"A relation cannot be both symmetric and antisymmetric — they pull in opposite directions."
Plausible, but wrong. A relation can be both. The catch is a sharp restriction: the only such relations are the ones whose pairs all lie on the diagonal. In other words, R \subseteq \{(a, a) \mid a \in A\}.
The canonical example is equality. The empty relation is another. Anything with even one off-diagonal pair automatically fails one of the two properties.
Chasing the contradiction down
Suppose R is both symmetric and antisymmetric. Take any pair (a, b) \in R.
- Symmetric gives (b, a) \in R.
- Antisymmetric, applied to the two pairs (a, b) \in R and (b, a) \in R, forces a = b.
So every pair in R has a = b. That is exactly the statement that R is a subset of the diagonal.
Why: symmetric forces "reverse is in," antisymmetric forces "if reverse is in, then the two elements were equal." Combining them: reverse is in, so the two elements were equal. Every pair in R is of the form (a, a). No off-diagonal pairs survive.
The picture
Concrete examples
1. Equality on any set. R = \{(a, a) \mid a \in A\} — precisely the diagonal. Symmetric: if a = a then a = a. Antisymmetric: if a = b and b = a then a = b. Both trivially hold.
2. The empty relation. R = \varnothing on any set. Symmetric vacuously (no pair to reverse), antisymmetric vacuously (no pair of reverses to force equality). Both hold.
3. Partial diagonal. A = \{1, 2, 3\}, R = \{(1, 1), (3, 3)\}. Only two diagonal cells filled. Symmetric: reversing a diagonal pair gives the same pair. Antisymmetric: the only reversal-pairs are (1, 1) with itself and (3, 3) with itself, and in each case the two coordinates are equal. Both hold.
What breaks when you add any off-diagonal pair
Add (1, 2) alone. Now symmetric fails — (2, 1) \notin R. Still antisymmetric.
Add (1, 2) and (2, 1) both. Now symmetric succeeds. But antisymmetric fails — both pairs are in R and 1 \neq 2. So symmetric and antisymmetric are mutually exclusive the moment a non-diagonal pair enters.
This is the whole content of the theorem: you have to choose. You can have symmetric without off-diagonal pairs. You can have antisymmetric with off-diagonal pairs as long as their reverses are out. To have both at once, keep R strictly on the diagonal.
Interactive: watch the edge case
Why this matters for JEE
Examiners set questions like "is the equality relation on \mathbb{R} an equivalence relation and also antisymmetric?" The answer is yes to both, and the reasoning hinges on the diagonal-only structure of equality. Students who memorise "symmetric and antisymmetric are opposites" give the wrong answer. Students who remember the characterisation — both hold iff R lives inside the diagonal — get full marks.
Equality is the trivial equivalence relation *and* a partial order
Take A any set and R = \{(a, a) \mid a \in A\}.
- Reflexive: every diagonal pair is in, by definition.
- Symmetric: reversing (a, a) gives (a, a).
- Transitive: the only chains are (a, a), (a, a) \to (a, a). All in.
- Antisymmetric: if (a, b) \in R and (b, a) \in R, then the pairs are (a, b) = (a, a) and (b, a) = (b, b), which forces a = b.
So equality is simultaneously an equivalence relation (reflexive, symmetric, transitive) and a partial order (reflexive, antisymmetric, transitive). This is the only way both classifications happen together, because the only relation that is both symmetric and antisymmetric on A is one whose off-diagonal content is empty.
One-line take-away
A relation is both symmetric and antisymmetric exactly when R \subseteq \{(a, a) \mid a \in A\} — every pair sits on the diagonal. Equality and the empty relation are the headline examples. Anything else forces you to pick one.
Related: Relations · Antisymmetric vs Not Symmetric · Equivalence Relations · Why 'Equals' Is the Trivial Equivalence Relation