A note on invariant transversals for normal subgroups

This paper investigates the existence of invariant transversals for normal subgroups and provides counterexamples to a conjecture in the specific case where the subgroup is abelian and the group is finite.

The Gist

Here is an explanation of the paper "A Note on Invariant Transversals for Normal Subgroups" using simple language and creative analogies.

The Big Picture: A Puzzle About Grouping

Imagine you have a giant, chaotic dance floor (this is your Group $G$). On this floor, there is a special, smaller group of dancers who always move in perfect unison (this is your Normal Subgroup $H$).

The mathematician's job is to organize the whole dance floor. They want to pick exactly one representative from every possible "dance formation" (called a coset) to create a master list. This list is called a Transversal.

The Twist: The mathematician wants this list to be "invariant." This means if you rotate the entire dance floor or change the perspective (a process called conjugation), the list of representatives should still look the same. It shouldn't matter who is looking at the dance floor; the list of representatives remains stable.

The Old Belief (The Conjecture)

For a long time, mathematicians believed a specific rule about these lists:

The Rule: If you can find a stable list of representatives for a group of dancers ($H$), then that group of dancers must be "pure." In math terms, they shouldn't be tangled up with the "chaos" or "noise" of the whole dance floor (the commutator subgroup, $G'$).

Think of it like this: If you can organize the dancers perfectly without them getting mixed up with the chaotic crowd, then the dancers must be very simple and quiet (Abelian) and not part of the messy, twisting movements of the whole room.

The author of this paper, Gerhard Hiss, set out to prove this rule was true. He thought, "If I can prove this, it will make classifying all these dance patterns much easier."

The Plot Twist: The Counterexamples

Instead of proving the rule, Hiss found exceptions. He discovered specific dance floors where:

1. The special group of dancers ($H$) is tangled up with the chaos of the room ($H \cap G' \neq \{1\}$).
2. Yet, you can still find a perfect, stable list of representatives.

The Analogy: Imagine a group of dancers who are actually part of the chaotic, twisting crowd. According to the old rule, you shouldn't be able to pick a stable list for them. But Hiss found a way to pick the list anyway. The old rule was wrong.

How Did He Find This? (The Detective Work)

Hiss didn't just guess; he used a mix of logic and computer power.

1. The Logic Check: He broke the problem down into smaller pieces. He realized that if the special dancers ($H$) are in the very center of the room (the center of the group), the problem becomes easier to analyze. He found a mathematical "recipe" (Proposition 2.3) to check if a stable list exists.
2. The Computer Search: He used a powerful math software called GAP (think of it as a super-fast calculator for group theory). He told the computer to check thousands of small dance floors.
   • He found that for small groups (fewer than 128 dancers), the old rule actually held up.
   • But once he got to groups of size 27 (with specific properties) and size 64, the computer found the "glitches"—the counterexamples.

The "Heavy Hitters" (Non-Solvable Groups)

The most exciting part of the paper is the discovery of two very complex, "non-solvable" dance floors (based on the group $PSL_3(4)$).

• These are like massive, intricate machines where the gears are locked together in a way that can't be taken apart easily.
• Hiss found that even in these incredibly complex, messy machines, there is a small, specific part ($H$) that is tangled with the mess, yet you can still organize a stable list for it.
• This proves that the old rule isn't just wrong for small, simple groups; it's wrong for the most complex groups in existence too.

Why Does This Matter?

You might ask, "Who cares about dance floors and lists?"

In mathematics, when you have a "Conjecture" (a guess that seems true), it acts like a map. If the map says "All roads lead to Rome," you can plan your journey based on that.

• Before this paper: Mathematicians thought, "If we see a stable list, we know the group is simple." This simplified their work.
• After this paper: The map has a hole in it. Mathematicians now know, "Wait, sometimes the group is messy even if the list is stable."

This forces mathematicians to update their theories. It's like realizing that not all birds can fly; some penguins are birds, but they swim. You have to adjust your understanding of what a "bird" (or in this case, a group with an invariant transversal) really is.

Summary

Gerhard Hiss tried to prove a rule about organizing groups of numbers. He expected the rule to be true. Instead, he used logic and computers to find specific cases where the rule fails. He showed that you can have a perfectly organized list even when the group is deeply tangled with chaos. This discovery breaks a long-held belief and forces the math world to rewrite its rules for these specific types of groups.

Technical Summary

Here is a detailed technical summary of the paper "A Note on Invariant Transversals for Normal Subgroups" by Gerhard Hiss.

1. Problem Statement

The paper investigates the existence of invariant transversals for normal subgroups within group theory. Specifically, it addresses Conjecture 1.1, proposed by the author and previously appearing in Ortjohann's Master's thesis.

• Definitions:
  • Let $G$ be a group and $H \leq G$ a subgroup. A transversal for the right cosets $H\backslash G$ is a set of representatives.
  • A subset $S \subseteq G$ is $L$-invariant if it is a union of $L$-conjugacy classes (invariant under conjugation by $L$).
  • A $G$-invariant transversal for $H\backslash G$ is a transversal that is invariant under conjugation by the entire group $G$.
• The Conjecture:

  Let $G$ be a finite group and $H \leq G$ be an abelian subgroup admitting a $G$-invariant transversal. Then $H \cap G' = \{1\}$, where $G'$ is the commutator subgroup of $G$.

The conjecture suggested that if an abelian normal subgroup admits a $G$-invariant transversal, it must have a trivial intersection with the derived subgroup. This would have simplified the classification of such pairs $(G, H)$. Previous work verified this for small groups (order $\leq 40$) and specific cases (e.g., Hall subgroups), but the general case remained open.

