On Notions of Expansivity for Operators on Locally Convex Spaces

Imagine you are watching a dance troupe perform on a stage. In the world of mathematics, this "stage" is a space where points (dancers) live, and the "dance moves" are performed by an operator (a rule that tells every dancer where to go next).

This paper is about a specific type of dance called Expansivity.

The Core Concept: The "Stretching" Dance

In a normal, calm dance, if two dancers start very close together, they might stay close, or drift apart slowly. But in an expansive dance, the rule is simple: No matter how close two dancers start, they will eventually be forced far apart.

The Rule: If you pick any two distinct dancers, there is a moment in the dance where the distance between them becomes larger than a specific "safety zone" (let's call it $c$ ).
The Goal: The paper asks: How can we tell if a dance is expansive just by looking at the choreography (the weights and the space)?

The authors are taking this idea, which was previously studied on simple, rigid stages (Banach spaces), and moving it to Locally Convex Spaces. Think of these as more flexible, complex stages where the rules of distance are a bit more fluid and can change depending on which "lens" you look through.

The Three Types of "Stretching"

The paper explores three different ways a dance can be considered "expansive":

Standard Expansivity (The "One Big Jump"):
- Analogy: Imagine two dancers start inches apart. At some point in the future (or even the past), they must jump so far apart that they are clearly separated.
- The Math: There is some time $n$ where the distance is huge.
Uniform Expansivity (The "Guaranteed Explosion"):
- Analogy: This is stricter. It's not just that some dancers jump apart; it's that every pair of dancers, no matter where they start, will be forced apart at a guaranteed rate. It's like a firework that always explodes with a specific minimum power, never fizzling out.
- The Math: The separation happens "uniformly" across the whole stage.
Average Expansivity (The "Long-Term Drift"):
- Analogy: This is the paper's main new contribution. Imagine two dancers who might stay close for a while, then drift apart, then come back together, then drift apart again. They don't necessarily stay far apart forever. However, if you take the average distance between them over a very long time, that average must be huge.
- The Math: Even if they hug occasionally, the average of their distances over time must blow up to infinity.

The Main Discoveries

1. The Weighted Shift (The Conveyor Belt)

The authors focus on a specific type of dance called a Weighted Shift.

Analogy: Imagine a conveyor belt with numbered spots. A dancer at spot 5 moves to spot 4, and a dancer at spot 4 moves to spot 3. But here's the twist: every time they move, they get multiplied by a "weight" (a number). If the weight is 2, they double in size (or distance from the center). If the weight is 0.5, they shrink.
The Question: How do we choose these weights so that the dance is expansive?

The Big Finding:
The authors created a "recipe" (a mathematical formula) to determine exactly which weights will make the dance expansive.

For Average Expansivity (the new concept), they found that if the weights grow fast enough on average, the dancers will drift apart over time, even if they occasionally get close.
They applied this to Fréchet sequence spaces and Köthe sequence spaces (which are just fancy names for specific types of infinite lists of numbers).

2. The Surprise: Chaos and Mixing

Usually, in math, if a system is "expansive" (dancers fly apart), it's hard for it to be "chaotic" (dancers mixing everywhere unpredictably) or "transitive" (dancers visiting every corner of the stage).

The Old Rule: Uniformly expansive dances were thought to be too orderly to be chaotic.
The New Discovery: The authors found a special dance (a weighted shift) that is Average Expansive AND Chaotic AND Mixing.
The Metaphor: It's like a dance where the dancers usually drift apart (satisfying the average rule), but they do it in such a wild, unpredictable way that they visit every corner of the room and never repeat a pattern. This was a surprise because it breaks the old intuition that "spreading out" and "mixing" are enemies.

3. The "Slow Growth" Surprise

In simple spaces (like standard Euclidean space), if a dance is uniformly expansive, the dancers must fly apart exponentially fast (like a virus doubling every hour).

The New Discovery: On these flexible, complex stages (Locally Convex Spaces), the authors found a dance that is Uniformly Expansive but the dancers only grow apart polynomially (like $n^2$ or $n^3$ ).
The Metaphor: Imagine a rubber band. In the old world, it snaps apart instantly. In this new world, it stretches slowly and steadily, but it still stretches forever and never lets the dancers get close again. This shows that the "flexible stage" allows for much more subtle types of expansion.

Why Does This Matter?

This paper is like upgrading the rulebook for a complex video game.

New Mechanics: It introduces "Average Expansivity," a new way to measure how systems separate.
New Levels: It solves puzzles about how these systems behave on complex, infinite stages (Köthe spaces).
Breaking Myths: It proves that you can have a system that is both "spreading out" and "chaotic" at the same time, and that "spreading out" doesn't always have to be explosive; it can be a slow, steady drift.

In short, the authors are mapping out the geography of how things move and separate in the most flexible mathematical worlds, showing us that the rules of "getting far apart" are much more interesting and varied than we previously thought.

Here is a detailed technical summary of the paper "On Notions of Expansivity for Operators on Locally Convex Spaces" by Bernardes Jr., Martínez-Giménez, and Rodenas.

1. Problem Statement and Context

The paper addresses the extension and characterization of expansivity concepts for linear operators within the framework of locally convex spaces (LCS), moving beyond the classical setting of Banach spaces.

