Subnormality of the quotients of $\mathbb T^d$-invariant Hilbert modules

This paper investigates the subnormality of quotients of $\mathbb T^d$-invariant Hilbert modules by homogeneous polynomials, establishing that such quotients are subnormal only if the polynomial is square-free and of degree at most one for standard spaces like the Hardy and Drury-Arveson modules, while also demonstrating that higher-degree examples exist for specific invariant modules like the Dirichlet module.

The Gist

Imagine you are a master architect working in a vast, multi-dimensional city called Complex Space. In this city, there are special buildings called Hilbert Modules. Think of these modules as giant, perfectly organized libraries filled with songs (mathematical functions).

The rules of this city are strict:

1. The Music: The songs are made of polynomials (like $z_1^2 + z_2$).
2. The Conductors: There are "conductors" (operators) who can shift the music around. If the music stays in perfect harmony when these conductors work, the library is called Subnormal. This is the "gold standard" of musical order.

The Big Question: The "Quotient" Problem

Now, imagine you want to build a smaller, specialized library inside one of these giant buildings. You do this by picking a specific rule (a polynomial, let's call it $p$) and throwing out every song that follows that rule.

• The Original Library: $H$ (The big, perfect library).
• The Rule: $p$ (e.g., "No songs where the first note equals the second note").
• The New Library: $H/[p]$ (The quotient). This is what's left after you remove all the songs that fit the rule.

The Mystery: If the original library was perfectly ordered (subnormal), does the new, smaller library keep that perfect order? Or does the act of cutting out the "rule-following" songs create chaos?

The authors of this paper are detectives trying to solve this mystery. They are asking: "Under what conditions does the new, smaller library remain perfectly ordered?"

---

The Detective's Findings (Simplified)

The paper explores different types of cities (domains) and different types of rules (polynomials). Here are their main discoveries, translated into everyday terms:

1. The "Square-Free" Rule

Imagine your rule is "No songs with the pattern $x^2$." If you try to remove songs based on a rule that has a "squared" part (like $x^2$), the resulting library usually falls apart.

• The Finding: For the new library to stay ordered, your rule ($p$) must be square-free.
• The Analogy: Think of a recipe. If you say "Remove all cakes with double chocolate," but your rule is actually "Remove cakes with chocolate squared," the math gets messy. The rule must be simple and unique (no repeated factors) to keep the structure stable.

2. The "Degree" Limit (The Size of the Rule)

The "degree" of a polynomial is like the complexity or length of the rule.

• Degree 1: A simple rule like "No songs where $z_1 = z_2$."
• Degree 2: A complex rule like "No songs where $z_1^2 = z_2$."

The Surprise:

• In some famous libraries (like the Hardy Space on a ball or the Drury-Arveson space), the new library is only ordered if the rule is very simple (Degree 1).
• The Metaphor: Imagine a high-security vault. If you try to change the lock with a complex, multi-toothed key (Degree 2), the vault breaks. It only works if you use a simple, single-tooth key (Degree 1).
• The Twist: The authors found a specific library (the Drury-Arveson module) where this is especially surprising. Even though the original library wasn't perfectly ordered to begin with, if you cut it with a simple rule, the result becomes perfectly ordered! It's like taking a slightly messy room, removing a specific messy corner, and suddenly the whole room looks spotless.

3. The "Tensor Product" Connection

One of the motivations for this research comes from a puzzle posed by a mathematician named N. Salinas.

• The Puzzle: If you take two perfect libraries and "glue" them together (a process called a Module Tensor Product), do you get a perfect library?
• The Connection: The authors realized that gluing two libraries together is mathematically the same as taking a giant library and cutting out the rule "$z_1 = z_2$."
• The Result: They proved that for certain types of libraries, gluing them together creates chaos unless the original libraries were already very specific. This solves a long-standing debate in the mathematical community.

4. The "Counter-Examples" (When Things Go Wrong)

The paper also shows that the rules aren't universal.

• The "Bad" Library: There are specific types of libraries (like the Dirichlet module) where even a simple rule (Degree 1) destroys the order.
• The "Weird" Library: In some very specific, custom-built libraries, you can actually use a complex rule (Degree 2) and still get a perfectly ordered result. This is like finding a house where you can knock down a load-bearing wall, and the house somehow stands up straighter than before!

---

Why Does This Matter?

You might ask, "Who cares about ordering libraries of songs?"