2. Methodology

Hiss employs a combination of structural group theory, character theory, and computational verification to disprove the conjecture.

• Structural Characterization (Section 2):

  • The author establishes necessary and sufficient conditions for the existence of a $G$-invariant transversal for a normal subgroup $H$.
  • Lemma 2.1 & Corollary 2.2: If $H$ is abelian and admits an $H$-invariant transversal, then $H \leq Z(G)$ (the center of $G$).
  • Proposition 2.3: A $G$-invariant transversal exists if and only if:
    1. $G = H C_G(H)$
    2. $C_{G/H}(Hg) = H C_G(g) / H$ for all $g \in G$.
  • Reduction: The problem is reduced to the case where $H$ is abelian and central ($H \leq Z(G)$). In this case, the existence of a transversal is equivalent to the condition that no non-trivial commutator of $G$ lies in $H$ (i.e., $H \cap G' = \{1\}$ is not strictly required if the central extension structure allows it).

• Connection to Cohomology and Character Theory (Section 3):

  • The problem is reframed using central extensions. If $H \leq Z(G)$, $G$ is a central extension of $\bar{G} = G/H$ by $H$.
  • A transversal $T$ defines a 2-cocycle $\gamma \in Z^2(\bar{G}, H)$.
  • The condition for the existence of an invariant transversal translates to the condition that every element in $\bar{G}$ is $\alpha$-regular (where $\alpha$ is a character of $H$ composed with the cocycle).
  • The author leverages Reynolds' work [13] and Blau's results [2] regarding the number of irreducible characters and the regularity of elements in central extensions. Specifically, if the number of irreducible characters of $G$ restricting to a specific character of $H$ equals the number of irreducible characters of $\bar{G}$, the transversal exists.

• Computational Verification:

  • The author uses the GAP (Groups, Algorithms, Programming) system to search for counterexamples among finite groups of small orders.
  • The search utilizes the character-theoretic condition $k(G) = k(G/H) \cdot k(H)$ (where $k(L)$ is the number of conjugacy classes), which is equivalent to the existence of the transversal when $H \leq Z(G)$.

3. Key Contributions and Results

A. Theoretical Characterization

Hiss provides a complete characterization for the existence of invariant transversals for normal subgroups. He proves that for abelian $H$, the existence of a $G$-invariant transversal is equivalent to specific conditions on the centralizer and the commutator subgroup, effectively linking the geometric property (transversal) to the algebraic structure (commutators and central extensions).

B. Disproof of Conjecture 1.1

The primary contribution is the construction of counterexamples to Conjecture 1.1, demonstrating that an abelian normal subgroup $H$ can admit a $G$-invariant transversal even if $H \cap G' \neq \{1\}$.

The paper identifies three distinct sources of counterexamples:

1. Metabelian Groups (Reynolds' Construction):
   • Citing Reynolds [13], Hiss notes the existence of metabelian groups of order $p^8$ with a central subgroup $H$ of order $p$ contained in $G'$, where the transversal condition holds.
2. Small $p$-Groups (Computational Discovery):
   • Using GAP, Hiss found 52 isomorphism classes of groups of order $27$($3^3$) containing a subgroup$H \leq Z(G) \cap G'$ of order 2.
   • These groups satisfy $k(G) = k(G/H) \cdot 2$, confirming the existence of the transversal despite $H \cap G' \neq \{1\}$.
   • Note: No counterexamples were found for groups of order $< 128$ with $H \leq Z(G)$, suggesting these phenomena appear at specific structural thresholds.
3. Non-Solvable Groups (Blau's Results):
   • The paper identifies two non-solvable counterexamples derived from covering groups of the projective special linear group $PSL_3(4)$.
   • Let $G_1$ and $G_2$ be covering groups of $PSL_3(4)$ with specific centers ($Z(G_1) \cong C_2 \times C_{12}$ and $Z(G_2) \cong C_2 \times C_4$).
   • In both cases, there exists a unique subgroup $H$ of order 2 in the center such that $H \cap G' \neq \{1\}$ (specifically, $H$ contains elements that are not commutators, yet the transversal exists).
   • These examples are significant because they show the conjecture fails even for quasisimple groups.

C. Reduction Lemma

Lemma 3.1 provides a "going down" result: if $H \leq Z(G)$ admits a $G$-invariant transversal and $Q \leq H \cap G'$ is a subgroup of prime order, then $Q$ admits a $P$-invariant transversal where $P$ is a Sylow $p$-subgroup of $G$. This allows the construction of counterexamples in $p$-groups from non-solvable ones.

4. Significance

• Refutation of a Conjecture: The paper definitively closes the question of Conjecture 1.1, showing it is false in general. This prevents future classification efforts from relying on the assumption that $H \cap G' = \{1\}$.
• Clarification of Loop Transversals: Since $G$-invariant transversals carry the structure of a loop, these results impact the understanding of which groups can be decomposed into specific loop structures.
• Interplay of Cohomology and Group Structure: The paper highlights the deep connection between the existence of invariant transversals, 2-cocycles, and the distribution of commutators within central extensions. It demonstrates that the condition "no non-trivial commutator lies in $H$" is sufficient but not necessary for the existence of such transversals.
• Computational Group Theory: The work serves as a case study for using computational tools (GAP) to verify theoretical conditions (like the conjugacy class count formula) and discover counterexamples in finite group theory.

In summary, Gerhard Hiss proves that the existence of a $G$-invariant transversal for an abelian normal subgroup $H$ does not imply that $H$ is disjoint from the commutator subgroup $G'$, providing explicit finite and infinite counterexamples across solvable and non-solvable groups.