Supersolvable Arrangements

• Supersolvable arrangements are central hyperplane arrangements defined by an intersection lattice admitting a maximal chain of modular elements.
• They exhibit linear factorization of invariants like the characteristic and Poincaré polynomials, ensuring freeness and a fiber-type structure.
• Graph-derived examples, such as chordal graphical arrangements, illustrate how combinatorial properties directly translate into algebraic and topological regularities.

A supersolvable arrangement is a central hyperplane arrangement whose intersection lattice $L(\mathcal{A})$ admits a maximal chain of modular elements. Introduced lattice-theoretically by Stanley and subsequently recast in arrangement language, supersolvability is one of the strongest combinatorial regularity conditions in arrangement theory. It forces linear factorization phenomena for characteristic and Poincaré polynomials, implies freeness and fiber-type structure, and in several major families admits exact combinatorial characterizations by chordality, modular flats of low rank, or recursive decomposition data (Liang et al., 2024, Hoge et al., 2014, Hoge et al., 2012).

1. Lattice-theoretic definition and basic forms

Let $\mathcal{A}$ be a finite central arrangement in a vector space $V$. Its intersection lattice is

$$L(\mathcal{A})=\left\{\bigcap_{H\in\mathcal{B}}H \mid \mathcal{B}\subseteq \mathcal{A}\right\},$$

ordered by reverse inclusion, with rank function $r(X)=\operatorname{codim}_V(X)$. An element $X\in L(\mathcal{A})$ is modular if, for every $Y\in L(\mathcal{A})$,

$r(X)+r(Y)=r(X\vee Y)+r(X\wedge Y),$

or equivalently, in the geometric-lattice formulation used for arrangements, if $X+Y\in L(\mathcal{A})$ for all $Y\in L(\mathcal{A})$ (Liang et al., 2024, Hoge et al., 2012). The arrangement is supersolvable when there exists a maximal chain

$\mathcal{A}$0

of modular elements, where $\mathcal{A}$1 and $\mathcal{A}$2 (Hoge et al., 2014).

A useful equivalent description, due to Björner–Edelman–Ziegler and used in the root-ideal setting, replaces the modular chain by a partition

$\mathcal{A}$3

such that $\mathcal{A}$4 has rank $\mathcal{A}$5 and no rank-2 flat of $\mathcal{A}$6 is contained in $\mathcal{A}$7 (Hultman, 2014). This block formulation is particularly effective in rank $\mathcal{A}$8, in root systems, and in simplicial settings.

Low-rank behavior is rigid. Every arrangement of rank at most $\mathcal{A}$9 is supersolvable (Hoge et al., 2012, Hoge et al., 2014). For rank $V$0, supersolvability is equivalent to the existence of a modular rank-2 element (Hoge et al., 2012). In the projective-plane language of line arrangements, this becomes the existence of a modular point: an intersection point $V$1 such that, for every other intersection point $V$2, the line $V$3 belongs to the arrangement (Dimca et al., 2017, Hanumanthu et al., 2019).

2. Algebraic consequences, nice partitions, and hierarchy

Stanley’s classical factorization theorem gives the most immediate invariant-theoretic signature of supersolvability: if $V$4 is supersolvable of rank $V$5, then its characteristic polynomial splits as

$V$6

where each $V$7 is a nonnegative integer (Liang et al., 2024). In the free case these integers coincide with the exponents, and for supersolvable arrangements they can be read from the modular chain (Hoge et al., 2012).

Supersolvability also sits inside Terao’s theory of factorizations of the Orlik–Solomon algebra. A partition $V$8 is nice if it is independent and every localization $V$9 contains a singleton block of the induced partition. Terao’s theorem identifies niceness with a tensor factorization

$$L(\mathcal{A})=\left\{\bigcap_{H\in\mathcal{B}}H \mid \mathcal{B}\subseteq \mathcal{A}\right\},$$0

and consequently

$$L(\mathcal{A})=\left\{\bigcap_{H\in\mathcal{B}}H \mid \mathcal{B}\subseteq \mathcal{A}\right\},$$1

Supersolvable arrangements are necessarily nice, and the modular-chain partition constructed from

$$L(\mathcal{A})=\left\{\bigcap_{H\in\mathcal{B}}H \mid \mathcal{B}\subseteq \mathcal{A}\right\},$$2

is a canonical nice partition (Hoge et al., 2014). The converse fails in general: nice does not imply supersolvable. However, if a partition is induced by a maximal chain, then it is nice if and only if that chain is modular, so in that restricted sense niceness recovers supersolvability (Liang et al., 2024).

