Chapter 48: Modules over a Principal Ideal Domain (PID)¶

Prerequisites: Vector Spaces (Ch04) · Polynomial Algebra (Ch00) · Rational Canonical Form (Ch13B)

Chapter Outline: From Vector Spaces to Modules → Definition of Modules over a Ring $R$ → Submodules, Quotients, and Homomorphisms → Free Modules & Rank → Basics of Principal Ideal Domains (PIDs) → The Fundamental Structure Theorem for Finitely Generated Modules over a PID → Decomposition into Free and Torsion Parts → Unified View: The Algebraic Essence of Jordan and Rational Canonical Forms → Application: Classification of Finitely Generated Abelian Groups

Extension: Module theory over a PID is the "Grand Unification" of linear algebra; it proves that all canonical form theories are essentially decompositions of module structures over the specific PID $F[x]$, serving as one of the pinnacles of modern abstract algebra.

When we relax the restriction in vector spaces that "scalars come from a field" to "scalars come from a ring," we obtain the concept of a Module. Over a Principal Ideal Domain (PID), such as $\mathbb{Z}$ or $F[x]$, the structure of modules exhibits exceptional symmetry and regularity. This chapter will prove that the Jordan and Rational Canonical Forms of matrices are essentially projections of the same deep algebraic theorem onto different contexts.

48.1 Definition and Foundations of Modules¶

Definition 48.1 (Module)

Let $R$ be a commutative ring with identity. An $R$-module $M$ is a set equipped with addition and $R$-scalar multiplication satisfying the 8 axioms similar to those of a vector space. - If $R$ is a field, an $R$-module is a vector space. - If $R = \mathbb{Z}$, an $R$-module is an Abelian group.

Definition 48.2 (Finitely Generated and Free Modules)

Finitely Generated: If $M$ can be spanned by a finite number of elements.
Free Module: If $M$ possesses a basis (linearly independent and spanning). Not all modules have bases.

48.2 The Structure Theorem over a PID¶

Theorem 48.1 (Structure Theorem for Finitely Generated Modules over a PID)

Let $R$ be a PID, and let $M$ be a finitely generated $R$-module. Then $M$ can be uniquely decomposed as: $$M \cong R^r \oplus R/(d_1) \oplus R/(d_2) \oplus \cdots \oplus R/(d_k)$$ where: - $R^r$ is the free part, and $r$ is the rank. - $R/(d_i)$ are the torsion parts, satisfying $d_1 \mid d_2 \mid \cdots \mid d_k$. These $d_i$ are the invariant factors of $M$.

48.3 A Unified Perspective on Canonical Forms¶

Operators as Module Actions

Given an $n \times n$ matrix $A$, we can treat $\mathbb{C}^n$ as a $\mathbb{C}[\lambda]$-module where the action is defined by: $p(\lambda) \cdot \mathbf{v} = p(A)\mathbf{v}$. - The Rational Canonical Form corresponds to the invariant factor decomposition above. - The Jordan Canonical Form corresponds to the elementary divisor decomposition (factoring each $d_i$ into powers of irreducible linear terms).

Exercises¶

1. [Basics] Prove that every Abelian group can be viewed as a $\mathbb{Z}$-module.

Solution

Proof: 1. Let $(G, +)$ be an Abelian group. 2. Define scalar multiplication $n \cdot a$: - If $n > 0$, $n \cdot a = a + a + \cdots + a$ ($n$ times). - If $n = 0$, $0 \cdot a = 0_G$. - If $n < 0$, $n \cdot a = -(|n| \cdot a)$. 3. It is easy to verify that this definition satisfies all module axioms (distributivity, associativity, etc.). Conclusion: The theory of Abelian groups is essentially the module theory over the ring of integers.

2. [Free] Give an example of a $\mathbb{Z}$-module that is not free.

Solution

Typical Example: Finite cyclic group $\mathbb{Z}_n$ (e.g., $\mathbb{Z}/2\mathbb{Z}$). Analysis: 1. In $\mathbb{Z}/2\mathbb{Z}$, for any element $a$, we have $2 \cdot a = 0$. 2. According to the definition of linear dependence, the scalar $2 \neq 0$ but $2 \cdot a = 0$ implies any single element is linearly dependent. 3. A free module requires basis vectors to be linearly independent. Conclusion: A module containing non-zero torsion elements (elements $m$ such that $r \cdot m = 0$ for some $r \neq 0$) cannot be a free module.

3. [Factors] If $M \cong \mathbb{Z} \oplus \mathbb{Z}/2\mathbb{Z} \oplus \mathbb{Z}/6\mathbb{Z}$, what are its rank and invariant factors?

