Ma · Mathematics

When Is the Union of Two Subspaces a Subspace?

Published By the 9to5Science editorial team

At a glance

  1. 1. Check containment

    Determine whether every vector in one subspace also belongs to the other.

  2. 2. Use the theorem

    The union is a subspace exactly when one of the two subspaces is contained in the other.

  3. 3. Test a failure

    If neither contains the other, choose one vector from each difference; their sum lies in neither subspace.

The union of two subspaces collects every vector that belongs to at least one of them. Unlike their intersection or their sum, that union is not automatically a subspace. The key test is whether one subspace is already contained in the other.

The criterion

Let UU and WW be subspaces of the same vector space VV. Then

U∪W is a subspace of V⟺U⊆W or W⊆U.U \cup W \text{ is a subspace of } V \quad\Longleftrightarrow\quad U \subseteq W \text{ or } W \subseteq U.

If one is contained in the other, their union is simply the larger subspace. If neither is contained in the other, the union fails the subspace test: it is not closed under addition.

Why containment is enough

Suppose U⊆WU \subseteq W. Every vector in UU is already in WW, so adding UU to WW contributes no new vectors:

U∪W=W.U \cup W = W.

Since WW is a subspace, the union is a subspace. The same argument applies if W⊆UW \subseteq U; then U∪W=UU \cup W=U.

Why neither containment makes the union fail

Suppose neither subspace contains the other. Then there are vectors

u∈U∖W,w∈W∖U.u \in U \setminus W, \qquad w \in W \setminus U.

Consider their sum u+wu+w. It cannot belong to UU: if it did, then w=(u+w)−uw=(u+w)-u would also belong to UU, because uu and u+wu+w are in UU and a subspace is closed under subtraction. That contradicts w∉Uw \notin U.

Likewise, u+wu+w cannot belong to WW. If it did, then u=(u+w)−wu=(u+w)-w would belong to WW, contradicting u∉Wu \notin W. Thus u+wu+w is in neither subspace, so it is not in U∪WU \cup W.

The union contains both uu and ww but not their sum. Therefore it is not closed under addition and cannot be a subspace.

Worked example: the coordinate axes

In R2\mathbb{R}^2, let

U={(x,0):x∈R},W={(0,y):y∈R}.U=\{(x,0):x\in\mathbb{R}\}, \qquad W=\{(0,y):y\in\mathbb{R}\}.

Here UU is the horizontal axis and WW is the vertical axis. Neither contains the other: (1,0)∈U(1,0)\in U but not WW, while (0,1)∈W(0,1)\in W but not UU. Their sum is (1,1)(1,1), which lies on neither axis. So U∪WU\cup W is not a subspace.

For comparison, let UU be the horizontal axis and let W=R2W=\mathbb{R}^2. Since U⊆WU\subseteq W, their union is WW, which is a subspace.

Common mistakes

  • Assuming a union behaves like a sum. The union U∪WU\cup W means “in UU or in WW.” The sum U+WU+W consists of vectors u+wu+w with u∈Uu\in U and w∈Ww\in W. The sum is always a subspace; the union usually is not.
  • Checking only scalar multiplication. A union can be closed under scalar multiplication and still fail to be a subspace because it is not closed under addition. The coordinate axes example shows this: scaling a vector on either axis keeps it on that axis, but adding (1,0)(1,0) and (0,1)(0,1) leaves the union.
  • Thinking the intersection must be trivial. The criterion does not depend on whether U∩WU\cap W is just the zero vector. The question is whether one entire subspace is contained in the other.
  • Comparing sets from different ambient spaces. The theorem concerns subspaces of the same vector space, so the union is being tested as a subset of that common space.

A quick way to decide

To determine whether a particular union is a subspace, first check containment. If U⊆WU\subseteq W or W⊆UW\subseteq U, the union is the larger subspace. Otherwise, find u∈U∖Wu\in U\setminus W and w∈W∖Uw\in W\setminus U; the proof above guarantees that u+wu+w is outside the union.

References

  1. Linear Algebra Done Right, Sheldon Axler

Spotted an error? Send a correction. Read how we write and check articles in our editorial policy.