##  [Triangularization](/triangularization-0) 

 Definition

The process of finding a basis in which a linear operator or matrix is represented by an upper-triangular matrix; the diagonal entries then list eigenvalues (counted with algebraic multiplicity) while entries above the diagonal encode generalized-eigenspace interactions.

 

 

 

 

 

 





## Principle

Principle

Triangularization exists once the characteristic (or minimal) polynomial splits over the base field: one constructs a flag of invariant subspaces so that the operator sends each subspace into itself and has scalar action on successive quotients, producing an upper-triangular matrix by similarity.

 

 

 

 

 





## Demonstration

Demonstration

Over C, Schur's theorem gives a unitary matrix U such that U* A U is upper triangular with eigenvalues on the diagonal; numerically this yields a stable reduction that reveals the spectrum without requiring full diagonalization.

 

 

 

 

## Misapplication

Misapplication

Treating an upper-triangular matrix as if it were diagonal—assuming the diagonal alone determines the operator's conjugacy class—ignores nilpotent couplings above the diagonal and can lead to wrong conclusions about powers or exponentials.

 

 

 

 

 





## Consequence

Consequence

Triangular form makes eigenvalues explicit and facilitates computation of characteristic polynomials and iterative numerical algorithms; it is a weaker but more generally available simplification than diagonalization.

 

 

 

 

## Reversal

Reversal

A reversal is diagonalization: when the off-diagonal entries above the diagonal vanish in some triangular representation, the operator is diagonalizable and the triangular form collapses to a diagonal one.

 

 

 

 

 





## Boundary

Boundary

Triangularization typically requires the polynomial to split in the base field; over fields where eigenvalues lie in extensions one may need a field extension. It does not in general provide canonical blocks (unlike Jordan or rational canonical forms).

 

 

 

 

 





## Semantic Tension

Semantic Tension

Tension appears between triangularization and Jordan form: both put eigenvalues on the diagonal, but Jordan refines the triangular structure into canonical nilpotent blocks while triangularization alone leaves noncanonical upper entries.

 

 

 

 

 





## Synthesis

Synthesis

Triangularization reduces an operator to a triangular action reflecting an invariant flag: it exposes the spectrum on the diagonal and encodes generalized-eigenspace relations above the diagonal, serving as a robust intermediate simplification when full diagonalization fails or is unavailable.