You have stared at the statement for ten minutes. Direct proof set off cleanly but stopped three lines in — you cannot turn the hypothesis into anything useful. So you tried contradiction: assume the negation, hunt for a wall. But the contradiction attempt is also unhappy — you keep generating true statements that do not contradict anything, because you have no specific target to crash into. Both tools have failed to latch onto the problem.
This is the moment to try contrapositive. The signature of this moment is very recognisable once you have seen it, and the three-question check below is built to spot it quickly.
The three-question check
Ask these in order. If all three say "yes," flip to the contrapositive.
- Is the statement an implication P \Rightarrow Q? Contrapositive only works on implications. If the claim is an equality, a set identity, or a forall-without-implies, skip this technique.
- Is the hypothesis P hard to use directly? That is, when you write down P, does algebra refuse to start — no clean variable to manipulate, no structure to decompose, no existence you can invoke?
- Is \lnot Q easy to work with? Negating the conclusion often turns a vague claim ("n is odd") into a concrete one ("n = 2m"). If \lnot Q gives you an equation, a decomposition, or a concrete object, the contrapositive route is paved.
Three yeses is the signal. You then prove the contrapositive "\lnot Q \Rightarrow \lnot P" by direct proof, and you are done.
Why contradiction stalls here but contrapositive succeeds
Both contradiction and contrapositive start from \lnot Q. The difference is what they ask you to produce.
Contradiction is a fishing expedition: you assume \lnot Q and P together, then manipulate until something breaks. You do not know in advance which statement will turn out to be the wall, so you have to try many directions. If no wall appears, you do not know if that means the statement is true (and you are missing the argument) or false (and the "proof" was doomed).
Contrapositive is targeted: you assume \lnot Q and you already know your destination is \lnot P. You are doing direct proof, just from a different starting point. The target is fixed, which means you can plan a route to it.
Why this matters when you are stuck: when direct proof fails because P is vague, contradiction inherits the same vagueness — you are still trying to use P, plus a new assumption \lnot Q, and hoping for a crash. Contrapositive drops P entirely from the working set. You only need to start from \lnot Q and derive \lnot P. That "drops P" move is often the breakthrough, because the whole reason direct proof stalled was that P was unusable.
A worked recognition: "if n^2 is odd, then n is odd"
Run the three-question check on this classic statement.
- Implication? Yes: P is "n^2 is odd," Q is "n is odd."
- Hypothesis hard to use? Yes. Assuming n^2 = 2k + 1, you would need to take a square root and hope the result has a nice form. It does not.
- Negation of conclusion easy? Yes. \lnot Q is "n is even" — a concrete form n = 2m that you can plug into things.
Three yeses. Flip. Prove instead: "if n is even, then n^2 is even." Writing n = 2m, you get n^2 = 4m^2 = 2(2m^2), which is even. One line. Done.
If you had tried contradiction on the original, you would have assumed n^2 is odd and n is even, and tried to crash. But the crash comes from computing n^2 = (2m)^2 = 4m^2, which is even — contradicting the assumed-odd n^2. That argument is actually identical to the contrapositive argument, just wrapped in contradictory clothing. The contrapositive version is cleaner because it does not require you to set up the extra layer of "assume both and look for a wall." You just start at \lnot Q and prove \lnot P.
When to back away from contrapositive
The three-question check is not a universal recommendation. If P itself was easy to use (question 2 says "no"), direct proof is almost always shorter than contrapositive. If \lnot Q is even vaguer than P (question 3 says "no"), contrapositive inherits the vagueness and will stall in the same way. And if the statement is not an implication at all (question 1 fails), contrapositive is the wrong tool and you need existence, contradiction, or construction.
The technique earns its place in the toolkit at the specific intersection where question 1 is yes, question 2 is yes, and question 3 is yes. That is the recognition signature.
The one-line takeaway
When direct proof has no traction on P and contradiction has no wall to find, check whether \lnot Q is concrete. If it is, switch to contrapositive — you now have a clear start and a clear target, which is what the other two were missing.
Related: Proof by Contrapositive · Contrapositive Gives You a Fixed Target — Unlike Contradiction, No Fishing · Signal: Hard-to-Use Hypothesis, Concrete Negated Conclusion — Flip to Contrapositive · Is Contrapositive Really Different from Contradiction, or the Same Trick in Disguise? · Proof by Contradiction