. And so on, without end.

Add up the distances: 100 + 10 + 1 + 0.1 + … = 111.1 meters, repeating. Add up the times: 10 + 1 + 0.1 + … = 11.1 seconds. Infinitely many stages, but a finite total: Achilles draws level after 11.1 seconds, at 111.1 meters, just as simple division says he should, 100 meters closed at 9 meters a second. The Stanford Encyclopedia of Philosophy calls this mathematics, made rigorous by Augustin-Louis Cauchy in the 19th century, the received view. It notes that philosophers still argue over parts of it, such as whether completing infinitely many tasks is coherent, the questions that make infinity hard.

Why do the liar and Russell's paradox matter?

They show contradictions arising from rules almost everyone accepts: that every sentence is true or false, and that any clear description picks out a collection.

The liar, "this sentence is false", was discussed by Greek philosophers of the Megarian school. Assume it's true: then what it says holds, so it's false. Assume it's false: then what it says fails, so it's true. Either way, contradiction. In 1901 Bertrand Russell found the same trap in mathematics. Some sets contain themselves (the set of all sets with more than three members has more than three members) and most don't. Now take R, the set of all sets that don't contain themselves. If R contains itself, it shouldn't; if it doesn't, it should.

Russell wrote to the logician Gottlob Frege on 16 June 1902, as Frege's book founding arithmetic on logic was going to press. Frege added a hastily written appendix conceding that the paradox undermined one of his basic laws, according to the Stanford Encyclopedia of Philosophy. Mathematicians repaired set theory by restricting how sets may be formed: in the system most use today, you can only carve a set out of a set that already exists, so R can't be built.

Still argued

The liar has no agreed solution. Alfred Tarski argued that a language can't consistently talk about its own truth, so "true" must be defined in a higher-level language. Saul Kripke proposed that sentences like the liar are neither true nor false. Graham Priest argues for dialetheism, the view that some contradictions are true and logic should be able to live with them. The Stanford Encyclopedia's entry on the liar calls how well these approaches succeed "a highly contentious issue".

Check yourself

• Without looking back, explain the difference between a veridical paradox, a falsidical one and an antinomy, with an example of each.
• Why does the birthday paradox need only 23 people, when matching your own birthday needs far more?
• How can infinitely many stages of Achilles' race take a finite time?

Sources

• Paradox — Wikipedia
• Zeno's Paradoxes — Nick Huggett, Stanford Encyclopedia of Philosophy, revised 2024
• Russell's Paradox — Stanford Encyclopedia of Philosophy, revised 2026
• Liar Paradox — Stanford Encyclopedia of Philosophy, revised 2026
• Birthday problem — Wikipedia