Induction Ladder — Climb From P(k) to P(k+1), One Rung at a Time

Dominoes are the standard metaphor for induction, but they hide one thing: the fact that the inductive step is an action you perform, not just a structure that exists. A ladder makes this action visible. Every rung is a statement P(k) for some positive integer k. Climbing from one rung to the next is exactly what the inductive step does — it is a move, executed by the proof, that transports the truth upward.

The visualisation below lets you climb a ladder rung by rung. Each click advances you from P(k) to P(k+1). There is no animation magic — you are performing the inductive step yourself, once per click, and the readout tracks how high you have climbed.

The climbing visualisation

Rung k:

A twelve-rung ladder. The bottom rung is the base case: you verify $P(1)$ directly. Every rung above is reached by executing the inductive step once — turning known $P(k)$ into newly-known $P(k+1)$. The climber cannot skip rungs; the step is atomic. Drag the slider or press "Climb one rung" to see how many steps you have applied.

Why "one rung at a time" is the right picture

An induction proof looks, from the outside, like it proves infinitely many things at once. That framing makes it seem mysterious — how can a finite argument establish infinitely many facts? The ladder resolves the mystery by reframing the work.

You do not prove all P(n) simultaneously. You prove one generic rung-climb — an argument of the form "from P(k), derive P(k+1)" — and you prove one specific base case — P(1) directly. Then the chain of rung-climbs transports you upward from the base case, one rung per step. Every specific P(n) is reached by a finite number of rung-climbs: n - 1 of them, to be precise, starting from P(1).

Why the proof is finite even though the claim is about all positive integers: for any specific n you care about, the proof of P(n) is finite and explicit. It is the base case plus n - 1 applications of the inductive step. The inductive step is proved once in a generic way, but applied as many times as needed to reach your target n. So the argument is reusable infrastructure, and the number of applications for a given n is finite — even though you can apply it as many times as you like to reach higher and higher n.

What the rung-climbing action actually is

Each click from rung k to rung k+1 is one execution of the inductive step. The step itself is an argument that looks like this:

Assume: P(k) is true (this is what you just established on rung k). Derive: P(k+1) is true (this is the rung you arrive at).

The content of the argument depends on what P(n) says. For example, if P(n) is the claim "1 + 2 + \dots + n = n(n+1)/2," then the inductive step is:

Assume 1 + 2 + \dots + k = k(k+1)/2.
Add k + 1 to both sides: 1 + 2 + \dots + k + (k+1) = k(k+1)/2 + (k+1).
Factor: k(k+1)/2 + (k+1) = (k+1)(k/2 + 1) = (k+1)(k+2)/2.
Recognise this as the right-hand side of P(k+1): (k+1)(k+2)/2.

This entire block is one rung-climb. You execute it once, and the result is that rung k+1 is now known. Repeat, substituting k+1 for k, to climb to rung k+2. And so on indefinitely.

The ladder distinguishes "can reach" from "has reached"

A subtle point the ladder makes sharper than dominoes: induction proves that every rung can be reached. It does not, in the physical sense, reach them all at once. What it provides is a guaranteed route from the base case to any chosen rung.

Suppose you are asked: does P(500) hold? The induction argument answers: yes, and here is how to verify it. Start at P(1) (which you proved directly). Apply the inductive step 499 times. The 499-th application lands you at P(500). The argument that P(500) holds is finite and explicit; you never need to physically execute those 499 steps, because the inductive step being valid for every k means the route is guaranteed to work.

This distinction — between the route existing and the route being traversed — is what makes induction efficient. You prove the step once, and the proof that P(500) holds is "apply the step 499 times from the base case," without needing to actually write out 499 applications.

What happens if a rung is broken

The ladder metaphor also makes failure modes visible. Suppose the inductive step fails at k = 7 — meaning the argument "from P(7), derive P(8)" does not go through. On the ladder, rung 8 is broken. You can climb from P(1) to P(7), but the rung you would step onto to reach P(8) is missing. You cannot proceed.

This is exactly the scenario illustrated in the "missing inductive step" variant with a gap in the domino chain, and the behaviour is the same: everything below the break is proved; everything above is unjustified.

The one-line takeaway

Induction is a ladder. The base case is the bottom rung you step onto. The inductive step is a single argument that turns any rung into the one above. You reach higher rungs by repeating that argument, not by having a separate argument for each rung.