3.4 Precision vs Readability

Mathematical writing must be precise enough to support proof and readable enough to support understanding. These goals are related, but they often pull in different directions.

Precision means that each statement has a definite meaning. The objects are declared, the assumptions are stated, and the conclusion follows under those assumptions.

Readability means that the reader can follow the argument without reconstructing too much hidden context.

A statement can be precise but unreadable.

\forall \varepsilon > 0,\ \exists \delta > 0,\ \forall x \in X,\ d_X(x,a) < \delta \Rightarrow d_Y(f(x), f(a)) < \varepsilon

This is exact, but it may be hard to parse when first introduced. A readable version can precede it:

The function $f$ is continuous at $a$ if points sufficiently close to $a$ are sent to points close to $f(a)$ .

Then the formal condition can follow. The prose gives the idea. The formula fixes the meaning.

Style	Strength	Risk
Fully symbolic	Exact and compact	Hard to read
Fully verbal	Easy to read	May be ambiguous
Mixed	Clear and precise	Requires discipline

Good mathematical writing usually uses a mixed style. It introduces an idea in prose, states it precisely in symbols, then explains how it will be used.

Poor:

Let $f$ be nice. Then the result follows.

Better:

Let $f : X \to Y$ be continuous. Since $X$ is compact and $Y$ is Hausdorff, the image $f(X)$ is compact in $Y$ .

The second version is longer, but it states the assumptions that make the theorem valid.

Precision matters most where a statement can change truth value under different contexts.

Ambiguous phrase	Precise replacement
For all $x$	For all $x \in X$
The function is bounded	$f$ is bounded on $A$
The sequence converges	$(a_n)$ converges in $X$
The map preserves structure	$f(a \cdot b) = f(a) \cdot f(b)$
The object is unique	Unique up to the specified equivalence

Readability matters most where a formal statement hides the main idea. A proof should not be a wall of symbols if the argument has a simple structure.

For example, instead of writing only:

\forall x,y \in V,\ \forall \lambda \in k,\ T(x+y)=T(x)+T(y),\ T(\lambda x)=\lambda T(x)

one can write:

The map $T : V \to W$ is linear: it preserves vector addition and scalar multiplication.

Then the equations can be supplied if needed.

Precision should not mean maximum formality at every line. It means enough formality to remove ambiguity.

Readability should not mean vague language. It means the structure of the argument is visible.

A useful pattern is:

Step	Purpose
State the setting	Declares objects and assumptions
Give the intuition	Explains why the statement should be true
State the formal claim	Fixes exact meaning
Prove in clear steps	Connects assumptions to conclusion
Name dependencies	Shows what was used

For example:

Let $G$ be a group and let $H \leq G$ . We want to know when cosets of $H$ can be multiplied. This works exactly when $H$ is normal. In that case, the quotient $G/H$ inherits a group structure from $G$ .

This paragraph is readable because it gives motivation. It is precise because $G$ , $H$ , normality, and $G/H$ have standard meanings in context.

The main failure mode of precision is overloading the reader with low-level detail too early. The main failure mode of readability is omitting assumptions that affect correctness.

A balanced text separates layers.

Layer	Example
Concept	Continuity means small input changes give small output changes
Definition	For every $\varepsilon > 0$ , there exists $\delta > 0$
Theorem	Continuous functions preserve compactness
Proof	Use an open cover of the image and pull it back
Formalization	Encode topological spaces, maps, covers, and compactness

Not every layer is needed every time. A textbook section may need all of them. A research paper may only need the theorem and proof. A reference page may emphasize definitions.

Symbol density should match the reader’s task. During calculation, dense notation is useful. During explanation, dense notation can obscure the idea.

Compare:

(a+b)^2 = a^2 + 2ab + b^2

This formula is short and readable because the notation is familiar.

But a long formula with many unnamed components should be broken into named parts. Naming reduces cognitive load.

Poor:

\phi(x)=\int_A K(x,y)g(y),dy+\lambda\int_B L(x,z)h(z),dz

Better:

Define

F(x)=\int_A K(x,y)g(y),dy

and

G(x)=\int_B L(x,z)h(z),dz.

Then

\phi(x)=F(x)+\lambda G(x).

The second version is longer, but easier to reason about. It exposes structure.

A similar principle applies to proofs. Long arguments should be decomposed into lemmas when the same idea appears more than once or when a subclaim has independent value.

Situation	Better structure
Repeated calculation	Define a notation
Repeated argument	Extract a lemma
Many assumptions	Group them in a setup
Many cases	Name each case
Long dependency chain	State an intermediate claim

Precision also requires explicit quantifier order. The order of quantifiers changes meaning.

\forall x \in X,\ \exists y \in Y,\ P(x,y)

means each $x$ may have its own $y$ .

\exists y \in Y,\ \forall x \in X,\ P(x,y)

means one $y$ works for every $x$ .

These are different statements. Prose should preserve this distinction.

“Each input has an output” corresponds to the first statement. “There is a single output that works for every input” corresponds to the second.

Readability improves when quantifiers are translated into clear prose, but the symbolic version should remain available when order matters.

Precision also depends on equality and equivalence. Saying that two objects are “the same” is often too vague. The text should say whether it means equal, isomorphic, homeomorphic, equivalent, congruent, or equal up to canonical identification.

Phrase	Better form
$A$ and $B$ are the same	$A = B$
The groups are the same	$G \cong H$
The spaces are the same	$X$ and $Y$ are homeomorphic
The functions are the same locally	They agree on a neighborhood
The objects differ only by coordinates	They are equivalent under change of basis

A readable text may use “same” informally after the precise relation has been stated. Before that, it should avoid the shortcut.

The best rule is local explicitness. Do not restate the entire foundation in every sentence, but state what the reader needs at the point where ambiguity would arise.

Mathematical writing should be exact at the boundaries and readable in the middle. Definitions, theorem statements, and final conclusions need high precision. Motivation and proof sketches can use more prose, provided they do not alter the claim.

Precision protects correctness. Readability protects transmission. A mathematical text fails if either one is missing.