Proof and Logical Reasoning
What Makes Mathematics Different
In science, the highest status a claim can reach is "well-supported by evidence so far." In mathematics, claims reach something stronger: proved. A proof is not an experiment, a survey, or an argument from authority — it is a sequence of logical deductions, each step following necessarily from the last, leading from accepted starting points (axioms and definitions) to a conclusion. Once verified, a proof settles the matter permanently. Euclid's proof that there are infinitely many primes is as valid today as it was 2,300 years ago, and no future discovery will ever overturn it.
This permanence comes from the nature of deduction itself. Inductive reasoning — "the sun rose every day so far, so it will rise tomorrow" — produces probable conclusions that new observations could revise. Deductive reasoning — "all men are mortal; Socrates is a man; therefore Socrates is mortal" — produces conclusions guaranteed by their premises. If the premises are true and each inference is valid, the conclusion cannot be false without breaking logic itself. Mathematics is built entirely from deduction, which is why its truths hold everywhere, for everyone, forever.
Anatomy of a Proof
A proof typically begins with definitions (what exactly do we mean by "prime," "even," "rational"?) and axioms (statements accepted without proof within the system). The body then chains together small, checkable steps. Each step may be trivial on its own — but their composition can carry you to conclusions that seem miraculous.
Proofs come in several standard forms:
- Direct proof: assume the hypotheses and reason straight through to the conclusion.
- Proof by contradiction (reductio ad absurdum): assume the claim is false and show this assumption leads to an impossibility; since logic cannot permit contradiction, the claim must be true.
- Proof by induction: to prove something about all natural numbers, show it holds for 1 and that if it holds for any number, it holds for the next — like dominoes, where proving the first falls and each knocks down the next topples the whole line.
- Counterexample: to disprove a universal claim, one example where it fails suffices — the fastest kill in mathematics.
Classic Proof One: √2 Is Irrational
The Pythagoreans believed all lengths were ratios of whole numbers. Then someone proved otherwise, and legend says they drowned the discoverer. Here is the argument, in full.
Suppose √2 were rational: √2 = a/b, where a and b are integers with no common factor (the fraction is in lowest terms). Squaring gives a² = 2b², which means a² is even. Now, if a were odd, a² would be odd (odd × odd = odd), so a must be even: write a = 2c. Substituting: 4c² = 2b², hence b² = 2c², which means b² is even, so b is even too.
Both a and b are even — meaning both divisible by 2 — contradicting our choice of a/b in lowest terms. The contradiction came purely from assuming √2 rational. Therefore √2 is irrational. Notice the structure: we never computed anything; we showed the alternative hypothesis self-destructs. This is proof by contradiction at its finest — ten lines that demolished an entire worldview.
Classic Proof Two: Infinitely Many Primes
Euclid's argument appears in his Elements around 300 BC and remains a model of economy. Primes are numbers greater than 1 divisible only by 1 and themselves: 2, 3, 5, 7, 11, ....
Suppose, for contradiction, that there are finitely many primes, and list them all: p₁, p₂, ..., pₙ. Now form the number N = p₁ × p₂ × ... × pₙ + 1 — the product of all primes, plus 1. Divide N by any prime pᵢ on the list: the product part divides evenly, leaving remainder 1. So no prime on the list divides N.
But every number greater than 1 has a prime factor (fundamental theorem of arithmetic). So N has some prime factor — and it cannot be any prime on our supposedly complete list. Contradiction. Therefore no finite list contains all the primes: they are infinite.
The beauty here is the constructive trick: rather than hunting for large primes directly, we manufacture a number that forces new primes into existence whenever anyone claims to have them all. (It also shows why "multiply all known primes and add 1" generates numbers whose factors include primes possibly beyond those known — Mersenne-style prime hunts descend from this idea.)
Proofs, Machines, and Limits
Proofs are so mechanical in principle that mathematicians long dreamed of automating them. That dream produced both triumphs and shocks. David Hilbert hoped all of mathematics could be grounded in a complete, consistent axiom system decidable by algorithm. Kurt Gödel destroyed the hope in 1931: in any consistent system rich enough to describe arithmetic, there exist true statements that cannot be proved within it. Completeness and consistency cannot coexist. Alan Turing then showed the same limit computationally — no algorithm can decide, for every statement, whether a proof exists.
These results did not end mathematics; they mapped its horizon. Most questions remain decidable, and proofs remain the gold standard. Today, computers assist both ways: formal verification systems like Lean and Coq check proofs line by line with perfect rigor (a four-color theorem controversy ended when a computer-verified proof appeared), while machine-learning systems have begun discovering proof steps humans missed.
Why It Matters
Proof is more than a technicality — it is a discipline of thought. Learning to prove trains you to state precisely what you mean, distinguish what you know from what you assume, and test ideas against the strictest possible standard. In an era of confident assertion, the ability to ask "how do you know?" and to build an answer link by link until nothing is left unexamined may be mathematics's most transferable gift.