---
title: Symmetric Strong Circuit Elimination (SSCE)
url: https://www.emergentmind.com/topics/symmetric-strong-circuit-elimination-property-ssce
type: topic
---

# Symmetric Strong Circuit Elimination (SSCE)

Searching arXiv for the specified papers and topic.
arXiv search query: 1002.2357 OR 2508.00132 SSCE matroid circuit elimination
The **Symmetric Strong Circuit Elimination Property (SSCE)** is a strengthening of the usual strong circuit elimination axiom in matroid theory. In its standard asymmetric form, strong circuit elimination asserts that if \(C_1\) and \(C_2\) are circuits, \(e\in C_1\cap C_2\), and \(e_1\in C_1-C_2\), then there is a circuit \(C_3\) such that \(e_1\in C_3\subseteq (C_1\cup C_2)-e\). SSCE replaces this one-sided preservation requirement with a symmetric one: whenever \(e_1\in C_1-C_2\), \(e_2\in C_2-C_1\), and \(e\in C_1\cap C_2\), there must exist a circuit \(C_3'\) with \(\{e_1,e_2\}\subseteq C_3' \subseteq (C_1\cup C_2)-e\) [2508.00132]. The property is not universal for matroids; rather, it characterizes a special class of connected matroids, and its study has led to both a structural classification and a modified circuit axiom system. A related but distinct development shows that, in both matroids and oriented matroids, elimination need only be imposed on **modular pairs** of circuits to recover the full theory [1002.2357].

## 1. Definition and relation to strong circuit elimination

The standard strong circuit elimination property recalled in the literature takes the following form: if \(C_1\) and \(C_2\) are circuits in a matroid \(M\), with
\[
e_1\in C_1-C_2 \quad\text{and}\quad e\in C_1\cap C_2,
\]
then \(M\) has a circuit \(C_3\) such that
\[
e_1\in C_3\subseteq (C_1\cup C_2)-e.
\]
This form is inherently asymmetric because it protects only a chosen element from \(C_1-C_2\) [2508.00132].

SSCE is defined by strengthening the conclusion symmetrically. A matroid \(M\) has the symmetric strong circuit elimination property if, whenever \(C_1\) and \(C_2\) are circuits and
\[
e_1\in C_1-C_2,\qquad e_2\in C_2-C_1,\qquad e\in C_1\cap C_2,
\]
there is a circuit \(C_3'\) such that
\[
\{e_1,e_2\}\subseteq C_3' \subseteq (C_1\cup C_2)-e
\]
[2508.00132].

The distinction is substantive rather than terminological. Ordinary strong elimination guarantees preservation of one designated element from one side of the overlap; SSCE requires simultaneous preservation of one designated element from each side. This makes SSCE strictly stronger than ordinary strong elimination. The difference is not formal only: there exist matroids in which ordinary elimination holds, as it must, but SSCE fails [2508.00132].

## 2. Structural characterization in connected matroids

For connected matroids, SSCE admits an exact structural characterization. The central theorem states that for a connected matroid \(M\), the following are equivalent:

1. \(M\) has the symmetric strong circuit elimination property;
2. \(M\) has no pair of skew circuits;
3. for all integers \(k\) and \(l\) exceeding two, \(M\) has no series minor isomorphic to
   \[
   S(U_{k-2,k},U_{l-2,l});
   \]
4. \(M^*\) is unbreakable [2508.00132].

This theorem identifies the precise obstruction to SSCE: the presence of two skew circuits. Sets \(X\) and \(Y\) in a matroid are **skew** if
\[
r(X)+r(Y)=r(X\cup Y).
\]
For circuits \(C_1\) and \(C_2\), this means that on their union the matroid restricts as a direct sum of the two circuit restrictions. The paper gives the explicit formulation that a matroid has \(k\) skew circuits if it has circuits \(\{C_1,\dots,C_k\}\) such that
\[
M\big| \left(\bigcup_{i=1}^k C_i\right)
=
(M|C_1)\oplus (M|C_2)\oplus\cdots\oplus (M|C_k)
\]
[2508.00132].

The equivalence with unbreakability of the dual places SSCE within an established connectivity framework. The relevant notion is that a matroid \(M\) is **unbreakable** if \(M\) is connected and \(M/F\) is connected for every flat \(F\) of \(M\) [2508.00132]. Combined with the cited result of Oxley–Pfeil, this yields
\[
\text{SSCE} \iff \text{no skew circuits} \iff M^* \text{ unbreakable}
\]
for connected \(M\) [2508.00132].