Hoge–Röhrle sharpened the structural position of supersolvable arrangements by proving

$$L(\mathcal{A})=\left\{\bigcap_{H\in\mathcal{B}}H \mid \mathcal{B}\subseteq \mathcal{A}\right\},$$3

with the converses failing in general (Hoge et al., 2014). Combined with the classical equivalence between supersolvable and fiber-type arrangements, this places supersolvability at the intersection of lattice theory, OS-algebra factorization, and recursive freeness (Hoge et al., 2012).

3. Graph-derived arrangements

For a simple graph $$L(\mathcal{A})=\left\{\bigcap_{H\in\mathcal{B}}H \mid \mathcal{B}\subseteq \mathcal{A}\right\},$$4, the graphical arrangement is

$$L(\mathcal{A})=\left\{\bigcap_{H\in\mathcal{B}}H \mid \mathcal{B}\subseteq \mathcal{A}\right\},$$5

Stanley’s theorem identifies supersolvability exactly with graph chordality: $$L(\mathcal{A})=\left\{\bigcap_{H\in\mathcal{B}}H \mid \mathcal{B}\subseteq \mathcal{A}\right\},$$6 is supersolvable if and only if $$L(\mathcal{A})=\left\{\bigcap_{H\in\mathcal{B}}H \mid \mathcal{B}\subseteq \mathcal{A}\right\},$$7 is chordal, equivalently if and only if $$L(\mathcal{A})=\left\{\bigcap_{H\in\mathcal{B}}H \mid \mathcal{B}\subseteq \mathcal{A}\right\},$$8 has a perfect elimination ordering (Liang et al., 2024). This remains one of the clearest combinatorial characterizations of supersolvability in arrangement theory.

Recent work tightened the graphical picture further. A graphical arrangement has a nice partition if and only if the graph is chordal, and for chordal graphs every nice partition is induced by a maximal modular chain of the supersolvable lattice. Thus, in the graphical case,

$$L(\mathcal{A})=\left\{\bigcap_{H\in\mathcal{B}}H \mid \mathcal{B}\subseteq \mathcal{A}\right\},$$9

and niceness ceases to be strictly more general than supersolvability (Liang et al., 2024).

Several graph-derived generalizations preserve this pattern with modified combinatorics. For $r(X)=\operatorname{codim}_V(X)$0-graphical arrangements, supersolvability is characterized by the existence of a vertex elimination order $r(X)=\operatorname{codim}_V(X)$1 such that each $r(X)=\operatorname{codim}_V(X)$2 is simplicial in the earlier induced subgraph and, whenever $r(X)=\operatorname{codim}_V(X)$3 and $r(X)=\operatorname{codim}_V(X)$4 is adjacent to $r(X)=\operatorname{codim}_V(X)$5, one has $r(X)=\operatorname{codim}_V(X)$6 (Mu et al., 2015). For the $r(X)=\operatorname{codim}_V(X)$7-Ish arrangements, the cone $r(X)=\operatorname{codim}_V(X)$8 is supersolvable if and only if $r(X)=\operatorname{codim}_V(X)$9 is a nest; in this class supersolvability, inductive freeness, and freeness are equivalent (Abe et al., 2014).

These graph-derived examples show that supersolvability is not merely a lattice abstraction. It is often the exact categorical translation of elimination-type combinatorics.

4. Reflection, root-ideal, and simplicial classifications

Family-wise classification results make supersolvability unusually explicit in several highly structured settings. The following criteria are established in the literature cited here (Hoge et al., 2012, 1311.0620, Hultman, 2014, Cuntz et al., 2017).

