Intermediate Arrangements

• Intermediate Arrangements are combinatorial, geometric, or algebraic structures that lie strictly between two established classes, integrating key properties from both rigid systems and flexible pseudo-line variants.
• They employ methodologies such as approaching pseudo-line definitions, plus-one generated hyperplane criteria, and arithmetic interpolations to control combinatorial invariants and duality properties.
• Their framework facilitates practical advances in algorithmic applications, geometric configurations, and network routing by transitioning between rigid and general arrangement systems.

An intermediate arrangement is a combinatorial, geometric, or algebraic structure that lies strictly between two established classes, typically interpolating properties from both. In the context of arrangements theory, several canonical intermediate objects have been studied, including approaching pseudo-line arrangements, plus-one generated hyperplane arrangements, and arithmetic interpolations between Coxeter (reflection-type) arrangements. The theory of intermediate arrangements seeks to identify and analyze classes which inherit substantial structure from a more restrictive family, yet exhibit increased generality or complexity characteristic of the broader class.

1. Approaching Pseudo-Line Arrangements as an Intermediate Class

An arrangement of approaching pseudo-lines in the Euclidean plane comprises $n$ bi-infinite, $x$-monotone curves $\ell_i$, each realized as the graph of a function $f_i:\mathbb{R}\to\mathbb{R}$, such that for any $i<j$ the function $x\mapsto f_j(x)-f_i(x)$ is monotonically decreasing and surjective. This definition ensures that for each pair, the corresponding curves "approach," cross once, and then "recede"—mirroring the simple crossing structure of classical line arrangements but allowing more flexibility than strict linearity (Felsner et al., 2020).

Crucially, the class of approaching pseudo-line arrangements is strictly intermediate: all classical line arrangements are included, but not all pseudo-line arrangements (general $x$-monotone curves crossing once) admit such an "approaching" representation. For example, the existence of non-realizable allowable sequences, such as those containing Asinowski's forbidden six-line suballowable, demonstrates strict containment.

2. Line-Like and Pseudo-Line-Like Properties

Approaching arrangements serve as a prototypical intermediate class by combining "line-like" geometric features with the combinatorial diversity of general pseudo-lines.

Line-like properties:

• Duality and generalized configurations: Each arrangement has a point–line dual configuration which remains within the approaching class; duality arguments traditionally restricted to lines extend to these arrangements (Felsner et al., 2020).
• Flip-graph connectivity: The flip graph, with nodes corresponding to arrangements and edges to local triangle flips, is connected—mirroring the property for line arrangements and enabling pathway transformations between any two instances.
• Bichromatic triangulation: In every nontrivial bichromatic coloring (i.e., red/blue coloring with at least one monochromatic crossing of each), there exists an empty triangle with at least one boundary from each color, paralleling Sylvester–Gallai theorems for lines.

Pseudo-line-like features:

• Enumeration: The number of isomorphism classes for simple approaching arrangements on $n$ elements is $2^{\Theta(n^2)}$, matching the exponential count for arbitrary pseudo-lines, yet greatly exceeding the $2^{\Theta(n\log n)}$ bound for straight lines.
• Allowable-sequence realizability: Deciding whether a combinatorial allowable sequence can be realized by an approaching arrangement is solvable in polynomial time ($\mathrm{P}$), in contrast to the $\exists\mathbb{R}$-hardness for straight lines. The constraints are encoded as a linear program with $O(n^3)$ inequalities.

These properties are summarized in the following table:

Property	Approaching Pseudo-lines	Straight Lines	General Pseudo-lines
Flip-graph connectivity	Yes	Yes	Yes
Duality w/ point configs	Yes	Yes	No
Enumeration growth	$2^{\Theta(n^2)}$	$2^{\Theta(n\log n)}$	$2^{\Theta(n^2)}$
Allowable sequence $\in$P	Yes	No ($\exists\mathbb{R}$-hard)	Yes

3. Intermediate Arrangements in Hyperplane Theory: Plus-One Generated Arrangements

In hyperplane arrangement theory, a structurally intermediate class is provided by plus-one generated arrangements. Given an essential arrangement $\mathcal{A}$ of rank $\ell$ in a vector space over a field $\mathbb{K}$, its logarithmic derivation module $D(\mathcal{A})$ is typically not free unless $\mathcal{A}$ itself is "free". A plus-one generated arrangement is one for which $D(\mathcal{A})$ admits a minimal free $S$-resolution with exactly one extra generator and a single linear syzygy. Such arrangements are generated by $\ell+1$ derivations connected through a linear relation involving a nonzero linear form (Abe, 2018).