## 3. Skew circuits as the obstruction

The equivalence between SSCE and the absence of skew circuits is the conceptual core of the theory. If SSCE fails, then there are circuits \(C_1,C_2\) and elements
\[
e\in C_1\cap C_2,\qquad e_1\in C_1-C_2,\qquad e_2\in C_2-C_1
\]
such that no circuit \(D\) satisfies
\[
\{e_1,e_2\}\subseteq D\subseteq (C_1\cup C_2)-e.
\]
Ordinary strong elimination can still be applied separately to obtain circuits \(D_1,D_2\) with
\[
e_1\in D_1\subseteq (C_1\cup C_2)-e,\qquad e_2\in D_2\subseteq (C_1\cup C_2)-e.
\]
If \(D_1\) and \(D_2\) are not skew, then \(M|(D_1\cup D_2)\) is connected, so there is a circuit \(D\subseteq D_1\cup D_2\) containing \(\{e_1,e_2\}\), contradicting the assumed failure. Therefore failure of SSCE forces the one-sided elimination circuits to form a skew pair [2508.00132].

This explains why skewness is exactly the missing condition needed to upgrade ordinary strong elimination to symmetric strong elimination. Ordinary strong elimination only requires enough internal interaction to preserve one chosen element. SSCE requires enough interaction between both noncommon parts of the original circuits to preserve one chosen element from each side after eliminating a shared element. When two circuits are skew, that coupling is absent: they behave rank-theoretically like separate summands on their union [2508.00132].

A plausible implication is that SSCE should be regarded less as an alternative axiom for arbitrary matroids than as a circuit-theoretic regularity condition detecting a strong form of internal connectivity among intersecting circuits.

## 4. Forbidden series minors and examples

The forbidden series-minor characterization in the connected case is given by the infinite family
\[
S(U_{k-2,k},U_{l-2,l})\qquad\text{for all integers }k,l>2
\]
[2508.00132]. Here \(S(\cdot,\cdot)\) denotes a series connection. The class of matroids satisfying SSCE is closed under series minors, meaning minors obtained by deletions and series contractions [2508.00132].

The basic counterexample is \(N_5\), described as the series connection of two copies of \(U_{1,3}\):
\[
N_5 \cong S(U_{1,3},U_{1,3}).
\]
It is obtained from a \(3\)-circuit \(\{e_1,e_2,e\}\) by adding \(f_i\) in parallel to \(e_i\) for \(i=1,2\). Taking
\[
C_1=\{e_1,f_2,e\},\qquad C_2=\{e_2,f_1,e\},
\]
one has
\[
e_1\in C_1-C_2,\qquad e_2\in C_2-C_1,\qquad e\in C_1\cap C_2,
\]
but there is no circuit contained in
\[
(C_1\cup C_2)-e
\]
that contains \(\{e_1,e_2\}\). Thus \(N_5\) fails SSCE [2508.00132]. The same example exhibits the obstruction explicitly: \(\{e_1,f_1\}\) and \(\{e_2,f_2\}\) are skew circuits [2508.00132].

The class is not minor-closed under arbitrary contraction. The paper notes that \(M(K_4)\) has no pair of skew circuits and hence satisfies SSCE, but every single-element contraction of \(M(K_4)\) is isomorphic to \(N_5\), so SSCE fails to be preserved under general contraction [2508.00132]. This distinction between series-minor closure and minor closure is a central technical limitation of the property.

In the binary setting, the paper also proves that a connected binary matroid with three skew circuits has a series minor isomorphic to \(M(G)\), where \(G\) is one of the graphs \(L_1,\dots,L_5\) in Figure 1 [2508.00132]. This is presented as a related structural theorem about skew circuits rather than a direct reformulation of SSCE.

## 5. Modified symmetric elimination as a matroid axiom

Although SSCE itself is not valid in all matroids, a closely related symmetric elimination principle does yield a universal axiom system. Let \({\mathscr C}\) be the set of circuits of a matroid \(M\). The paper proves that \({\mathscr C}\) obeys the following condition:

> **(C3)\(^{\prime\prime}\)** Let \(C_1\) and \(C_2\) be members of \({\mathscr C}\) with
> \[
> e_1\in C_1-C_2,\qquad e_2\in C_2-C_1.
> \]
> If
> \[
> e\in C_1\cap C_2
> \]
> and
> \[
> (C_1-e_1)\cup (C_2-e_2)
> \]
> contains no member of \({\mathscr C}\), then \({\mathscr C}\) contains a member \(C_3\) such that
> \[
> \{e_1,e_2\}\subseteq C_3\subseteq (C_1\cup C_2)-e.
> \]
> Furthermore, \(C_3\) is the unique circuit of \(M\) contained in \((C_1\cup C_2)-e\) [2508.00132].

The added hypothesis is that
\[
(C_1-e_1)\cup (C_2-e_2)
\]
contains no circuit; equivalently, that union is independent in the stated sense [2508.00132]. This condition repairs the failure of naive SSCE and makes the symmetric conclusion universally valid.

The resulting cryptomorphism is that a collection \({\mathscr C}\) of nonempty pairwise incomparable subsets of a finite set \(E\) is the set of circuits of a matroid on \(E\) if and only if \({\mathscr C}\) satisfies **(C3)\(^{\prime\prime}\)** [2508.00132]. Thus the usual circuit axioms can be replaced by nonemptiness, pairwise incomparability, and this modified symmetric elimination condition.