In the real world, this math helps us understand systems of equations and quantum mechanics.

• Operators: The "conductors" in our story represent physical forces or signals.
• Stability: Knowing when a system stays "subnormal" (stable and predictable) after we remove certain variables is crucial for engineering and physics.
• Geometry: It helps mathematicians understand the shape of multi-dimensional spaces.

The Takeaway

The paper is a map of a mathematical landscape. It tells us:

1. Simplicity is key: Usually, to keep a system stable after making changes, the changes must be simple (Degree 1).
2. Context matters: What works for one type of library (Hardy space) might fail for another (Dirichlet space).
3. Surprises exist: Sometimes, breaking a system (removing a part) actually makes it more stable, defying our intuition.

The authors have essentially drawn a boundary line: "Here is where the order holds, and here is where it breaks." And in doing so, they've solved a puzzle that had mathematicians scratching their heads for decades.

Technical Summary

Here is a detailed technical summary of the paper "Subnormality of the Quotients of $T_d$-Invariant Hilbert Modules" by K. S. Amritha, S. Bera, S. Chavan, and S. S. Sequeira.

1. Problem Statement and Motivation

The paper investigates the subnormality of quotient modules derived from $T_d$-invariant Hilbert modules over the polynomial ring $\mathbb{C}[z_1, \dots, z_d]$.

• Context: In multivariable operator theory, a Hilbert module $H$ is defined by a commuting $d$-tuple of operators $T = (T_1, \dots, T_d)$. A submodule $M$ is a closed subspace invariant under $T$. The quotient module $H/M$ is defined on the orthogonal complement $H \ominus M$ with the compressed operators $T_j|_{H \ominus M}$.
• Core Question: Given a $T_d$-invariant Hilbert module $H_\kappa$ and a principal homogeneous submodule $[p]$ generated by a homogeneous polynomial $p$, under what conditions is the quotient module $H_\kappa/[p]$ subnormal?
• Motivation: This problem is deeply connected to a question posed by N. Salinas regarding the subnormality of module tensor products. Specifically, if $p(z_1, z_2) = z_1 - z_2$, the subnormality of the quotient $\widehat{H}_{\kappa_1} \otimes H_{\kappa_2} / [p]$ is equivalent to the subnormality of the module tensor product $H_{\kappa_1} \otimes_{\mathbb{C}[z]} H_{\kappa_2}$. The authors aim to classify when these quotients are subnormal, particularly for classical spaces like the Hardy space ($H^2$) and the Drury–Arveson space ($H^2_d$).

2. Methodology and Framework

The authors employ a blend of operator theory, complex analysis, and algebraic geometry.

• Definitions:
  • $T_d$-invariant Hilbert Module: A reproducing kernel Hilbert space $H_\kappa$ on a Reinhardt domain $\Omega$ where the kernel $\kappa$ is invariant under the torus action ($\kappa(\zeta \cdot z, \zeta \cdot w) = \kappa(z, w)$).
  • Submodule $[p]$: The closed linear span of $\{z^\alpha p : \alpha \in \mathbb{Z}_+^d\}$.
  • Square-free Polynomials: Polynomials with no repeated irreducible factors.
• Key Tools:
  • Stieltjes Moment Problem: The authors utilize the characterization of subnormal operators via moment sequences. A commuting tuple is subnormal if and only if its moments correspond to a positive Borel measure.
  • $m$-isometries: The paper leverages the property that multiplication operators on certain spaces (like $H^2_d$) are $m$-isometries. This imposes strict constraints on the spectrum and the degree of polynomials defining subnormal quotients.
  • Algebraic Decomposition: A crucial step involves decomposing homogeneous polynomials in two variables into linear factors (Proposition A), a technique that fails in dimensions $d \geq 3$.
  • Spectral Analysis: The authors analyze the Taylor spectrum of the compressed operators and use properties of normal extensions to derive contradictions for non-subnormal cases.

3. Key Contributions and Results

The paper provides a comprehensive classification of subnormal quotient modules for specific domains and modules.

A. General Necessary Conditions

• Square-free Condition (Proposition 2.3): If $H_\kappa/[p]$ is subnormal, then the polynomial $p$ must be square-free. If $p$ has a repeated factor, the quotient cannot be subnormal.
• Degree Constraints: For specific spaces, the degree of $p$ is strictly bounded.

