---
title: Simple-Minded Systems in Representation Theory
url: https://www.emergentmind.com/topics/simple-minded-system
type: topic
---

# Simple-Minded Systems in Representation Theory

Simple-minded systems are finite orthogonal families of indecomposable objects that serve, in stable and related triangulated categories, as analogues of the simple objects of an abelian category. In the stable module category of a self-injective algebra, they are defined by a brick condition and an extension-generation condition, and they occur naturally as images of simple modules under stable equivalences [1009.1427]. In derived and negative Calabi–Yau settings, the adjacent notions of simple-minded collections and \(w\)-simple-minded systems place the same idea into a broader framework of \(t\)-structures, exact structures, mutation, and reduction [2010.11799], [1808.02519].

## 1. Definition and basic formalism

Let \(A\) be a finite-dimensional self-injective algebra over an algebraically closed field \(k\), and let \(A\text{-}\underline{\mathrm{mod}}\) or \(A\text{-}\stmod\) denote the stable module category, whose morphisms are module maps modulo those factoring through projectives. In this category, a nonzero object \(M\) is a **stable brick** if
\[
\End_{A\text{-}\stmod}(M)\cong k.
\]
A finite family \(\mathcal S=\{S_1,\dots,S_t\}\) of stable bricks is an **orthogonal system** if
\[
\Hom_{A\text{-}\stmod}(S_i,S_j)=0 \qquad (i\neq j).
\]
Its extension closure is defined recursively by
\[
(\mathcal S)_0=\{0\},\qquad (\mathcal S)_1=\mathcal S\cup\{0\},\qquad
(\mathcal S)_n=(\mathcal S)_{n-1}\star(\mathcal S\cup\{0\}),
\]
where \(\mathcal X\star\mathcal Y\) consists of the middle terms of triangles \(X\to Z\to Y\to X[1]\) with \(X\in\mathcal X\) and \(Y\in\mathcal Y\). The system \(\mathcal S\) is a **simple-minded system** if it is orthogonal and
\[
\mathcal F(\mathcal S)=\bigcup_{n\ge 0}(\mathcal S)_n = A\text{-}\stmod
\]
or, in the more general Artin-algebra formulation, every indecomposable non-projective module can be obtained from \(\mathcal S\) by finitely many extensions up to projective summands [2006.14289], [1009.1427].

The definition has a triangulated analogue beyond stable module categories. In a Hom-finite triangulated category \(\mathcal C\), a \(w\)-orthogonal collection \(\mathcal S=\{S_1,\dots,S_e\}\) satisfies Schurity together with
\[
\Hom_\mathcal C(S_i,\Sigma^\ell S_j)=0 \qquad \text{for } \ell\in\{-w+1,\dots,-1\},
\]
and it is a \(w\)-simple-minded system if
\[
\mathcal C=\langle \mathcal S\rangle * \Sigma^{-1}\!\langle \mathcal S\rangle * \cdots * \Sigma^{-w+1}\!\langle \mathcal S\rangle.
\]
For \(w=1\), this recovers the usual simple-minded collection formalism in derived categories [2010.11799].

Related weakened notions also appear. In the representation-finite self-injective setting, generation can be tested by the condition that for every non-projective indecomposable \(X\) there exists \(S\in\mathcal S\) with \(\Hom_{A\text{-}\stmod}(X,S)\neq 0\) [2006.14289]. Zhang defines a **weakly simple-minded system** by requiring that for every nonzero \(X\) there exists \(S\in\mathcal S\) with \(\Hom(S,X)\neq 0\) [2606.21881].

## 2. Stable equivalence and invariant-theoretic significance

The original motivation is that simple modules behave well under passage to stable categories. If
\[
a\colon \underline{\bmod}A \xrightarrow{\ \simeq\ } \underline{\bmod}B
\]
is a stable equivalence, then the image under \(a\) of a complete set of non-projective simple \(A\)-modules is again a simple-minded system. Orthogonality is preserved because stable equivalences preserve stable Hom-spaces, and generation is preserved because the extension-filtration defining an sms is transported across the equivalence [1009.1427].

Koenig–Liu prove the corresponding invariance theorem: if \(a\colon\underline{\bmod}A\to\underline{\bmod}B\) is a stable equivalence and \(\mathcal S\) is a simple-minded system in \(\bmod_{\mathrm p}A\), then \(a(\mathcal S)\) is a simple-minded system over \(B\). Consequently, the set of simple-minded systems, up to isomorphism, depends only on the stable category and not on the algebra itself [1009.1427].

