Why the √2 Proof Needs 'p and q Share No Common Factor' — What Breaks If You Skip It

Open any textbook proof that \sqrt{2} is irrational and the first line reads something like: "Assume \sqrt{2} = p/q with p and q sharing no common factor other than 1 — that is, the fraction is in lowest terms." A natural student reaction is to suspect this clause is cosmetic. After all, every rational number can be written in lowest terms, so why spell it out? What would go wrong if you left that assumption out?

The answer is that without the lowest-terms clause the entire proof stops dead. There is no contradiction to find. The sentence is not decoration — it is the trap that the argument springs later.

Recap: the proof's shape

The contradiction proof for \sqrt{2} assumes \sqrt{2} = p/q with p, q integers and q \neq 0, squares both sides, and derives:

p^2 = 2q^2

From this you deduce p^2 is even, hence p is even, say p = 2k. Substituting:

4k^2 = 2q^2 \implies q^2 = 2k^2

so q^2 is even, hence q is even. The proof ends by pointing out that both p and q are divisible by 2, which contradicts the lowest-terms assumption.

Why the last line is the engine: everything before it is just algebra — divisibility facts about p and q. The contradiction is the collision between "both divisible by 2" and "lowest terms." Take away the lowest-terms setup and the collision has nothing to collide with.

What "without lowest terms" actually gives you

Suppose you skip the clause. Start with: assume \sqrt{2} = p/q for some integers p, q with q \neq 0. Run through the same algebra. You still derive:

p is even.
q is even.

Now pause. Is "p and q are both even" a contradiction by itself? No. It is a perfectly consistent, ordinary fact about two integers. For example, p = 14 and q = 10 are both even — nothing is wrong with them. They do not satisfy \sqrt{2} = 14/10, of course, but the algebraic conclusion "both even" is not itself impossible. The proof has produced a true statement, not an absurdity.

Without the lowest-terms clause, you have no second statement to clash with "both even." The argument terminates with a consistent observation and no contradiction. The proof, as a proof, fails.

Why lowest terms is the essential second premise

Put the two premises side by side:

(From the assumption) p and q share no common factor other than 1. (Lowest terms.)
(From the algebra) p and q are both divisible by 2.

These two statements cannot both be true. If a common factor of 2 exists, then 1 is not the only common factor. Premise (1) says all common factors are 1; premise (2) exhibits a common factor of 2. That is a logical contradiction of the form "A and not A" — exactly the kind of impossibility that proof by contradiction is engineered to deliver.

Why this specific pairing: the proof's algebra guarantees a common factor will appear. The lowest-terms clause guarantees no common factor can appear. Those two guarantees are on a collision course the moment the algebra starts — and that collision is what makes the assumption \sqrt{2} = p/q untenable.

Visualising: what the two versions produce

Drag the dot below between without and with the lowest-terms clause. Watch the conclusion change from a harmless observation to a genuine contradiction.

Without the lowest-terms clause the algebra still runs, but its output — "both even" — is a consistent fact, not an impossibility. The lowest-terms clause supplies the incompatible partner. When the two collide, the contradiction snaps into place and $\sqrt{2}$ is forced to be irrational.

An alternative rescue: infinite descent

There is a second way to complete the proof without invoking "lowest terms" explicitly — but it leans on a different (and equally strong) principle.

Suppose you drop the lowest-terms clause and derive that p, q are both even. Write p = 2p_1 and q = 2q_1. Then \sqrt{2} = p_1/q_1 too. Repeat the algebra. Now p_1, q_1 are both even: p_1 = 2p_2, q_1 = 2q_2. Repeat again. Forever.

What you have produced is an infinite descending sequence of positive integers p > p_1 > p_2 > \cdots. But positive integers cannot descend forever — the positive integers have a least element (namely 1), so any descending chain must eventually hit bottom. This principle is called the well-ordering of the positive integers, and it is the sub-principle the lowest-terms trick exploits. Descending forever is impossible; contradiction.

Why this is the same proof in disguise: "lowest terms" is the one-step version of "descent is blocked." If p/q is already reduced to the point where you cannot divide out any more common factors, you cannot descend any further — you have landed on the bottom rung. Dropping the lowest-terms clause just means you have to do the descent step-by-step yourself and invoke well-ordering at the end.

Watching "both even" become harmless without the trap

Take p = 6, q = 4. Check: p is even, q is even. Is this a contradiction? Nope — 6 and 4 are just two even integers. Nothing about them forbids coexistence.

But wait: is \sqrt{2} = 6/4 = 1.5? No; 6/4 = 1.5, and \sqrt{2} \approx 1.414. So the hypothetical assumption \sqrt{2} = p/q with p = 6, q = 4 is actually false. The algebra derived "both even" from an assumption that was never satisfied — and without the lowest-terms clause, there is no way for the proof to detect that the assumption failed.

Adding the lowest-terms clause forces the only remaining candidates to be ones where "both even" is impossible — and exactly those candidates are where the contradiction lands.

The general pattern

This is a widespread pattern in contradiction proofs: the assumption you negate is often accompanied by a second, seemingly minor, clause whose job is to create the collision. Without the second clause, the negation by itself produces a consistent hypothesis, and the contradiction never materialises.

Other examples where a secondary clause does this work:

Infinitely many primes (Euclid): you assume \{p_1, \ldots, p_n\} is the complete list. Without "complete," the algebra just finds a new prime and that is a non-event; with "complete," the new prime contradicts the premise.
No rational square root of 3: same structure as \sqrt{2} — lowest terms plus derived divisibility by 3.
Uniqueness proofs: assume two distinct objects satisfy the property, then show they must be equal. The word distinct is the trap — without it, proving equality is just a fact.

The short version

Without "lowest terms," the algebra concludes "p and q are both even" — a consistent observation, not an impossibility.
The "lowest terms" clause says "p and q share no common factor" — which "both even" violates.
The collision between those two is the contradiction. Remove the clause, and the collision disappears.
Alternatives like infinite descent reach the same contradiction by invoking well-ordering instead, but they implicitly carry the same idea: descent must terminate somewhere.

The lowest-terms clause is not a decoration. It is the carefully placed trip-wire that the algebra walks into. Without it, the algebra runs, says something true, and leaves the original assumption untouched. With it, the same algebra delivers a proof.