B. Main Theorems for Specific Spaces

The authors establish complete characterizations for the Hardy space on the bidisc ($H^2(\mathbb{D}^2)$), the Hardy space on the unit ball ($H^2(\mathbb{B}^2)$), and the Drury–Arveson space ($H^2_2$).

Theorem 2.1 (Hardy Space on Bidisc):
For $H^2(\mathbb{D}^2)$ and a nonconstant homogeneous $p \in \mathbb{C}[z_1, z_2]$, the following are equivalent:

1. $H^2(\mathbb{D}^2)/[p]$ is subnormal.
2. $\deg p = 1$.
3. The compressed operators satisfy a linear dependence: $a T_{z_1} = b T_{z_2}$ for some $a, b \in \mathbb{C}$ (not both zero).

Theorem 2.2 (Hardy Space on Ball & Drury–Arveson Space):
For $H = H^2(\mathbb{B}^2)$ or $H = H^2_2$ (Drury–Arveson), the same equivalence holds:

1. $H/[p]$ is subnormal.
2. $\deg p = 1$.
3. $a T_{z_1} = b T_{z_2}$.

Significance of Theorem 2.2:

• Surprising Result for $H^2_2$: The Drury–Arveson module $H^2_d$ is not subnormal for $d \geq 2$. However, the paper proves that its quotient by a linear polynomial is subnormal. This highlights that subnormality can emerge in quotients even when the parent module lacks it.
• Degree Bound: The proof relies on the fact that the multiplication tuple on these spaces is an $m$-isometry (specifically, a 1-isometry for $H^2(\mathbb{B}^2)$ and a $d$-isometry for $H^2_d$). This forces $\deg p \leq m$.

C. Counterexamples and Limitations

The paper demonstrates that these results are sensitive to the underlying geometry and invariance properties:

• Failure for General $T_2$-invariant Modules: There exist $T_2$-invariant Hilbert modules where $H_\kappa/[p]$ is subnormal even if $\deg p = 2$ (Example 5.2).
• Failure for $U_2$-invariant Modules: While Theorem 2.2 holds for $H^2(\mathbb{B}^2)$, the implication "$\deg p = 1 \implies$ subnormal" fails for general $U_2$-invariant modules (Example 5.3 shows a case where $\deg p=2$ yields a subnormal quotient).
• Dirichlet Module: The quotient of the Dirichlet module $D^2(\mathbb{B}^2)$ by a linear polynomial is not subnormal (Example 5.4), showing that the result does not hold for all weighted spaces.

D. Structural Results

• Principal Submodules: The authors prove that for $T_2$-invariant modules and square-free $p$, the submodule generated by $p$ coincides with the ideal of functions vanishing on the zero set of $p$ (i.e., $[p] = [p]_0$).
• Tensor Product Connection: The paper rigorously establishes the link between the subnormality of the quotient by $z_1 - z_2$ and the subnormality of the module tensor product, providing a negative answer to Salinas's question in the general case (as the tensor product is not always subnormal).

4. Significance and Impact

1. Classification: The paper provides a definitive classification for the subnormality of quotients in the most prominent function spaces (Hardy and Drury–Arveson), showing that linearity of the defining polynomial is a necessary and sufficient condition in these contexts.
2. Resolution of Salinas's Problem: It clarifies the relationship between algebraic constructions (module tensor products) and analytic properties (subnormality), showing that the tensor product of subnormal modules is not necessarily subnormal.
3. Dimensional Sensitivity: The results highlight a sharp distinction between $d=2$ and $d \geq 3$. While the classification is complete for $d=2$, the paper notes that for $d \geq 3$, the subnormality of $H^2(\mathbb{D}^d)/[p]$ for linear $p$ remains an open problem.
4. Methodological Innovation: The use of $m$-isometry properties to bound the degree of polynomials in subnormal quotients offers a new technique for analyzing operator tuples in several complex variables.

5. Conclusion

The paper concludes that for classical Hilbert modules on the bidisc and unit ball, the subnormality of a quotient module $H/[p]$ is extremely restrictive: it occurs if and only if the defining homogeneous polynomial $p$ is of degree 1. This contrasts with more general $T_d$-invariant modules where higher-degree polynomials can yield subnormal quotients. The work bridges commutative algebra (square-free polynomials, factorization) with operator theory (subnormality, moment problems), providing a robust framework for future investigations into multivariable operator theory.