12.19.1 Tensor Addition Closure Property

Tensor Addition Closure Property ensures that adding tensors of the same type results in another tensor of the same type.

Tensor Addition Closure Property is the guarantee that adding any two tensors of the same fixed variance type (p,q) on a fixed space V always produces another tensor of that identical type (p,q) on the identical space V, so that the set of all tensors of a given type never needs to be enlarged, altered, or reinterpreted merely to accommodate the result of a sum. It is the most elementary closure fact in tensor algebra, underlying the recognition that tensors of a fixed type form a vector space in their own right, and it is the precondition that must hold before any further vector-space reasoning — bases, dimension, linear independence — can be applied to tensors of that type at all.

Statement and Verification of Closure

The Closure Statement

T, S \in T_{q}^{p} (V) ​ ​ \Rightarrow ​ ​ T + S \in T_{q}^{p} (V)

For tensors T and S, both of type (p,q) on the same space V, their sum T + S — defined by adding corresponding coordinates, or equivalently by adding the multilinear maps pointwise on every tuple of arguments — is again a tensor of type (p,q) on V. No new slot, no change in rank, and no departure from V occurs as a result of the addition.

Verification via the Defining Multilinear Property

(T + S) (a u + b u^{'}, \dots) = a \cdot (T + S) (u, \dots) + b \cdot (T + S) (u^{'}, \dots)

Viewing tensors as multilinear maps, the pointwise sum (T+S)(v₁,...,vₖ) = T(v₁,...,vₖ) + S(v₁,...,vₖ) is itself multilinear whenever T and S individually are, since multilinearity of each argument slot is inherited directly from the multilinearity of T and S in that slot combined with ordinary distributivity of the base field or ring. This confirms that T + S genuinely satisfies the definition of a tensor of type (p,q), not merely resembles one numerically.

Consequences of the Closure Property

Tensors of Fixed Type Form a Vector Space

T_{q}^{p} (V) ​ ​ is a vector space

Because addition is closed, together with the tensor space already possessing a zero tensor, additive inverses, and scalar multiplication satisfying the usual axioms, the set of all type-(p,q) tensors on V forms a genuine vector space in its own right. Closure is the specific axiom among these that guarantees the addition operation does not need to be restricted to some proper subset of T⁰ₖ(V) to remain meaningful.

Dimension Counting Relies on Closure

The formula dim(T⁰ₖ(V)) = (dim V)^k for the space of covariant k-tensors, and its analogues for mixed types, presupposes that the tensor space is closed under addition and scalar multiplication, since dimension is only a meaningful invariant of a genuine vector space; without closure, "the space of type-(p,q) tensors" would not itself be a vector space subject to a well-defined dimension count.

Diagram of Closure Under Addition

What Closure Does Not Guarantee

No Closure Across Different Types or Spaces

T \in T_{1}^{0} (V), ​ ​ S \in T_{2}^{0} (V) ​ ​ \Rightarrow ​ ​ T + S ​ ​ undefined

Addition is only closed, and indeed only defined at all, within a single fixed type (p,q) on a single fixed space V; attempting to add a rank-1 covariant tensor to a rank-2 covariant tensor has no meaning, since the two objects accept different numbers of arguments and cannot be combined by the ordinary pointwise-sum formula. Closure is therefore a statement about a fixed type, not a claim that all tensors whatsoever can be added together.

Closure Is Distinct from Closure Under Other Operations

The tensor product of two type-(p,q) tensors is generally not of type (p,q) but of the larger combined type (2p,2q), so the tensor product operation is not closed on a fixed type space in the way addition is. Tensor addition closure property refers specifically to the addition operation and carries no implication about closure of the tensor product, contraction, or any other tensor operation, each of which must be assessed independently for its own closure behavior.

Role in the Broader Theory of Tensor Operation Closure

The Simplest Instance of a General Pattern

Tensor addition closure is the most elementary example within the broader study of which tensor operations remain within a given space of tensors and which instead produce tensors of a different type or on a different space. It serves as the base case establishing that tensor spaces of fixed type are well-behaved vector spaces, a fact subsequently relied upon whenever closure properties of more elaborate operations, such as the tensor product or contraction, are analyzed in relation to it.