Solution

Analysis: 1. Free Part: The term $\mathbb{Z}$ appears once, so the rank $r = 1$. 2. Torsion Part: Composed of $\mathbb{Z}/2\mathbb{Z}$ and $\mathbb{Z}/6\mathbb{Z}$. 3. Check the divisibility chain: $2 \mid 6$, which matches the standard form of the structure theorem. Conclusion: The rank is 1, and the sequence of invariant factors is $(2, 6)$.

4. [Smith Form] How are invariant factors in module theory related to the Smith form of $\lambda$-matrices?

Solution

Essential Unity: 1. Let $A$ be a square matrix. Transform its characteristic matrix $\lambda I - A$ into Smith Normal Form. 2. The non-trivial polynomials on the diagonal are $d_1(\lambda), \ldots, d_k(\lambda)$. 3. The vector space viewed as a $\mathbb{C}[\lambda]$-module has the structure $\bigoplus \mathbb{C}[\lambda]/(d_i(\lambda))$. Conclusion: The algebraic significance of the Smith Normal Form is revealing the module decomposition of the space under the action of a linear operator.

5. [PID] Why is module theory over $F[x, y]$ much more complex than over $F[x]$?

Solution

Algebraic Context: 1. $F[x]$ is a PID, where every ideal is generated by a single element, ensuring the simplicity of the structure theorem. 2. $F[x, y]$ is not a PID. For example, the ideal $(x, y)$ cannot be generated by a single polynomial. 3. The lack of the PID property means modules cannot be decomposed into a simple direct sum of cyclic modules. Conclusion: Representation theory over polynomial rings in more than one variable is the domain of algebraic geometry, extending far beyond linear algebra.

6. [Eigenvalues] If $M$ is a pure torsion module over $\mathbb{C}[\lambda]$, what is its physical meaning?

Solution

Explanation: 1. A pure torsion module means there is no free part ($r=0$). 2. In matrix theory, this implies the dimension of the vector space is finite. 3. Every vector is "annihilated" by some polynomial (i.e., $p(A)v = 0$). Conclusion: Every linear operator on a finite-dimensional vector space makes the corresponding module a pure torsion module.

7. [Divisors] Decompose $\mathbb{Z}/12\mathbb{Z}$ into its elementary divisor form.

Solution

Steps: 1. Perform prime power factorization on the modulus: $12 = 2^2 \cdot 3^1 = 4 \cdot 3$. 2. Use the Chinese Remainder Theorem (or the elementary divisor decomposition theorem): Conclusion: $\mathbb{Z}/12\mathbb{Z} \cong \mathbb{Z}/4\mathbb{Z} \oplus \mathbb{Z}/3\mathbb{Z}$. The elementary divisors are $4$ and $3$.

8. [Rank] Prove that for a finite-dimensional vector space, $r = \dim V$ and the torsion part is zero.

Solution

Proof: 1. A vector space is a module over a field $F$. 2. The only ideals of a field $F$ are $\{0\}$ and $F$. 3. Any $F/(d_i)$ where $d_i \neq 0$ collapses to the zero space (since non-zero elements in a field are invertible). 4. Thus, the torsion part must be zero. 5. The entire module is $F^r$, where $r$ is clearly equal to the dimension of the space.

9. [Cyclic] What is a cyclic module? What structure in matrices does it correspond to?

Solution

Definition: A module generated by a single element, taking the form $R/(d)$. Matrix Correspondence: It corresponds to a companion matrix block. This means the entire space can be spanned by a single vector $\mathbf{v}$ through repeated application of the operator $A$: $\{ \mathbf{v}, A\mathbf{v}, A^2\mathbf{v}, \ldots \}$.

10. [Unity] Why is this theory called the "Grand Unification" of linear algebra?

Solution

Philosophical Insight: 1. It unifies seemingly isolated properties of integers (prime factorization) and polynomials (irreducible factorization) under the PID framework. 2. It proves that all matrix similarity theories (Jordan, Rational, Smith) are merely applications of the same module decomposition theorem under different scalar rings. 3. It provides a unified logical starting point for moving from linear systems to broader algebraic structures (like Lie algebras and ring theory).

Chapter Summary¶

Module theory over a PID achieves the "Grand Closure" of linear algebra logic:

High Degree of Unification: It proves that the factorization of integers and the factorization of polynomials are algebraically identical, establishing a universal structural model.
Origin of Canonical Forms: Jordan and Rational forms are no longer isolated tricks but inevitable consequences of the PID module decomposition theorem under specific ring actions.
Structural Resolution: Through the division into free and torsion parts, module theory clearly defines the boundary between "infinite degrees of freedom" and "restricted cyclic structures" in linear systems.