Key facts include:

• Every deletion of a hyperplane from a free arrangement yields either another free arrangement or a strictly plus-one generated arrangement.
• In rank 3, free plane arrangements are surrounded (in the deletion-addition poset) exclusively by free and strictly plus-one generated arrangements.

This positions plus-one generated arrangements as the canonical next-to-free class, providing tight algebraic and combinatorial control, and making them the natural intermediate family for the algebraic theory of arrangements.

4. Arithmetic Interpolation: Intermediate Arrangements in Reflection/Coxeter Theory

The type D reflection arrangement and its arithmetic restrictions yield a further canonical example of intermediate arrangements. The rank-$n$ type D arrangement $\mathcal{D}_n$ comprises hyperplanes $x_i\pm x_j=0$, and by intersecting with $s$ coordinate hyperplanes $x_k=0$, one obtains restrictions $\mathcal{D}_{n,s}$. These interpolate between type D ($s=0$) and type B ($s=n$) (Degen et al., 16 Nov 2025).

For these intermediate arrangements, the $h$-polynomial and Chow polynomial invariants behave in a strictly arithmetic way for fixed rank $n$ and $s\in\{0,\ldots,n\}$: $$h_{\mathcal{D}_{n,s}}(t) = (1-\frac{s}{n})h_{\mathcal{D}_n}(t) + \frac{s}{n}h_{\mathcal{B}_n}(t),$$

$$\mathrm{Ch}_{\mathcal{D}_{n,s}}(x) = (1-\frac{s}{n})\mathrm{Ch}_{\mathcal{D}_n}(x)+\frac{s}{n}\mathrm{Ch}_{\mathcal{B}_n}(x).$$

This convex-linear interpolation results in $\gamma$-vector and Chow-polynomial components that vary exactly linearly in $s$, interpolating chamber enumeration and intersection-theoretic invariants linearly between the endpoints.

5. Intermediate Arrangements in Algorithmic and Combinatorial Contexts

Intermediate arrangements also arise algorithmically, such as in the manipulation of multistage interconnection networks using fundamental arrangements. Here, a fundamental arrangement (FA) is an explicit sequence of $N-1$ perfect matchings for $N$ channels (with $N$ even), such that every unordered pair is realized exactly once. Transitioning any initial permutation to any target in at most $2(N-1)$ stages is possible by first routing to the FA and then from FA to the target, with each stage affecting only one unique pair (Gur et al., 2010). The FA thus acts as an intermediate permutation arrangement providing structural control for rearrangeable nonblocking routing, facilitating efficient network algorithms.

6. Structural and Theoretical Role of Intermediate Arrangements

Intermediate arrangements serve as testbeds for conjectures and theorems aiming to distinguish the boundary separating rigid geometric/algebraic families (lines, free arrangements, Coxeter systems) from maximally flexible or combinatorially general classes (pseudo-lines, arbitrary arrangements, matroids). Any theorem that holds for an intermediate arrangement but fails for the most general class, while remaining nontrivial over the intermediate one, is a candidate for isolating the essential properties inherent to the geometric notion at hand (Felsner et al., 2020, Abe, 2018, Degen et al., 16 Nov 2025).

These classes arise naturally: approaching pseudo-lines as a monotonicity interpolation, plus-one generated arrangements as a single-step enlargement of freeness, and arithmetic restrictions as convex combinations in Coxeter theory. Each provides deep insight into local-to-global behavior and the algebraic/combinatorial invariants of arrangements, and exposes the minimal critical obstructions lying between rigid and wild cases.