This immediately links simple-minded systems to the Auslander–Reiten conjecture. That conjecture asks whether stably equivalent Artin algebras must have the same number of non-isomorphic non-projective simple modules. Koenig–Liu formulate the associated question of whether every simple-minded system has cardinality equal to the number of non-projective simple modules; if so, invariance of sms’s would imply the conjecture. They verify this in several classes: triangular algebras, one-point extensions, Nakayama algebras, and group algebras of \(p\)-groups [1009.1427].

The same paper identifies explicit class-specific behavior. For triangular algebras \(A=kQ/I\) with \(Q\) acyclic, there is exactly one simple-minded system, namely the set of all non-projective simple modules. For one-point extensions
\[
A=\begin{pmatrix}B&M\\0&k\end{pmatrix},
\]
the sms’s of \(A\) are precisely those obtained by adjoining the new simple injective to an sms of \(B\). For Nakayama algebras with non-projective simples \(S_1,\dots,S_n\), any sms has exactly \(n\) elements and the sets of socles and tops of its terms are both permutations of \(\{S_1,\dots,S_n\}\). For group algebras of \(p\)-groups in characteristic \(p\), the sms’s are exactly the endo-trivial modules [1009.1427].

## 3. Representation-finite self-injective algebras

For representation-finite self-injective algebras, the definition simplifies substantially. Let \(A\) be indecomposable, basic, representation-finite self-injective, let \(\ell(A)\) be the number of isomorphism classes of simple \(A\)-modules, and let \(\nu= D\Hom_A(-,A)\) be the Nakayama functor on \(A\text{-}\stmod\). Guo–Liu–Ye–Zhang prove that a family \(\mathcal S\) in \(A\text{-}\stmod\) is a simple-minded system if and only if it satisfies the three conditions
1. \(\mathcal S\) is an orthogonal system,
2. \(|\mathcal S|=\ell(A)\),
3. \(\nu(\mathcal S)=\mathcal S\),
so in this context orthogonality, the correct cardinality, and Nakayama-stability already imply generation [2006.14289].

Their proof combines torsion-pair methods with covering theory on the repetitive quiver \(k(\mathbb Z\Delta)/(\tau^N)\). Starting from \(\mathcal Y=\mathcal F(\mathcal S)\) and perpendicular subcategories \(\mathcal X\) and \(\mathcal Z\), they show that the “negative extension” space \(\mathcal Z\star \mathcal S\) must vanish. This forces \(\mathcal Z=\{0\}\), hence \(\mathcal F(\mathcal S)=A\text{-}\stmod\) [2006.14289].

The same paper establishes an extendibility theorem: every Nakayama-stable orthogonal system in \(A\text{-}\stmod\) can be enlarged through a finite chain of Nakayama-stable orthogonal systems until it becomes a simple-minded system. The construction proceeds by selecting a nonzero indecomposable in the negative part of an associated torsion pair, extracting a brick or one of its \(\nu\)-iterates that is orthogonal to the current system, adjoining its \(\nu\)-orbit, and repeating until termination, which is guaranteed by finite representation type [2006.14289].

Chan–Koenig–Liu place this result into a classification picture. For a representation-finite self-injective algebra \(A\), every simple-minded system in \(\stmod A\) is realized as the image of the simple modules of some self-injective algebra \(B\) under a stable equivalence of Morita type, and \(B\) is unique up to Morita equivalence. They also show that every sms lifts to a Nakayama-stable simple-minded collection in \(D^b(\mathrm{mod}\,A)\) [1305.2576].

This classification is encoded combinatorially by **Riedtmann configurations**. If the stable Auslander–Reiten quiver is \(\Gamma=\mathbb ZQ/\langle \tau^h\rangle\), then configurations of \(\Gamma\) correspond exactly to sms’s in \(\stmod A\). On the derived side, the liftability result implies that sms’s can be generated algorithmically via mutation, and the sms quiver is connected for representation-finite self-injective algebras [1305.2576].

A particularly explicit instance is the self-injective Nakayama case. Guo–Liu–Ye–Zhang give a complete construction of all sms’s of \(A(n,l+1)\) using non-crossing partitions, first in the symmetric case \(A(n,dn)\) and then in general via a covering from a symmetric algebra \(A(e,ed)\) where \(e=\gcd(n,l)\). Each sms belongs to one of two uniform families \(\mathcal L'_{p,k}\) or \(\mathcal S'_{p,k}\), obtained by pulling back corresponding constructions from the symmetric cover [2006.14287].

## 4. Torsion pairs, mutation, reduction, and gluing

Dugas formulates sms-theory in any Hom-finite Krull–Schmidt triangulated category \(T\). Given an sms \(S\subset T\) and a subset \(X\subseteq S\), he constructs torsion pairs associated to \(X\), proves functorial finiteness of the relevant extension-closed subcategory, and defines left and right mutations \(\mu_X^+(S)\) and \(\mu_X^-(S)\) using minimal triangles determined by those torsion pairs. If \(T\) has a Serre functor \(\nu\) and both \(S\) and \(X\) are invariant under \(\nu[1]\), then both mutations are again simple-minded systems [1207.7338].

