Here is a subtle but common misreading of induction. A student writes:
"The inductive step says: if P(k) is true, then P(k+1) is true. Combined with the base case, this means P(k) is true."
The conclusion is a half-step wrong in a way that matters. The inductive step does not prove P(k). It proves a conditional: if P(k), then P(k+1). That conditional says nothing, by itself, about whether P(k) is true. To conclude P(k) for an arbitrary k, you need the base case plus the chain of conditionals firing one after another — starting from the base case and propagating all the way up to P(k).
This article pins down the distinction.
The shape of the inductive step
Write the inductive step fully explicitly:
This is a universally-quantified conditional. It says: if you ever know P(k), you can conclude P(k+1). It does not say P(k) is actually true. The "if" is doing all the work — the statement is a promise about what follows from knowing P(k), not an assertion that P(k) itself is established.
An everyday analogy: "If it is raining, the streets are wet" does not tell you it is raining. It only tells you what to conclude if you ever learn it is raining. To actually conclude that the streets are wet, you need the additional fact that rain is occurring.
Induction's "additional fact" is the base case: P(1) is true. With P(1) in hand, the inductive step (applied at k = 1) now fires and gives you P(2). With P(2), the inductive step (applied at k = 2) fires and gives you P(3). And so on.
Why the misreading happens
The misreading is seductive because the inductive step is about P(k), and when a student writes the step, they typically say "assume P(k) holds." They then spend a paragraph working with P(k) as if it were a known fact. It is easy to walk away thinking "we showed P(k)."
What actually happened is this: inside the inductive step's proof, you assumed P(k) in order to derive P(k+1). The assumption is not a claim of truth — it is a conditional working hypothesis. The final output of the inductive step is the conditional P(k) \Rightarrow P(k+1), not the claim P(k).
Why this distinction matters in practice: if the inductive step alone could establish P(k), you would not need a base case. The whole reason induction works is the chain: base case provides P(1) as a ground-level fact; the inductive step provides an arrow from each P(j) to P(j+1); applying the arrow k - 1 times gets you to P(k). Remove either ingredient and nothing is proved. The inductive step on its own is a family of conditionals with no ground-level input.
A diagram of the chain
A concrete check: can the inductive step alone prove a false statement?
Yes — and this is the cleanest way to see that the inductive step is not, by itself, a proof of P(k).
Take the claim "n = n + 1 for all positive integers n." The inductive step is:
- Assume k = k + 1.
- Add 1 to both sides: k + 1 = k + 2, which is P(k+1).
The step is valid. The conditional "if P(k) then P(k+1)" is true for every k. But P(k) itself is false for every k — no integer equals its successor. If the inductive step alone established P(k), we would have "proved" a falsehood.
What saves induction from producing this nonsense is the base case. When you try P(1): is 1 = 2? No. The base case fails. Without a valid base case, the inductive step's arrows have no input to transport, and the proof collapses exactly where it should.
This is a rigorous demonstration that the inductive step does not prove P(k) — because the step can be valid on falsehoods, and only the base case's ground-level truth makes the chain produce anything useful.
Restating the inductive step carefully
Here is language that does not invite the misreading:
- Not: "By the inductive step, P(k) is true."
- Yes: "By the base case and repeated application of the inductive step, P(k) is true."
Or more formally:
- Not: "P(k) is true and P(k) \Rightarrow P(k+1), therefore P(k+1)."
- Yes: "P(1) is true, and repeatedly applying the inductive step from P(1) yields P(2), P(3), \dots, P(k), P(k+1)."
The key is naming the mechanism that actually produced P(k): it was not the inductive step alone, it was the base case plus the chain. The step is an arrow. Arrows need an input.
What to say in your own proofs
When you finish an induction proof, the conclusion sentence should look like one of these:
- "By the principle of induction, P(n) is true for every n \geq 1."
- "By induction (base case at n = 1, inductive step for k \geq 1), P(n) holds for all positive n."
Notice neither sentence says "by the inductive step, P(n) is true." The phrase is "by induction" — meaning the whole machinery, base case plus step plus principle — and not "by the step," which would be the misreading.
The one-line takeaway
The inductive step is a conditional: it transports the truth of P(k) into the truth of P(k+1). It does not produce the truth of P(k) out of nothing. The base case produces P(1); the inductive step then propagates to P(2), P(3), \dots The two together prove P(k) — the step alone cannot.
Related: Mathematical Induction · Missing Base Case — The Domino Chain That Never Starts · Isn't Induction Circular? You're Assuming What You Want to Prove · Induction Ladder — Climb From P(k) to P(k+1), One Rung at a Time · Missing Inductive Step — The Domino Chain With a Gap at Tile 7