Family	Supersolvability criterion	Consequence
Reflection arrangements $X\in L(\mathcal{A})$0	Every irreducible factor has rank $X\in L(\mathcal{A})$1, or is type $X\in L(\mathcal{A})$2, $X\in L(\mathcal{A})$3, or $X\in L(\mathcal{A})$4 with $X\in L(\mathcal{A})$5	Irreducible case is equivalent to existence of a modular rank-2 element
Restrictions of reflection arrangements	Beyond restrictions of supersolvable reflection arrangements, only $X\in L(\mathcal{A})$6 and the sporadic $X\in L(\mathcal{A})$7 occur up to isomorphism	For irreducible restrictions of rank $X\in L(\mathcal{A})$8, supersolvability is equivalent to a modular 1-dimensional element
Root ideal arrangements $X\in L(\mathcal{A})$9	$Y\in L(\mathcal{A})$0 is chain peelable	Supersolvability is preserved under subideals; $Y\in L(\mathcal{A})$1 is Koszul iff $Y\in L(\mathcal{A})$2 is supersolvable
Irreducible supersolvable simplicial arrangements	Rank $Y\in L(\mathcal{A})$3: $Y\in L(\mathcal{A})$4 or $Y\in L(\mathcal{A})$5; rank $Y\in L(\mathcal{A})$6: $Y\in L(\mathcal{A})$7, $Y\in L(\mathcal{A})$8, or $Y\in L(\mathcal{A})$9	In rank $r(X)+r(Y)=r(X\vee Y)+r(X\wedge Y),$0, irreducible supersolvable simplicial arrangements are crystallographic

For reflection arrangements, a particularly striking simplification occurs: for irreducible $r(X)+r(Y)=r(X\vee Y)+r(X\wedge Y),$1, supersolvability is equivalent to the existence of a modular rank-2 element in $r(X)+r(Y)=r(X\vee Y)+r(X\wedge Y),$2 (Hoge et al., 2012). For restrictions of reflection arrangements, the analogous detection mechanism drops to dimension $r(X)+r(Y)=r(X\vee Y)+r(X\wedge Y),$3: an irreducible restriction is supersolvable if and only if its lattice contains a modular 1-dimensional element (1311.0620).

Root ideal arrangements exhibit a different but equally sharp combinatorial criterion. Hultman proved that $r(X)+r(Y)=r(X\vee Y)+r(X\wedge Y),$4 is supersolvable if and only if the root-poset ideal $r(X)+r(Y)=r(X\vee Y)+r(X\wedge Y),$5 is chain peelable; the only minimal obstructions are the $r(X)+r(Y)=r(X\vee Y)+r(X\wedge Y),$6 star ideal and the $r(X)+r(Y)=r(X\vee Y)+r(X\wedge Y),$7 height-4 ideal (Hultman, 2014). In this class, the Orlik–Solomon algebra is Koszul exactly in the supersolvable cases, giving a precise positive instance of the general “Koszul implies supersolvable?” problem (Hultman, 2014).

5. Rank-three and projective line arrangements

In $r(X)+r(Y)=r(X\vee Y)+r(X\wedge Y),$8, supersolvability becomes especially concrete. A line arrangement is supersolvable if and only if it has a modular point (Dimca et al., 2017, Hanumanthu et al., 2019). This specialization has generated a detailed rank-three theory linking modularity to Jacobian syzygies, splitting types, and enumerative constraints.

If $r(X)+r(Y)=r(X\vee Y)+r(X\wedge Y),$9 is a supersolvable line arrangement of $X+Y\in L(\mathcal{A})$0 lines with modular point $X+Y\in L(\mathcal{A})$1, and $X+Y\in L(\mathcal{A})$2 is the minimal degree of a Jacobian syzygy, then

$X+Y\in L(\mathcal{A})$3

hence

$X+Y\in L(\mathcal{A})$4

In particular, for supersolvable line arrangements the subtle invariant $X+Y\in L(\mathcal{A})$5 is determined combinatorially by the multiplicity of a modular point (Dimca et al., 2017).

