Strict or Not? That One Question at the Endpoint Decides Round vs Square Bracket

You finished the algebra. The inequality reduced to something clean — maybe 1 \le x < 7 or -2 < x \le 5 — and now you have to transcribe it into interval notation. This is where careful students lose marks for nothing. They did all the hard work, then wrote (1, 7) when the answer was [1, 7), or [−2, 5) when it was (−2, 5]. A whole problem's credit burned on a bracket character.

The fix is a reflex. At every endpoint, ask one question and nothing else:

Is the inequality sign strict (<, >) or non-strict (\le, \ge) at this endpoint?

Strict means the endpoint is excluded — write a round bracket. Non-strict means the endpoint is included — write a square bracket. That is the entire decision. You do it once for the left endpoint, once for the right endpoint, and you do them independently. The two endpoints do not have to agree, which is the whole reason half-open intervals like [a, b) exist.

The rule in one table

Symbol at endpoint	Endpoint status	Bracket	Picture
< or > (strict)	excluded	round ( or )	hollow circle
\le or \ge (non-strict)	included	square [ or ]	filled circle

That is it. Everything else in this article is drilling the reflex into your hand so that under exam pressure it happens automatically.

Why the two endpoints are independent

An interval has a left end and a right end. Each end came from its own inequality in the original problem. The inequality at the left end has nothing to do with the inequality at the right end — they were produced by different pieces of algebra, or different conditions in a system.

Take the compound condition x \ge 1 and x < 7. You get two separate statements:

• Left end: x \ge 1. The symbol is \ge (non-strict). So 1 is included — square bracket.
• Right end: x < 7. The symbol is < (strict). So 7 is excluded — round bracket.

Put them together: [1, 7). Neither bracket "knows" about the other one. You asked the question twice, independently, and combined the answers.

This is the single most important realisation. Once you stop expecting the brackets to match, you stop trying to "pick one" and start deciding each endpoint on its own merits.

The memory hook: shape matches meaning

If the reflex does not stick from rules alone, use the shape of the bracket as a mnemonic:

• A square bracket [\ ] has a flat wall. The endpoint hits the wall — it is included.
• A round bracket (\ ) curves away. The endpoint slips past — it is excluded.

Now pair this with the symbol:

• \le and \ge have a bar underneath. The bar is the wall. Square.
• < and > are pure arrows. No bar, no wall. Round.

The underline of \le is literally the flat side of [. That small visual match is what makes the reflex sticky: you see the bar, your hand reaches for the wall-bracket. You see no bar, your hand reaches for the curve-bracket.

Worked example: both endpoints, independently

Solve and write the answer in interval notation:

This is a compound inequality — a single expression 2x + 1 is squeezed between two bounds. You solve it the same way as an equation, operating on all three parts at once. Crucially, both inequality symbols travel unchanged through positive operations and the strictness of each is preserved independently.

Step 1. Subtract 1 from every part.

The left symbol stays \le (non-strict). The right symbol stays < (strict). Subtraction does not touch the strictness of either.

Step 2. Divide every part by 2 (positive — no flip).

Same strictness as before: \le on the left, < on the right.

Step 3. Apply the reflex at each endpoint.

• Left endpoint is -2. Symbol: \le. Non-strict, include, square bracket.
• Right endpoint is 2. Symbol: <. Strict, exclude, round bracket.

Answer. [-2,\ 2).

Notice what did not happen: you did not stop to think "do both brackets need to match?" You did not glance at the left bracket to inform the right. Each endpoint was decided by the single symbol sitting next to it in the final inequality. Two independent questions, two independent answers.

Boundary check — the sanity test that catches slips

After you write the interval, test each endpoint against the original inequality.

For [-2, 2) against -3 \le 2x + 1 < 5:

• At x = -2: 2(-2) + 1 = -3. Is -3 \le -3? Yes (equality counts). Is -3 < 5? Yes. Both conditions hold, so -2 is a solution — correctly included as a square bracket.
• At x = 2: 2(2) + 1 = 5. Is -3 \le 5? Yes. Is 5 < 5? No (strict). So 2 is not a solution — correctly excluded as a round bracket.

The brackets match the boundary behaviour. If the test had failed — say, x = 2 had satisfied the inequality but you had written it out with a round bracket — that would be a transcription error, caught in thirty seconds by plugging in.

Do this check in the exam. It is the cheapest insurance policy in algebra.

The three slips this reflex prevents

1. Matching brackets on autopilot. Students who believe intervals must be "open" or "closed" as a whole will write [-2, 2] or (-2, 2) instead of [-2, 2). The reflex of asking two questions breaks that habit.

2. Copying the bracket from the symbol direction. The symbol < points left, which tempts students to write the bracket on the left as (. But < at the right endpoint means the right bracket is round. The direction of the arrow does not choose the side — the endpoint's position does. Strictness chooses the shape.

3. Forgetting the infinity rule. \infty is not a real number and cannot be included, so it always gets a round bracket. Even if you derive x \ge 3 — a non-strict inequality — the interval is [3, \infty): square on 3, round on \infty. The strictness rule applies to the finite endpoint; the round bracket at \infty is unconditional.

One more pattern: the empty middle

Sometimes the reflex flags a bracket that, combined with the other endpoint, makes the interval empty. Consider a solution 3 < x < 3. Left end strict, right end strict. So you write (3, 3) — which is formally a valid open interval, but it is the empty set because no real number is simultaneously greater than 3 and less than 3. The brackets did their job (round-round, because both were strict). What the brackets cannot tell you is that the interval is empty — that is a separate observation about the two endpoint values coinciding.

Similarly, [3, 3] is a valid closed interval — the single point \{3\}. Square brackets, one point, one element. Again, the brackets are correct; the content is what you notice next.

The reflex, restated

At every endpoint of a solved inequality:

1. Look at the symbol next to that endpoint.
2. If it is < or > — round bracket, hollow dot, excluded.
3. If it is \le or \ge — square bracket, filled dot, included.
4. Do this for each endpoint independently.
5. Remember: \infty always gets a round bracket.

That is the whole reflex. Six seconds per interval, and you never lose marks for bracket transcription again. For the deeper geometric reading — why "included" and "excluded" are the natural words here — see [Round Bracket "(" vs Square Bracket "" in Interval Notation. For the exact same distinction stated symbol-first, see < vs ≤ — Strict vs Non-Strict Inequality in Solution Sets.