In the stable category of a self-injective algebra, this mutation formalism parallels the mutation of simple-minded collections in derived categories. Dugas shows that mutating the set of simple \(\Lambda\)-modules relative to \(X\) yields the images of the simple \(\Gamma\)-modules under a stable equivalence \(\underline{\mathrm{mod}}\Gamma\to\underline{\mathrm{mod}}\Lambda\), where \(\Gamma\) is the tilting mutation of \(\Lambda\) relative to \(X\) [1207.7338].

For \(w\)-simple-minded systems in \((-w)\)-Calabi–Yau triangulated categories, Coelho Simões–Pauksztello develop an analogous reduction theory. If \(S\) is \(w\)-orthogonal and \((S)\) is functorially finite, then the bi-perpendicular
\[
Z=\{d\in D \mid \Hom_k(S,d)=0=\Hom_k(d,S)\text{ for }0\le k<w-1\}
\]
inherits a triangulated structure, and there is a bijection
\[
\{\,w\text{-simple-minded systems }T\subseteq D \text{ with }S\subseteq T\,\}
\longleftrightarrow
\{\,w\text{-simple-minded systems }R\subseteq Z\,\}
\]
given by \(T\mapsto T\setminus S\) and \(R\mapsto R\cup S\) [1808.02519].

A direct SMS reduction-and-gluing theory for stable module categories of self-injective algebras is developed in 2026. If \(\mathcal S\) is a Nakayama-stable orthogonal system with \(\Omega(\mathcal S)\subseteq\mathcal F(\mathcal S)\) and \(\mathcal F(\mathcal S)\) functorially finite, then the pairs \(({}^\perp\mathcal S,\mathcal F(\mathcal S))\) and \((\mathcal F(\mathcal S),\mathcal S^\perp)\) are stable \(t\)-structures, hence induce a recollement. For a Nakayama-stable subset \(\mathcal S_0\subset\mathcal S\), the bi-perpendicular
\[
\mathcal D={}^\perp(\mathcal S_0)^\perp
\]
carries a triangulated structure and is triangle equivalent to \(B\text{-}\underline{\mathrm{mod}}\) for \(B=eAe\). There is then a bijection between sms’s in \(A\text{-}\underline{\mathrm{mod}}\) containing \(\mathcal S_0\) and sms’s in \(\mathcal D\), and conversely one can glue sms’s from the outer terms of the recollement to obtain an sms in the middle category [2606.16081].

These mutation and reduction formalisms are structurally parallel to positive cluster-tilting theory. The literature repeatedly describes sms-theory, especially the \(w\)-theory, as a “negative” counterpart to cluster-tilting, with generation in negative shifts replacing positive-shift vanishing and with abelian subcategories or reduced triangulated categories replacing quotient constructions [2010.11799], [1808.02519].

## 5. Derived categories, \(w\)-systems, and abelian subcategories

The closely related notion in derived categories is a **simple-minded collection**. If \(A\) is a finite-dimensional algebra, then the simple \(A\)-modules form a simple-minded collection in \(\operatorname D^b(\operatorname{mod}A)\); their extension closure is exactly \(\operatorname{mod}A\), hence abelian [2010.11799]. More generally, in an essentially small \(k\)-linear Hom-finite triangulated category \(\mathcal C\) with split idempotents, a simple-minded collection is a \(1\)-orthogonal collection whose smallest extension-closed subcategory containing it is all of \(\mathcal C\), equivalently a finite family with no negative self-extensions that generates under extensions and shifts [2010.11799].

Jørgensen proves that if \(w>2\) and \(\mathcal S\) is a \(w\)-orthogonal collection in \(\mathcal C\), then its extension closure \(\langle \mathcal S\rangle\) is a **proper abelian subcategory** of \(\mathcal C\), every object of \(\langle \mathcal S\rangle\) has finite length, and its simple objects are exactly the members of \(\mathcal S\). The proof uses Quillen exact categories and Dyer’s theorem, which applies once one knows
\[
\Hom_\mathcal C(\langle\mathcal S\rangle,\Sigma^{-1}\langle\mathcal S\rangle)=0.
\]
When \(\mathcal S\) is a \(w\)-simple-minded system, this abelian subcategory supports an internal tilting theory, even though for \(w\ge 2\) it is typically not the heart of a \(t\)-structure [2010.11799].

