5.18 Tensor Product Ring Context

Tensor Product Ring Context is the generalization of the tensor product construction from vector spaces over a field to modules over an arbitrary ring, replacing the scalar field F with a ring R and the vector spaces V, W with R-modules, while preserving the same underlying quotient-by-relations recipe. This context is essential because many of the most important algebraic objects, such as ideals, abelian groups, and modules arising in algebraic geometry and number theory, are not vector spaces over a field but modules over a more general ring, and the tensor product construction must be adapted accordingly to remain applicable.

From Fields to Rings: What Changes

Modules Replace Vector Spaces

In the ring context, V and W are taken to be modules over a ring R, meaning they support addition and an action of R by "scalar" multiplication satisfying the module axioms, but without necessarily having the strong structural guarantees, such as every module having a basis, that vector spaces over a field enjoy.

The Tensor Product Over a Ring

Given a commutative ring R and R-modules M and N, the tensor product M ⊗_R N is constructed exactly as before, as a quotient of the free module R(M × N) by the submodule generated by the same four families of bilinear relations, now interpreted with scalars r ∈ R:

where S is generated by the analogous additivity and R-homogeneity relations.

Complications Absent in the Field Case

Failure of the Basis-Based Dimension Formula

Unlike vector spaces, a general R-module need not have a basis, so the convenient formula dim(V ⊗ W) = dim(V) · dim(W) has no direct analogue in the ring context; instead, more delicate invariants, such as the number of generators or the structure of torsion, must be tracked.

Tensor Products Can Collapse Unexpectedly

Over a general ring, tensoring can produce a much smaller module than naive expectation suggests, or even the zero module, as illustrated by ℤ/2ℤ ⊗_ℤ ℤ/3ℤ = 0, a phenomenon with no counterpart when tensoring nonzero vector spaces over a field, where the tensor product of two nonzero spaces is always nonzero.

Right and Left Module Considerations for Noncommutative Rings

When R is not commutative, the tensor product M ⊗_R N requires M to be a right R-module and N to be a left R-module, and the resulting tensor product is generally only an abelian group, not itself an R-module, unless additional bimodule structure is present.

Diagram Comparing Field and Ring Contexts

The Universal Property in the Ring Context

Bilinear Maps of Modules

The universal property carries over faithfully: for R-bilinear maps β: M × N → P into an R-module P, meaning β is additive and R-homogeneous in each argument, there exists a unique R-module homomorphism f: M ⊗_R N → P such that f(m ⊗ n) = β(m, n).

Preservation of the Categorical Structure

As in the field case, M ⊗_R N represents the functor sending P to the set of R-bilinear maps Bil_R(M × N, P), and this representability continues to guarantee uniqueness up to canonical isomorphism, independent of which specific quotient construction is used.

Special Cases Recovering Familiar Constructions

Tensoring Over the Integers

When R = ℤ, modules are simply abelian groups, and M ⊗_ℤ N computes the tensor product of abelian groups, a construction fundamental in algebraic topology for computing homology with different coefficient groups via the universal coefficient theorem.

Recovering the Vector Space Case

When R is a field F, every R-module is automatically a vector space with a basis, and the ring-context tensor product construction reduces exactly to the vector space tensor product described throughout the rest of this theory, confirming that the ring context is a strict generalization rather than a different construction.

Practical and Theoretical Significance

Foundation for Algebraic Geometry and Commutative Algebra

Tensor products over rings are indispensable in algebraic geometry, where fiber products of schemes correspond to tensor products of coordinate rings, and in commutative algebra, where flatness, a property defined in terms of exactness of the tensor product functor, governs the behavior of modules under base change.

Motivation for Studying Flatness and Torsion

The unexpected vanishing and collapsing behavior possible in the ring context motivates the study of properties such as flatness, which characterizes modules N for which tensoring with N preserves injective maps, a condition automatically satisfied by every vector space over a field but not by modules over a general ring.