Open any proofs textbook and you will notice the same ritual opening line in almost every direct proof: "Let n be an arbitrary integer." Not "let n be an integer." Not "take any integer n." Not "suppose n = 4." The specific word arbitrary shows up so reliably that it feels like a magic incantation. A natural question: "Why that word? What does 'arbitrary' add that the sentence does not already say?"
It adds a promise — and that promise controls what you are allowed to do for the rest of the proof.
What "arbitrary" actually means
In everyday English, arbitrary means random or unjustified — "an arbitrary decision" means a decision without a good reason. In mathematics, the word has a precise, almost opposite meaning:
Arbitrary (in mathematics)
An object is arbitrary if it was chosen without any special property, so the argument must work regardless of which specific object you chose.
When you write "let n be an arbitrary integer," you are promising the reader: "I will not rely on anything special about this n. Every step I take will be justifiable for any integer you could have picked in its place."
That promise is load-bearing. It is the reason a proof about a single variable n can cover infinitely many integers.
What the word rules out
Say the word arbitrary disappears from the opening line: "let n be an integer." Now imagine a student reads this and silently assumes n = 2 — because 2 is a nice number and the algebra will be easier. Five lines in, they have computed "n^2 + n = 6," and their "proof" hinges on the fact that 6 is divisible by 2. The argument only works for n = 2, but the claim was about every integer.
Inserting arbitrary is what stops this. It signals: no shortcut, no convenient choice. You cannot silently pick n = 2; the letter n stays as a placeholder for every integer at once. Every line of algebra must survive replacement of n by any integer whatsoever — -17, 0, 1 000 000, or the fussy ones the grader will pick when checking your work.
Why "arbitrary" is stronger than "any": in everyday English, any is ambiguous. "Pick any integer" could mean "pick some integer of your choice" (existential) or "for every integer" (universal). Arbitrary collapses that ambiguity toward the universal reading — any integer, so the proof must work for all of them.
A picture of the promise
A concrete contrast
Without the word, read sloppily:
"Let n be an integer. Then n^2 \geq 0 because n is between -10 and 10."
The justification "because n is between -10 and 10" is a silent assumption that n is small. The proof does not cover n = 15 or n = -1000. The sloppy read slipped in a restriction.
With the word, read carefully:
"Let n be an arbitrary integer. Then n^2 \geq 0 because the square of any integer is non-negative."
Now the justification is a universal fact that does not depend on n's size. The word arbitrary signalled that any justification must survive the full range of integers, not a convenient subrange.
"Arbitrary" is also how you know you are done
When you reach the conclusion, re-read your proof. The test for correctness is: could any step have secretly assumed something special about n? If every step survives the replacement of n by any integer, the proof is complete. If a step would fail for, say, n = 0 or negative n, you have a case you forgot to handle.
Examples of hidden assumptions the word arbitrary helps you notice:
- "Divide both sides by n" — assumes n \neq 0. Not valid for arbitrary integer n.
- "\sqrt{n^2} = n" — assumes n \geq 0. Not valid for arbitrary integer n; you need \sqrt{n^2} = |n|.
- "n/2 is an integer" — assumes n is even. Not valid for arbitrary integer n.
Each hidden assumption is a crack where the proof leaks. The word arbitrary in the opening line is a constant reminder to scan for these cracks before you sign off.
Related words: "any," "some," "for all," "there exists"
The vocabulary of arbitrariness has four siblings worth distinguishing.
- Arbitrary n — you chose n, but with no special property. Equivalent to for all n. Proof must work in every case.
- Any n — ambiguous in English; context decides. In proofs, usually means arbitrary.
- Some n — there exists an n with the stated property. You only need to exhibit one.
- A specific n — you have fixed n to a particular value. Conclusions are about that value only.
When a proof opens "let n be an arbitrary integer", it enters the for-all world. When a proof opens "there exists an integer n such that...", it enters the there-exists world. The two worlds require different techniques. Direct proofs of implications live in the for-all world, and arbitrary is the sign at the entrance.
Why textbooks write the full phrase
Textbooks could just write n \in \mathbb{Z} and call it a day. They spell out "let n be an arbitrary integer" for three reasons.
1. It trains the reader. Students learning proofs need the universal quantifier visible in English before it becomes second nature in symbols. The full phrase is pedagogy.
2. It is a commitment device. Writing arbitrary forces you, the proof-writer, to keep the commitment. Every line you write is implicitly being audited against "would this survive replacement of n by any integer?" The word is a leash on your own shortcuts.
3. It makes the scope clear to the reader. A proof that opens "Let n be an arbitrary integer" and closes "\blacksquare" has unambiguously established a universal claim. A proof that opens "Let n = 5" has established a claim about 5. The opening word fixes the scope of the conclusion.
The short summary
- Arbitrary means no special property assumed — the argument must work for every element of the class.
- The word is not decorative. It is a promise that shapes every subsequent line.
- It rules out silent shortcuts: picking convenient values, assuming sign, assuming non-zero, assuming size.
- It is also the self-check at the end: did any step secretly rely on something special?
- It places the proof firmly in the for-all world, as opposed to the there-exists world.
Next time you write a direct proof, make arbitrary the first word after let. Then read every subsequent line with the question "would this survive any integer?" burned into your eyes. If the answer is yes throughout, you have a proof.
Related: Mathematical Proof — Direct Proof · What Can You Assume When You Assume P? · Why 20 Examples Aren't Enough · Quantifier Flipper