원전 소개 — Lang

Summary

Lang's Linear Algebra is the theoretical, proof-first treatment of the subject. From the first page it works over an arbitrary field $K$ rather than just $\mathbb{R}$, so every definition and theorem is stated in full generality.¹ The book's pedagogical order is itself a thesis: linear maps are the primary objects, and matrices appear as their coordinate representations. Vector spaces (Ch I) and matrices (Ch II) are set up, then linear mappings are studied abstractly (Ch III) before the correspondence between maps and matrices is established (Ch IV).²

The middle of the book builds the structural machinery: scalar products, orthogonality, and the dual space (Ch V), then determinants developed axiomatically as multilinear alternating forms with existence and uniqueness proofs (Ch VI). The climax is spectral theory: symmetric, Hermitian, and unitary operators (Ch VII), eigenvalues and the characteristic polynomial (Ch VIII), and the spectral theorem (diagonalization of symmetric/Hermitian operators).

The final chapters reach toward graduate algebra. Lang treats polynomials of operators and the polynomial ring $K[t]$ as a Euclidean domain (Ch IX, XI), proving the Cayley–Hamilton theorem via triangulation (Ch X) and the primary decomposition and Jordan normal form through unique factorization in $K[t]$ (Ch XI). A final chapter on convex sets culminates in the Krein–Milman theorem (Ch XII) — unusual for a linear algebra text and a bridge to functional analysis.

Key Takeaways

• Everything is over a field $K$ (defined concretely as a subfield of $\mathbb{C}$, generalized later) — not just real or complex.³
• A vector space is axiomatized by VS 1–VS 8; subspaces, linear combinations, and the subspace generated by vectors follow.⁴
• Dimension is well defined because any two bases have the same number of elements (Thm 3.2), proved from the exchange theorem (Thm 3.1).⁵
• Linear maps before matrices: kernel, image, and the rank–nullity relation $\dim\ker f + \dim\operatorname{im} f = \dim V$ are developed for abstract maps; matrices come in Ch IV as the representation in chosen bases.
• Determinants are characterized as the unique multilinear, alternating, normalized function of the columns.
• Spectral theorem: a symmetric (real) / Hermitian (complex) operator has an orthonormal eigenbasis; unitary operators are diagonalizable.
• Canonical forms: Cayley–Hamilton, primary decomposition, and Jordan form all flow from the structure of $K[t]$.

Entities & Concepts

• 체와 벡터공간 — fields, the VS axioms, subspaces (Ch I)
• 기저와 차원 — linear independence, bases, invariance of dimension (Ch I)
• 합과 직합 — sums, direct sums, direct products (Ch I)
• 행렬과 선형방정식 — the space of matrices, linear systems, multiplication (Ch II)
• 선형사상 — maps, kernel, image, rank–nullity (Ch III)
• 선형사상과 행렬 — the map↔matrix correspondence, change of basis (Ch IV)
• 스칼라곱과 직교성 — scalar products, orthogonal bases, rank (Ch V)
• 쌍대공간과 쌍선형형식 — dual space, bilinear/quadratic forms, Sylvester's theorem (Ch V)
• 행렬식 — multilinear alternating forms, existence/uniqueness, Cramer (Ch VI)
• 고윳값과 특성다항식 — eigenvectors, eigenvalues (Ch VIII)
• 대칭·에르미트·유니터리 연산자 — the three operator classes (Ch VII)
• 스펙트럼 정리 — diagonalization of symmetric/Hermitian operators (Ch VIII)
• 연산자의 다항식 — $K[t]$, minimal polynomial, Cayley–Hamilton (Ch IX–X)
• 삼각화와 일차분해 — Schur, primary decomposition (Ch X–XI)
• 조르당 표준형 — the Jordan canonical form (Ch XI)
• 볼록집합과 크레인–밀만 정리 — convex sets, hyperplanes, Krein–Milman (Ch XII)
• 읽기 경로 — chapter-by-chapter route through this book

Relation to Other Wiki Pages

Lang is the sole, theoretically-oriented source for this wiki. Its abstract over-a-field approach contrasts with computational linear-algebra texts (Anton, Strang) that lead with $\mathbb{R}^n$ and row reduction. Its later chapters ($K[t]$, primary decomposition, Jordan form) point toward module theory and abstract algebra; its convex-sets chapter toward functional analysis.