'Does a Root Exist?' and 'Does a Fixed Point Exist?' Are Completeness Questions in Disguise

Some problems ask you to compute. Some ask you to simplify. And then there is a third kind — the kind that asks you to prove something exists at all. "Show x^3 = x + 1 has a real solution." "Show this continuous map from [0, 1] to itself has a fixed point." "Show every positive real has a square root." These feel less like algebra and more like promises that need verifying. And once you recognise the shape of the promise, you almost always reach for the same drawer: completeness.

This page is the recognition drill. When a question asks for a witness — a point where a condition holds — the proof strategy is rarely about solving. It is about trapping the point between two conditions and invoking a theorem that says the trap must catch something real. Every such theorem, when you dig down, runs on completeness.

The shape of an existence question

Existence questions tend to look like: does f(x) = 0 have a solution on [a, b]? Does there exist x with g(x) = x? Do two graphs meet somewhere in [a, b]? Does a continuous function on [a, b] attain its max? Does \sqrt{c} exist as a real number?

None ask you to compute a number. They ask you to confirm a witness lives on the number line — not in a hole. Only a complete number system guarantees that.

The master tool: the Intermediate Value Theorem

The workhorse for root-existence on the real line is the Intermediate Value Theorem (IVT):

Intermediate Value Theorem (IVT)

If f is continuous on [a, b] and f(a) and f(b) have opposite signs, then there exists some c \in (a, b) with f(c) = 0.

In words: a continuous curve that starts below the x-axis and ends above it — or vice versa — must cross the axis somewhere in between. You cannot jump over a line if you are required to draw your path without lifting the pen.

The IVT picture. The curve enters the slab $[a, b]$ below the axis at $f(a)$ and leaves above at $f(b)$. Continuity means no jumps, and completeness of $\mathbb{R}$ means the crossing point $c$ is actually *on* the number line — not hiding in a gap.

The word "continuous" does work. But the real engine, one layer below, is completeness. IVT on \mathbb{R} is proved by taking c = \sup\{x \in [a, b] : f(x) \le 0\} and showing f(c) = 0; the supremum exists because \mathbb{R} is complete. Strip completeness away and the theorem collapses.

Why IVT fails on \mathbb{Q}

Consider f(x) = x^2 - 2 on [1, 2] \cap \mathbb{Q}. Then f(1) = -1 < 0 and f(2) = 2 > 0. The function is continuous. Everything looks set up for IVT.

But f has no rational root. Its only zero is \sqrt{2}, which is irrational. The set \{x \in \mathbb{Q} \cap [1, 2] : x^2 \le 2\} has no rational supremum — the only candidate is \sqrt{2}, and that candidate sits in a hole. IVT would produce a crossing if \mathbb{Q} were complete; it is not, so the theorem does not apply. See What Completeness Means.

Worked example: x^3 = x + 1 has a real root

Claim: the equation x^3 = x + 1 has a solution in \mathbb{R}.

Step 1 — rephrase as a root problem. Move everything to one side: let f(x) = x^3 - x - 1. A solution of the original equation is exactly a root of f.

Step 2 — check continuity. f is a polynomial, so it is continuous on all of \mathbb{R}.

Step 3 — find a sign change. Compute at two convenient points:

f(1) = 1 - 1 - 1 = -1 < 0.
f(2) = 8 - 2 - 1 = 5 > 0.

Step 4 — apply IVT. Since f is continuous on [1, 2] and f(1) < 0 < f(2), there exists c \in (1, 2) with f(c) = 0. That c is the real root. Done.

Notice what the proof does not do. It does not compute c, does not bound it more tightly than "between 1 and 2", does not even tell you whether c is rational — though the rational root theorem quickly shows it is not. Existence is cheaper than construction, and IVT cashes in that cheapness.

The same story over \mathbb{Q}

Run the same steps with f(x) = x^3 - x - 1 but insist on rational answers. The sign check still gives f(1) = -1 and f(2) = 5. Step 4 now asks for a rational c with f(c) = 0. By the rational root theorem, the only candidates are \pm 1, and neither works. So there is no rational c. IVT relies on completeness, and completeness is what you gave up when you shrank to \mathbb{Q}.

Fixed-point theorems are completeness in disguise too

A fixed point is just a root in friendlier clothing. If g(x) = x, then f(x) = g(x) - x has a zero; conversely, any root of f gives a fixed point of g = f + x. Every fixed-point theorem is secretly a root-existence theorem, and the same completeness machinery shows up.

Brouwer (1-D form). Every continuous g : [0, 1] \to [0, 1] has a fixed point. Proof: f(x) = g(x) - x satisfies f(0) \ge 0 and f(1) \le 0, so by IVT (hence by completeness) some c gives g(c) = c.

Banach contraction mapping. On a complete metric space, a contraction |g(x) - g(y)| \le k|x - y| with k < 1 has a unique fixed point. The proof iterates g to build a Cauchy sequence and uses completeness to conclude it converges. Drop completeness and the sequence can converge to a ghost point outside the space.

Bolzano–Weierstrass, the extreme value theorem, the mean value theorem — same family. Existence guarantees, all resting on \mathbb{R} being complete.

The three-step reading reflex

Is it an existence claim about a point on \mathbb{R}? A root, a fixed point, a maximum, a crossing. If yes, you are in completeness territory.
Can you wedge the point between two checked conditions? A sign change for IVT, a shrinking interval for Cauchy sequences, a bounded non-empty set for the supremum axiom. The wedge is what you write down.
Is the ambient space complete? \mathbb{R} yes, \mathbb{Q} no. If the problem is secretly over \mathbb{Q}, the approach fails.

Most existence proofs in first-year analysis are three lines once the wedge is set up: state continuity, check two values, invoke the theorem. You do not need to construct the point — you need to corner it. That cornering is what completeness sells you.