A scalar-valued function assigned to a square linear operator or matrix that records oriented volume scaling and indicates invertibility; algebraically it is an alternating multilinear form of the columns (or rows).

Principle

The determinant equals the product of eigenvalues (counted with algebraic multiplicity) and is multiplicative across composition: det(AB)=det(A)det(B); it vanishes precisely when the linear transformation is singular.

Demonstration

For a 2×2 matrix [[a,b],[c,d]] the determinant ad−bc measures the signed area scaling of parallelograms under the linear map; for a diagonal 3×3 matrix with entries λ1,λ2,λ3 the determinant is λ1λ2λ3 and equals the volume scale factor.

Misapplication

Using the determinant for non-square matrices, or treating a small determinant as a reliable numerical indicator of near-singularity without accounting for conditioning and scaling, leads to errors.

Consequence

Correct use yields multiplicative volume change under change of variables (Jacobian) and a simple invertibility test (det≠0); algebraic consequences include characteristic polynomial relations and orientation detection.

Reversal

Replacing the determinant by the permanent removes sign alternation and orientation information; inverting the sign convention flips orientation but preserves magnitude of volume scaling.

Boundary

Defined for endomorphisms of finite-dimensional vector spaces (or matrices over a commutative ring); not directly defined for arbitrary linear operators on infinite-dimensional spaces except via specialized constructions (Fredholm determinants, regularized determinants) with extra hypotheses.

Semantic Tension

Determinant competes with trace and spectrum: trace gives additive eigenvalue information, the spectrum lists eigenvalues, while the determinant compresses multiplicative eigenvalue information and orientation into one scalar.

Synthesis

The determinant is the scalar summary of how a square linear map rescales oriented volume; it is computed from an alternating multilinear form, equals the product of eigenvalues in finite dimensions, and is central to invertibility, change-of-variables, and orientation questions.