Closure Quiz: Which Operation Escapes Which Set?

Your textbook has a terse line that goes something like: "The naturals are closed under addition and multiplication but not under subtraction." That sentence feels like something to memorise. It is not. It is a description of the reason entire new number systems had to be invented. This satellite turns the closure table into something you can poke at — pick a set, pick an operation, see whether you escape.

Play with the table

Pick a set S on the left axis, and an operation on the top. The cell tells you whether S is closed under that operation — and if not, hands you the specific example that breaks it.

The closure table for $\{+, -, \times, \div\}$ against the four nested number systems. Each red cell was the historical reason the next bigger set was invented: subtraction escaped $\mathbb{N}$ so humans built $\mathbb{Z}$; division escaped $\mathbb{Z}$ so humans built $\mathbb{Q}$; $\mathbb{Q}$ fails one more closure — on roots and limits — which is why $\mathbb{R}$ exists.

Reading each row

\mathbb{N} — the naturals. Two green cells, two red. You can add and multiply counting numbers freely and stay in the counting numbers. But 3 - 5 is -2, which is not a counting number. And 3 \div 2 is 1.5, also not a counting number. So subtraction and division can leave the naturals. Any time a student says "you can't subtract a bigger number from a smaller one," they are really saying "the naturals aren't closed under subtraction" — a perfectly correct statement within \mathbb{N}, and also exactly the complaint that launched the integers.

\mathbb{Z} — the integers. Now subtraction works: 3 - 5 = -2 is an integer. Addition and multiplication still work. The only hold-out is division: 7 \div 4 = 1.75 is not an integer. So \mathbb{Z} plugs one of the two holes in \mathbb{N}.

\mathbb{Q} — the rationals. Division now works too — as long as you don't divide by 0. 7 / 4 is the rational number 7/4, not a problem. The only surviving exception is the single forbidden divisor 0. Every other closure holds.

\mathbb{R} — the reals. Looks the same as \mathbb{Q} in this table. The reals don't fix any closure of \{+, -, \times, \div\}; those were already closed on the rationals. The reason \mathbb{R} exists is to close a different operation — taking limits and roots. \sqrt{2} is not rational but is real. That is a closure failure of \mathbb{Q} under "root of a positive rational," and patching it forces the reals into existence.

A sharper view: the counter-examples

Each red cell in the table is a worked-out escape. Memorising the table is possible but fragile; remembering one counter-example per red cell is much better, because the example proves the failure on its own.

Five counter-examples that tell the whole story

\mathbb{N} under -: 3 - 5 = -2 \notin \mathbb{N}.
\mathbb{N} under \div: 3 \div 2 = 1.5 \notin \mathbb{N}.
\mathbb{Z} under \div: 7 \div 4 = 1.75 \notin \mathbb{Z}.
\mathbb{Q} under \div (by 0): 3 \div 0 — no rational answer.
\mathbb{Q} under "take square root": \sqrt{2} \notin \mathbb{Q} — this is the escape that motivates \mathbb{R}.

Each one is a single expression that takes two inputs from a set and hands you an answer outside that set. That single instance is all it takes to break closure — closure is a "for all pairs" claim, and one bad pair refutes it.

Why the pattern is a ladder, not random

The table has a very specific staircase structure: each larger set fixes exactly the closures that the smaller one broke. This is not coincidence. Humans built the number systems on purpose to fix them. Whenever a closure failed inside some set, the fix was to invent just enough new elements to close it.

\mathbb{N} fails under -. The fix: include additive inverses of every natural. You get the integers \mathbb{Z}.
\mathbb{Z} fails under \div. The fix: include multiplicative inverses of every non-zero integer. You get the rationals \mathbb{Q}.
\mathbb{Q} fails under "root of a positive rational" (and more generally under taking limits of Cauchy sequences). The fix: fill in all the gaps on the number line. You get the reals \mathbb{R}.
\mathbb{R} fails under "square root of a negative number." The fix: include a new element i with i^2 = -1. You get the complex numbers \mathbb{C}.

Same pattern, five times. The closure table you are poking at isn't a list of facts — it is the skeleton of the number system, read from bottom to top.

Does the pattern keep going?

A natural worry: if every closure failure forces a new set, is there any end? What operation could the reals fail at that forces us to go further?

The answer is \sqrt{-1}. The equation x^2 = -1 has no solution in \mathbb{R}, so taking square roots of negatives is the closure failure. The patch is \mathbb{C}, the complex numbers. And after that? It turns out that in \mathbb{C}, every polynomial equation of positive degree has a root (the Fundamental Theorem of Algebra). So \mathbb{C} is algebraically closed, and the ladder of number-system extensions driven by polynomial closures stops there. There are other extensions (quaternions, octonions) driven by different needs, but the main arithmetic staircase ends at the complex numbers.

How to spot closure questions on an exam

A closure question usually reads like: "Is the set of odd integers closed under addition?"

Treat it as a two-line test:

Pick two elements of the set. The worked-out pair is 3 and 5, two odd integers.
Apply the operation. Is the answer in the set? 3 + 5 = 8. Eight is even, not odd. So the set of odd integers is not closed under addition.

If the answer stays in the set for every pair you can dream up, you might have closure — but to prove it, you have to argue from the definitions (here: "odd = not divisible by 2"), not from examples. Exhibiting one escape is enough to disprove closure; proving closure requires a general argument. That asymmetry is what makes closure one of the cleanest types of algebra question on a paper.