Complex supersolvable line arrangements are strongly constrained. A nontrivial complex line arrangement cannot have more than $X+Y\in L(\mathcal{A})$6 modular points, and if all crossing points have multiplicity $X+Y\in L(\mathcal{A})$7 or $X+Y\in L(\mathcal{A})$8, then the arrangement has no modular points and therefore is not supersolvable (Hanumanthu et al., 2019). For $X+Y\in L(\mathcal{A})$9-homogeneous supersolvable line arrangements, where every modular point has the same multiplicity $Y\in L(\mathcal{A})$0, one has the sharp upper bound

$Y\in L(\mathcal{A})$1

with equality realized by the full monomial arrangement

$Y\in L(\mathcal{A})$2

(Abe et al., 2019). When there are at least two modular points, these arrangements are classified up to lattice-isotopy by explicit subarrangements $Y\in L(\mathcal{A})$3 of the full monomial arrangement, and their complements and Milnor fibers are combinatorially rigid in the sense of lattice-isotopy invariance (Abe et al., 2019).

The double-point problem is another rank-three theme. For real non-pencil supersolvable line arrangements with $Y\in L(\mathcal{A})$4 lines and a modular point of multiplicity $Y\in L(\mathcal{A})$5,

$Y\in L(\mathcal{A})$6

where $Y\in L(\mathcal{A})$7 denotes the number of double points (Hanumanthu et al., 2019). In the complex $Y\in L(\mathcal{A})$8-homogeneous case, the conjectural lower bound $Y\in L(\mathcal{A})$9 is proved in several large regimes, including $\mathcal{A}$00, $\mathcal{A}$01, and all cases with at least two modular points (Abe et al., 2019).

A nearby notion, nearly supersolvability, underscores how robust the modular paradigm is in rank $\mathcal{A}$02: a nearly supersolvable line arrangement is not supersolvable but has a nearly modular point, and such arrangements are always either free or nearly free (Dimca et al., 2017).

6. Abelian, toric, and recent developments

Supersolvability extends beyond linear arrangements. For abelian arrangements in $\mathcal{A}$03, Bibby–Cohen–Delucchi define supersolvability via a chain of $\mathcal{A}$04-ideals in the poset of layers, and strict supersolvability via $\mathcal{A}$05-ideals (Bibby et al., 30 Sep 2025). In the toric case, the complement sits atop a tower of fiber bundles, and the monodromy of each supersolvable bundle factors through an Artin braid group; in the strictly supersolvable case it factors further through the pure braid group. This stronger factorization determines invariants such as the cohomology ring and the lower central series Lie algebra of the fundamental group (Bibby et al., 30 Sep 2025).

Cones over Dirichlet arrangements provide a different bridge between combinatorics and homological algebra. If $\mathcal{A}$06 is obtained from the underlying graph by adding edges between all boundary vertices, then for the cone $\mathcal{A}$07 the following are equivalent: $\mathcal{A}$08 (Lutz, 2018). This places Dirichlet arrangements among the small number of classes where Koszulness of the Orlik–Solomon algebra is known to be exactly equivalent to supersolvability.

A recent combinatorial-topological development is that the tope graphs of all supersolvable hyperplane arrangements, and even supersolvable oriented matroids, admit Hamiltonian cycles. The proof is constructive and uses the inductive decomposition theorem of Björner–Edelman–Ziegler: the regions above a supersolvable rank-$\mathcal{A}$09 subarrangement form path fibers, and an alternating traversal of these fibers lifts a Hamiltonian cycle inductively (Körber et al., 20 Aug 2025). This situates supersolvability not only as an algebraic and lattice-theoretic condition, but also as a source of strong Gray-code-type structure in region graphs.

Supersolvable arrangements therefore occupy a rare position in arrangement theory. They are defined purely combinatorially, yet they control freeness, OS-algebra factorization, fiber-type topology, braid-monodromy factorization, and in several major families admit exact classification. The principal subtlety is that many nearby properties—niceness, freeness, Koszulness—coincide with supersolvability only in special classes; outside those classes the converse implications fail or remain open (Hoge et al., 2014, Lutz, 2018).