The paper also records an essential limitation: one cannot weaken \((\mathrm{C3})''\) by merely asking \(e_1\in C_1\) and \(e_2\in C_2\), rather than requiring
\[
e_1\in C_1-C_2,\qquad e_2\in C_2-C_1.
\]
A counterexample comes from the cycle matroid of \(K_{2,3}\) [2508.00132]. This shows that the asymmetry in where the designated elements are drawn from remains structurally important even in the repaired symmetric axiom.

## 6. Relation to modular elimination and oriented matroids

A distinct line of work shows that elimination axioms can often be imposed only on **modular pairs** of circuits without loss of generality. For a family \(S\) of incomparable subsets, one forms
\[
U(S):=\{\bigcup T \mid T\subseteq S\},
\]
ordered by inclusion. Two members of \(S\) are a modular pair if they are a modular pair in the atomic lattice \(U(S)\); equivalently, their join has lower interval of length \(2\) [1002.2357].

In the matroid setting, if \(C\) is a collection of incomparable finite subsets of a set \(E\), the paper defines the elimination property
\[
E(C_1,C_2,C): \textrm{ for all } e\in C_1\cap C_2 \textrm{ there is } C_3\in C\textrm{ with }C_3\subseteq (C_1\cup C_2)\setminus \{e\}.
\]
Its main theorem states:

> Let \(C\) be a collection of incomparable finite subsets of a set \(E\). If \(E(A,B,C)\) for all modular pairs \(A,B\in C\), then \(E(A,B,C)\) for all pairs \(A,B\in C\) [1002.2357].

For finite \(E\), this yields a circuit characterization of matroids using elimination only on modular pairs [1002.2357]. The significance for SSCE is indirect but substantial. SSCE is a strengthening of circuit elimination, while modular-elimination theory shows that, in ordinary matroid axiomatization, one can drastically reduce the set of circuit pairs on which elimination must be checked.

The oriented-matroid analogue uses signed circuits \(X:E\to\{-1,0,+1\}\), support
\[
supp(X):=\{e\in E\mid X(e)\neq 0\},
\]
\(\mathbb Z_2\)-symmetry under \(X\mapsto -X\), and the strong oriented elimination axiom
\[
\begin{array}{rl}
\OOE (X,Y,C):& \textrm{for all } e,f \textrm{ with } X(e)=-Y(e)\neq 0, X(f)\neq Y(f)\ \textrm{there is } Z\in C \textrm{ with } Z(e)=0, Z(f)\neq 0,\\
& \textrm{and } Z(g)\in\{0,X(g),Y(g)\} \textrm{ for all }g\in E.
\end{array}
\]
The paper proves that it is enough to require \(\OOE(X,Y,C)\) only when \(supp(X)\) and \(supp(Y)\) form a modular pair [1002.2357].

The term “Symmetric Strong Circuit Elimination Property” does not appear there, but the comparison is informative. In oriented matroids, the elimination statement just displayed is already symmetric under swapping \(X\) and \(Y\) in the evident sense, whereas the 2025 work isolates a genuinely stronger symmetric requirement in the unoriented matroid setting: simultaneous preservation of distinguished elements from both sides of a circuit overlap [1002.2357; 2508.00132]. This suggests that “symmetry” has different technical roles in the two contexts—coordinatewise symmetry in oriented elimination versus bilateral preservation in SSCE.

## 7. Significance and scope

SSCE occupies an intermediate position between standard circuit axioms and specialized structural properties. It is not a cryptomorphic axiom for all matroids, because it fails in examples such as \(N_5\) [2508.00132]. At the same time, it is not merely a curiosity: in connected matroids it admits a complete structural description in terms of skew circuits, forbidden series minors, and dual unbreakability [2508.00132].

Its main significance lies in clarifying what additional circuit interaction is required beyond ordinary strong elimination. The characterization
\[
\boxed{ M \text{ has SSCE } \iff M \text{ has no pair of skew circuits } \iff M \text{ has no series minor } S(U_{k-2,k},U_{l-2,l})\ (k,l>2) \iff M^* \text{ is unbreakable} }
\]
for connected matroids gives a precise answer [2508.00132].

A second contribution is methodological. The modified symmetric elimination axiom \((\mathrm{C3})''\) shows that the failure of naive SSCE can be repaired by inserting the right independence hypothesis, yielding a full circuit cryptomorphism [2508.00132]. In parallel, modular-elimination theory shows that elimination conditions may often be localized to modular pairs without changing the resulting theory [1002.2357].

Taken together, these developments place SSCE within a broader program in matroid theory: refining elimination principles to identify which local circuit interactions are genuinely essential, which are redundant, and which capture special structural classes rather than the entire category of matroids [1002.2357; 2508.00132].

Source: https://www.emergentmind.com/topics/symmetric-strong-circuit-elimination-property-ssce