Mathematics.

algebraic representations

Quiver Representations

Representation Theory75 minDifficulty8 out of 10

You should know: group representations, modules

Overview

A quiver is a directed graph, and a representation of a quiver Q assigns a vector space to each vertex and a linear map to each arrow. Quiver representations provide a unified framework for studying linear algebra problems (Jordan normal form, flags, etc.) and encode module categories over path algebras. Gabriel's theorem classifies quivers of finite representation type: they are exactly those whose underlying undirected graph is a Dynkin diagram of type A, D, or E.

Intuition

A quiver representation is like a system of vector spaces connected by linear maps — think of it as a 'wiring diagram' for linear algebra. The path algebra kQ encodes all composition of arrows, and a quiver representation is exactly a module over kQ. Gabriel's theorem is a beautiful surprise: the indecomposable representations of a Dynkin quiver are in bijection with positive roots of the root system.

Formal Definition

Definition

A quiver Q = (Q₀, Q₁, s, t) consists of a set Q₀ of vertices, a set Q₁ of arrows, and source/target maps s, t: Q₁ → Q₀. A representation M of Q over a field k assigns:

M:Q0{k-vector spaces},M:Q1{k-linear maps},i.e.  MiVectk,  Mα:Ms(α)Mt(α)M: Q_0 \to \{\text{k-vector spaces}\}, \quad M: Q_1 \to \{\text{k-linear maps}\}, \quad \text{i.e.} \; M_i \in \mathbf{Vect}_k, \; M_\alpha: M_{s(\alpha)} \to M_{t(\alpha)}
Quiver representation: vector spaces and linear maps
Rep(Q)kQ-mod(category of representationsmodules over the path algebra)\mathrm{Rep}(Q) \cong kQ\text{-mod} \quad (\text{category of representations} \cong \text{modules over the path algebra})
Equivalence with path algebra modules
dimM=(dimkMi)iQ0Z0Q0(dimension vector)\mathbf{dim}\, M = (\dim_k M_i)_{i \in Q_0} \in \mathbb{Z}_{\geq 0}^{Q_0} \quad (\text{dimension vector})
Dimension vector
χQ(d,e)=iQ0dieiαQ1ds(α)et(α)(Euler form)\chi_Q(\mathbf{d}, \mathbf{e}) = \sum_{i \in Q_0} d_i e_i - \sum_{\alpha \in Q_1} d_{s(\alpha)} e_{t(\alpha)} \quad (\text{Euler form})
Tits/Euler form of the quiver

Notation

NotationMeaning
QQQuiver (directed graph)
kQkQPath algebra of Q over k
dimM\mathbf{dim}\, MDimension vector of representation M
Rep(Q)\mathrm{Rep}(Q)Category of k-representations of Q

Properties

Simple representations

ForeachvertexiQ0,thereisasimplerepresentationSiwith(Si)i=k,(Si)j=0forji,andallmapszero.For each vertex i ∈ Q₀, there is a simple representation Sᵢ with (Sᵢ)ᵢ = k, (Sᵢ)ⱼ = 0 for j ≠ i, and all maps zero.

Projective indecomposables

TheindecomposableprojectivekQmodulePicorrespondingtovertexihasdimensionvectorgivenbythenumberofpathsfromitoeachvertex.The indecomposable projective kQ-module Pᵢ corresponding to vertex i has dimension vector given by the number of paths from i to each vertex.

Theorems

Theorem 1: Gabriel's Theorem
A connected quiver Q (with no oriented cycles) is of finite representation type (finitely many indecomposable representations up to isomorphism) if and only if the underlying undirected graph is a Dynkin diagram of type Aₙ, Dₙ (n≥4), E₆, E₇, or E₈. In this case, the indecomposable representations are in bijection with positive roots of the corresponding root system.
Theorem 2: Krull–Schmidt Theorem for Quivers
Every finite-dimensional representation of a finite quiver decomposes uniquely (up to isomorphism and permutation of summands) as a direct sum of indecomposable representations.

Worked Examples

  1. 1

    A representation assigns vector spaces V₁, V₂ and a linear map f: V₁ → V₂. By the classification of linear maps (rank-nullity), indecomposables arise from f being 0, injective, or the identity on 1-dim spaces.

    Rep(A2):V1fV2\text{Rep}(A_2): \quad V_1 \xrightarrow{f} V_2
  2. 2

    Indecomposables: S₁ = (k → 0) with dim vector (1,0), S₂ = (0 → k) with dim vector (0,1), P₁ = (k →^{id} k) with dim vector (1,1).

    Indecomposables:S1:k0,S2:0k,P1:kidk\text{Indecomposables:} \quad S_1: k \to 0, \quad S_2: 0 \to k, \quad P_1: k \xrightarrow{\mathrm{id}} k
  3. 3

    There are 3 indecomposables, matching the 3 positive roots of A₂: α₁, α₂, α₁+α₂ (Gabriel's theorem).

    Positive roots of A2:α1,  α2,  α1+α2(3 roots=3 indecomposables)\text{Positive roots of } A_2: \alpha_1,\; \alpha_2,\; \alpha_1+\alpha_2 \quad (3 \text{ roots} = 3 \text{ indecomposables})

✓ Answer

A₂ has 3 indecomposable representations, corresponding to the 3 positive roots of A₂.

Practice Problems

Mediumfree response

For the quiver D₄ (one central vertex with 3 outgoing arrows to leaves), how many positive roots does the D₄ root system have, and therefore how many indecomposable representations does the D₄ quiver have?

Hardfree response

Describe the path algebra kQ for the quiver Q: 1 → 2 → 3 (type A₃). List a basis and explain the multiplication.

Quiz

Gabriel's theorem states that a quiver Q has finite representation type if and only if its underlying graph is:
A representation of a quiver Q is equivalent to a module over:
The number of indecomposable representations of a Dynkin quiver equals:

Summary

  • A quiver representation assigns vector spaces to vertices and linear maps to arrows.
  • Rep(Q) ≅ kQ-mod: representations are modules over the path algebra.
  • Gabriel's theorem: finite representation type iff underlying graph is a Dynkin diagram (A, D, E).
  • For Dynkin quivers, indecomposables are in bijection with positive roots.
  • The Jordan quiver (one vertex, one loop) has infinitely many indecomposables (Jordan blocks).

References

  1. BookAssem, I., Simson, D. & Skowroński, A. — Elements of the Representation Theory of Associative Algebras (2006)
  2. BookKirillov, A. Jr. — Quiver Representations and Quiver Varieties (2016)