2.5.4 · D5Number Theory (Intermediate)

Question bank — Irrational numbers — √2, π, e — proof that √2 is irrational

1,936 words9 min readBack to topic

Five pictures that anchor every trap below

Before the quick-fire drill, lock in the five images the questions keep returning to. Each figure IS the reasoning — the words only point at it.

1 · What "terminating" vs "repeating" actually looks like. A rational number's decimal either stops (terminating) or locks into a loop forever (repeating). Irrationals do neither — they wander with no cycle. The repeat is not a coincidence: long division of has only possible remainders, so a remainder must eventually repeat, and once it does, the digits loop.

Look at the green block: the moment the same remainder returns, the whole tail must copy itself — that forced loop is why every fraction repeats.

2 · Why we demand lowest terms (). A fraction like hides a shared factor of 2. "Lowest terms" means we have already cancelled every common factor, so and share nothing. That is the tripwire: if the algebra later forces a shared factor of 2, we have contradicted a cancellation we already performed.

The red crossings show common factors being removed; the final coprime pair on the right is what the proof starts from — so "both even" at the end is a factor that should not exist.

3 · Why squaring turns a root into an integer relation. has a stubborn root sign. Squaring both sides sends and , then clearing the denominator lands on the clean integer equation — the doorway to parity arguments.

4 · Parity as pairing squares. "Even" means the dots pair up with none left over. A square of dots for an odd number always leaves a lone diagonal dot; a square for an even number pairs perfectly. This is the visual meaning of " even even" — and the Pythagorean unit-square diagonal is where comes from in the first place, via Pythagoras.

5 · Dense vs complete on the number line. Rationals crowd arbitrarily close to from both sides (density), yet the exact point is a hole in the rationals. Filling every such hole is what "complete" means.

The blue rationals swarm the yellow target but never land on it — that gap is exactly the irrational number.


True or false — justify

A rational number always has a terminating decimal
False — it may repeat forever (like ); the rule (figure 1) is that a rational has a terminating OR repeating decimal, while irrationals do neither.
Every non-terminating decimal is irrational
False — never terminates yet equals ; only a non-terminating and non-repeating decimal is irrational.
is irrational because a calculator shows no repeating pattern
False — a calculator only shows finitely many digits, which can never prove non-repetition; the contradiction proof is what actually establishes it.
The proof would still work if we forgot to reduce to lowest terms
False — the whole contradiction is " both even yet " (figure 2); without assuming lowest terms there is no GCD fact to contradict.
The proof shows √2 equals no fraction of the specific form even/even
False — it starts from an arbitrary lowest-terms fraction and derives that both must be even, so it rules out all fractions, not a special subclass.
is transcendental
False — it satisfies , a polynomial with integer coefficients, so it is algebraic (irrational but not transcendental).
Between and there are no rational numbers
False — rationals are dense on the real line (figure 5); infinitely many fractions lie in that tiny gap, but none of them equals .
If is even then is even
True — squaring preserves parity (figure 4): an odd gives , which is odd, so an even square forces an even root.
If is divisible by 3 then is divisible by 3
True — this is the same prime argument as parity but with the prime 3; it is exactly why the identical proof shows is irrational.
Because rationals can approximate √2 to any accuracy, √2 is "basically rational"
False — approximation is not equality; , and no fraction squares to exactly 2.

Spot the error

"Assume . Then , so is even, done — contradiction."
Incomplete, not wrong: showing even alone is no contradiction; you must push on to show is also even, which then breaks .
" means is even, therefore is odd."
Nonsense — is a multiple of 2, hence even; both sides of an equation share the same parity, so they cannot disagree.
"Since is even, write ; then , so ."
Algebra error: gives , so (not ); dropping the factor of 4 destroys the step that proves even.
"The same proof shows is irrational, since makes even."
Error: leads to , i.e. and — a valid integer, no contradiction; the argument only bites when the number is not a perfect square.
"√2 is irrational, and irrationals aren't real numbers."
Error: irrationals are real numbers; they fill the gaps between rationals on the number line (figure 5) — "irrational" means "not a ratio," not "not real."
", and 2 is rational, so √2 must be rational too."
Error: a product of two irrationals can be rational; that says nothing about whether each factor is rational, just as times itself cancels the irrationality.
"The proof uses Pythagoras, so it only works for right triangles, not for numbers."
Error: Pythagoras is just one motivation (the diagonal in figure 4); the algebraic proof from needs no geometry at all.

Why questions

Why must we assume before starting?
So that " and are both even" becomes a genuine contradiction (figure 2); without the coprime assumption there is nothing for the "both even" conclusion to violate.
Why does the proof need 2 to be prime?
Primeness is what lets ""; for a non-prime like 4, a square can be divisible by 4 without its root being — so the cascade breaks.
Why does squaring both sides () actually help?
It removes the square root and turns the problem into the pure integer equation (figure 3), where parity/factorization arguments become available.
Why is "√2 has a non-repeating decimal" equivalent to "√2 is irrational"?
A repeating (or terminating) decimal can always be rewritten as a fraction of integers (figure 1); so non-repeating forever is exactly the failure to be any such ratio.
Why does the same argument fail for ?
is a perfect square, so satisfies the equation with no contradiction; every prime under the root appears an even number of times.
Why can we phrase the contradiction as "infinite descent"?
If worked, both being even lets us halve to a smaller positive fraction , and repeat forever — but positive integers cannot shrink endlessly, which is impossible.
Why is being dense (rationals filling every gap) not the same as being complete?
Density means fractions get arbitrarily close to (figure 5), but the exact point is still a hole in the rationals — completeness is what filling those holes with irrationals achieves.

Edge cases

Family A — sign & trivial values

Is negative-square-root a separate case the proof missed?
No — is rational iff is, so the positive-root proof covers both signs at once.
Does the proof handle (i.e. √2 an integer)?
Yes — is a lowest-terms fraction; the derivation still forces even, but 1 is odd, an immediate contradiction, so no integer equals .
Is a counterexample to "roots are irrational"?
No — is rational; the claim is only that roots of non-perfect-squares are irrational, and is a perfect square.
Does with but slip past the proof?
No — gives , so that case is excluded before the argument even begins.

Family B — structure of the number

Is evidence that irrationals sum to rationals in general?
It shows irrationals can sum to a rational, but only in special cancelling cases; stays irrational, so no general rule.
Is irrational for the same reason as ?
No — is not the root of any integer polynomial (it is transcendental), so the parity/algebraic trick fails; its irrationality needs Lambert's calculus-based proof.

Family C — representation

What if someone uses base-3 instead of base-10 decimals — does √2 become rational?
No — rationality is base-independent; changing base changes the digits but not whether an exact fraction exists.

The whole logic in one schema

Every trap above lives somewhere on this flow. The green path is the honest proof; the red branches are where the misconceptions try to escape and get blocked.

perfect square

stop too early

Assume sqrt2 = p over q in lowest terms

Square both sides

p squared = 2 q squared

p squared is even so p is even

substitute p = 2m

q squared = 2 m squared so q is even

both p and q even

shares factor 2 but gcd was 1

CONTRADICTION so sqrt2 is irrational

no contradiction see sqrt4

not yet a contradiction

Recall One-line summary of every trap

The proof works because 2 is prime (figure 4) and is coprime (figure 2); it fails for perfect squares because the prime under the root appears an even number of times; and approximation, density (figure 5), or a partial "both even" is never the same as an actual contradiction.