Flip open a Class 11 textbook and you will see proofs written almost entirely in English: "Since n is even, n = 2k for some integer k. Then..." Flip to a higher-level text and you see the same argument in symbolic shorthand: \forall n \in \mathbb{Z}\ \text{Even}(n) \Rightarrow \exists k \in \mathbb{Z}\ (n = 2k) \Rightarrow \dots Both are proofs. Which one should you write? And how much formality is enough?
The honest answer: a mathematical proof is judged by whether every step is unambiguous and justified — not by the ratio of English words to symbols. You can write a fully valid proof in pure English, in pure symbolic logic, or in a mix, and they are all legitimate as long as the logic is tight.
Three registers, all legal
You can think of proof-writing as happening at three registers.
Register 1 — Pure English.
"Assume n is an even integer. Since n is even, we can write n as two times some integer k. Squaring, we get n squared equals four k squared, which equals two times 2k^2. Since 2k^2 is an integer, n squared is twice an integer, and therefore n squared is even."
Every sentence is a full English sentence. No symbols beyond equations.
Register 2 — Mixed (most common in school).
"Assume n is even. Then n = 2k for some integer k. Then n^2 = (2k)^2 = 4k^2 = 2(2k^2). Since 2k^2 \in \mathbb{Z}, n^2 is even."
English sentences with symbols and equations embedded. This is what textbooks and exam answers use 95% of the time.
Register 3 — Pure symbols.
n \in 2\mathbb{Z} \Rightarrow \exists k \in \mathbb{Z}\ (n = 2k) \Rightarrow n^2 = 4k^2 = 2(2k^2) \Rightarrow \exists m \in \mathbb{Z}\ (n^2 = 2m) \Rightarrow n^2 \in 2\mathbb{Z}.
Almost no English. Everything is logical notation.
All three are valid direct proofs of "if n is even, then n^2 is even." The mathematics is identical; only the surface language changes.
What actually matters: unambiguity and justification
The grader is not counting symbols. They are checking two things on every line.
- Unambiguity. Could a careful reader mis-read this sentence? If yes, fix it.
- Justification. Does the line follow from the previous lines (or from a definition, axiom, or theorem)? If the connection is not clear, make it clear — either with a word like "by definition" or "by the distributive law", or by inserting the missing step.
Any register that meets these two criteria is formal enough. Pure English can be rigorous; pure symbols can be sloppy. The register is a style choice; the rigour is not.
Why unambiguity is the real criterion: a proof is a chain of claims. If a reader can misread any link, the chain breaks. Symbols often remove ambiguity (think of the precision of \forall x \in \mathbb{R} vs "for any real number"), but well-written English can be just as precise — and is usually easier to follow.
The formality dial
Choose the register based on audience and convention
Three factors usually decide where you sit on the dial.
1. The exam or textbook convention. Indian board papers (CBSE, ICSE, state boards) overwhelmingly use the mixed register. That is the default for school proofs and what you should write unless told otherwise. University-level algebra and analysis textbooks sometimes drift toward more symbols; olympiad solutions vary wildly.
2. The complexity of the logical structure. For a simple chain of implications, English reads cleanly and symbols feel like clutter. For a statement with nested quantifiers ("for every \varepsilon > 0 there exists \delta > 0 such that for every x..."), symbols \forall \varepsilon > 0\ \exists \delta > 0\ \forall x \ldots are often easier to parse than the English. Use symbols where they clarify and English where it clarifies.
3. Your reader. A proof written for your class teacher is usually mixed. A proof written for an advanced research journal leans more symbolic. A proof written for a 15-year-old encountering the topic for the first time leans more English. Meet the reader where they are.
What cannot be omitted
Regardless of register, certain elements must appear in a school-level proof.
- A statement of what you are proving — either the full claim or a reference like "Claim: if n is even, then n^2 is even."
- The hypothesis assumption line — "Assume n is even" or "Let n be an arbitrary even integer."
- A justification for every non-obvious step — one-word tags are fine: "by definition," "by algebra," "by the previous theorem."
- The conclusion line — explicitly stating that you reached the claim's conclusion.
- A closing mark — \blacksquare, \square, or QED. This tells the reader the proof is complete.
If any of these is missing, the proof is incomplete at any register.
Registers that are not OK
A few failure modes to avoid.
"Symbol soup" with no sentences. Writing n \Rightarrow n=2k \Rightarrow n^2 = 4k^2 with no commentary is not a proof. The \Rightarrow arrows do not tell the reader what is being assumed, what is being derived, or how. A pure-symbol proof needs at least minimal scaffolding: \forall, \exists, \Rightarrow in the right logical places, and some words explaining the structure.
"Vibes-only English" with no algebra. Writing "Well, obviously if n is even then n squared is also even because evenness carries through squaring" is not a proof either. The English sentence gives no mechanism — no n = 2k, no algebra, no re-packaging. English is fine, but the algebraic content must still be there.
Mixing sloppily. Writing "Assume \forall n is even" is ungrammatical in both English and logic. If you use \forall, use it in a formally correct logical expression. If you use English, write full sentences. Don't smear the two.
Using informal English where precision matters. "n is pretty much even" or "roughly, n equals 2k" are not precise enough. The register can be English, but the English must be precise.
A rule of thumb
"Write the proof the way the textbook writes it."
That is the simplest advice. Look at the worked examples and proofs in your textbook for the chapter you are studying. Match that register. If the textbook uses "Let n be an even integer. Then n = 2k...", you write "Let n be an even integer. Then n = 2k...". The textbook's register is tuned to your reader (the grader) and your exam.
If you are unsure whether your proof is formal enough, apply the rewrite test: give your proof to a friend studying the same chapter. If they can follow every line without asking a question, your register is fine. If they get stuck, it is not the register — it is a missing justification or an ambiguous phrase.
The short summary
- Formality is about clarity and justification, not the symbols/English ratio.
- Three registers are legitimate: pure English, mixed (default), pure symbols.
- The mixed register is the textbook default and what you should use in school exams unless told otherwise.
- Every proof needs a hypothesis line, justified steps, a conclusion, and a closing mark — at any register.
- Symbol soup without sentences and vibes-only English without algebra both fail.
- Match the textbook's register.
Proofs are read, not decoded. Write for clarity, use symbols where they clarify, use English where it clarifies, and don't confuse ornamentation with rigour.
Related: Mathematical Proof — Direct Proof · What Does QED Mean? · Do I Need to State Every Definition? · Proof 'Because' vs 'If' Inside a Proof