This perspective clarifies the relation between sms’s and hearts. In the derived \(w=1\) case, the extension closure of a simple-minded collection is the heart of a bounded \(t\)-structure. In the higher negative case, one retains an abelian subcategory or exact structure but not necessarily a heart. Jørgensen explicitly describes this as “negative cluster tilting theory,” contrasting the negative vanishing
\[
\Hom_\mathcal C(\mathcal S,\Sigma^j\mathcal S)=0 \qquad (-w+1\le j\le -1)
\]
with the positive vanishing used in higher cluster theory [2010.11799].

Further refinements connect mutation of simple-minded collections, \(w\)-simple-minded systems, and HRS-tilting. Broomhead–Coelho Simões–Pauksztello–Woolf characterize when a simple HRS tilt of a length heart is again a length heart, prove sufficient conditions for infinite iterability of simple-minded mutation via simple-minded reduction, and give a common mutation-pair framework covering both simple-minded collections and \(w\)-simple-minded systems [2401.02947]. Jin’s reduction theory for simple-minded collections shows that SMS reduction in singularity categories is the “shadow” of SMC reduction in an ambient triangulated category, paralleling the relation between Calabi–Yau reduction and silting reduction [1907.05114].

## 6. Geometric restrictions, special classes, and recent developments

The Auslander–Reiten geometry of a stable category imposes strong restrictions on possible sms’s. Chan–Liu–Zhang prove that if \(\mathcal C\) is a stably quasi-serial component of rank \(n\) in the Auslander–Reiten quiver of a self-injective algebra and \(\mathcal S\) is an sms, then
\[
|\mathcal S\cap\mathcal C|<n,
\qquad
\mathrm{ql}(M)<n \text{ for each } M\in \mathcal S\cap\mathcal C.
\]
In particular, no module in a homogeneous tube can belong to a simple-minded system [1909.04440].

Domestic Brauer graph algebras provide a representation-infinite setting in which sms’s can nevertheless be characterized sharply. Zhang proves that for a domestic Brauer graph algebra an orthogonal system \(\mathcal S\subseteq A\text{-stmod}\) is a simple-minded system if and only if it contains at least one non-periodic module, equivalently one object from each Euclidean component, and satisfies the syzygy-closure condition
\[
\Omega(\mathcal S)\subseteq \mathcal F(\mathcal S).
\]
The same work shows that if a finite orthogonal system is weakly simple-minded and \(\Sigma(\mathcal S)\subseteq \mathcal F(\mathcal S)\), then it is already an sms [2606.21881].

For \(2\)-domestic Brauer graph algebras, a 2025 construction gives all sms’s explicitly. The stable Auslander–Reiten quiver has two Euclidean components, four exceptional quasi-tubes, and infinitely many homogeneous tubes; an sms is produced by choosing a maximal orthogonal set of non-periodic bricks in one Euclidean component, filling out the quasi-tubes via stable bi-perpendicular triangular regions, and recovering the remaining Euclidean terms by \(\Omega\). The resulting orthogonal system has size equal to the number of non-projective simples and satisfies \(\Omega^{-1}(\mathcal S)\subset \mathcal F(\mathcal S)\), hence is an sms. The same paper shows that every weakly simple-minded system of finite cardinality is an sms in this setting [2509.24184].

For \(1\)-domestic Brauer graph algebras, Zhang reduces classification to the \(2\)-domestic case by covering theory. If
\[
\overline F\colon C\text{-}\underline{\mathrm{mod}}\to A\text{-}\underline{\mathrm{mod}}
\]
is the covering functor from a \(2\)-domestic Brauer graph algebra \(C\), then \(\varphi\)-stable sms’s in \(C\) correspond exactly to sms’s in \(A\). The resulting classification can be described as a “seed + bi-perp” construction: choose one object in the Euclidean component, compute its stable bi-perpendicular inside the Euclidean and quasi-tube components, and obtain a maximal orthogonal system that is automatically an sms [2606.16085].

A different line of development relates sms-theory to coherent rings. Zhang shows that for an orthogonal system \(\mathcal S\) with extension-closed preimage generated by a finite generator \(M\), coherence of \(\operatorname{End}_A(M)\) and finendo hypotheses force covariant finiteness of \(\mathcal F(\mathcal S)\), and under additional weak-sms assumptions this implies that \(\mathcal S\) is an sms. In particular, over left pure-semisimple rings, every weakly simple-minded system is already a simple-minded system [2403.07619].

Taken together, these developments present simple-minded systems as a unifying notion linking stable equivalence, Auslander–Reiten geometry, mutation theory, recollement and reduction, \(t\)-structures and exact categories, and explicit classification problems in self-injective representation theory.

Source: https://www.emergentmind.com/topics/simple-minded-system