Background: Expansivity is a fundamental concept in dynamical systems, originally defined for homeomorphisms on metric spaces. In linear dynamics, it characterizes operators where distinct points separate under iteration. Previous works (e.g., Eisenberg, Hedlund, Mazur, Bernardes et al.) established characterizations for Banach spaces, including uniform expansivity and average expansivity.
The Gap: While Bernardes et al. [5] recently extended standard expansivity and uniform expansivity to LCS, several critical questions remained open:
1. The concept of average expansivity had not been generalized to arbitrary locally convex spaces.
2. A complete characterization of uniformly expansive weighted shifts on Fréchet sequence spaces was missing (specifically, Problem E in [5] remained open).
3. The relationship between average expansivity and chaotic dynamics (Li-Yorke, distributional chaos) in the LCS setting was not fully explored, particularly regarding whether average expansivity precludes topological transitivity (unlike uniform expansivity).

2. Methodology

The authors employ a combination of functional analysis techniques and constructive examples:

Generalization of Definitions: They define average expansivity for operators on arbitrary LCS using families of seminorms, ensuring the definition is topological (independent of the specific choice of seminorms).
Conjugacy Arguments: To handle weighted shifts, they utilize isomorphisms to transform weighted shifts into unweighted shifts (or shifts with different weights) on isomorphic spaces, allowing them to transfer results from unweighted cases to weighted ones.
Köthe Sequence Spaces: They focus heavily on Köthe sequence spaces ( $\lambda_p(A, J)$ ), a rich class of Fréchet spaces, to derive concrete characterizations. They analyze the behavior of the weights and the Köthe matrix entries.
Constructive Counterexamples: To demonstrate the divergence between Banach space intuition and LCS behavior, they construct specific weighted shifts with carefully designed block structures (using sequences of blocks $A_j, B_j, C_j$ ) to control growth rates and density properties.
Direct Sums and Restrictions: They utilize properties of direct sums and restrictions to generalize results from individual components to product spaces.

3. Key Contributions and Results

A. Extension of Average Expansivity (Section 2)

Definition: The authors define an operator $T \in GL(X)$ on an LCS $X$ as average expansive if for every non-zero $x$ , there exists a seminorm $\|\cdot\|_\alpha$ such that the Cesàro mean of the norms diverges:
$\sup_{n \in \mathbb{N}} \left( \frac{1}{2n+1} \sum_{j=-n}^{n} \|T^j x\|_\alpha \right) = \infty$
Hierarchy: They establish the implication chain: Uniform Expansivity $\implies$ Average Expansivity $\implies$ Expansivity.
Characterization on Fréchet Sequence Spaces:
- Theorem 4: Provides a complete characterization for bilateral weighted backward shifts ( $B_w$ ) on Fréchet sequence spaces with a canonical basis. $B_w$ is average expansive if and only if the weighted averages of the norms of the canonical basis vectors diverge in at least one direction (forward or backward).
- Corollary 5: Applies this to Köthe sequence spaces, giving explicit conditions involving the Köthe matrix entries $a_{j,k}$ and the weights $w_j$ .

B. Uniform Expansivity on Köthe Spaces (Section 3)

Partial Solution to Problem E: The authors provide a complete characterization of uniformly expansive weighted forward shifts on Köthe sequence spaces (Theorem 11).
- The condition involves the existence of indices $\ell$ for every $k$ such that specific infimum limits of the ratio of matrix entries and weight products tend to infinity.
- Three mutually exclusive cases (A, B, C) are identified, corresponding to expansion occurring in the positive direction, negative direction, or both (splitting the space).
Polynomial vs. Exponential Growth (Example 14):
- A significant finding is that in LCS, uniform expansivity does not imply exponential growth of trajectories.
- They construct a uniformly expansive operator on the space of rapidly decreasing sequences $s(\mathbb{Z})$ where trajectories grow only polynomially ( $O(n^k)$ ), contrasting with the exponential growth required in Banach spaces.

C. Dynamics and Chaos (Section 2 & 3)

Topological Transitivity: The authors prove that unlike uniformly expansive operators (which are never topologically transitive), average expansive operators can be topologically transitive.
Distributional Chaos (Theorem 9): They construct a bounded weighted shift on classical spaces ( $c_0(\mathbb{Z})$ $c_{0} (Z)$ or $\ell_p(\mathbb{Z})$ $ℓ_{p} (Z)$ ) that is:
1. Topologically transitive (hypercyclic).
2. Average positively expansive.
3. Distributionally chaotic (admitting distributionally irregular vectors).
- This demonstrates that average expansivity is compatible with strong chaotic behavior, a property not shared by uniform expansivity.

D. General Properties (Section 4)

The paper establishes that expansivity, average expansivity, and uniform expansivity are preserved under:
- Inversion ( $T \iff T^{-1}$ ).
- Rotation by scalars on the unit circle ( $T \iff \lambda T$ ).
- Powers ( $T \iff T^k$ ).
- Restrictions to invariant subspaces.
- Topological conjugacy.
- Direct sums and product operators.

4. Significance and Impact

Theoretical Advancement: This work significantly broadens the scope of linear dynamics from Banach spaces to the more general and complex setting of locally convex spaces. It clarifies that many "expansive" behaviors in LCS are qualitatively different from those in normed spaces (e.g., the lack of exponential growth requirement).
Resolution of Open Problems: It solves the open problem regarding uniform expansivity for weighted shifts on Köthe spaces and introduces the missing concept of average expansivity in this setting.
Dynamical Complexity: By showing that average expansivity allows for topological transitivity and distributional chaos, the paper refines the classification of linear dynamical systems. It suggests that "average" conditions are weaker and more flexible than "uniform" conditions, allowing for richer dynamical behaviors.
Methodological Utility: The techniques used (conjugacy, block constructions, and seminorm analysis) provide a toolkit for future research into other dynamical properties (like shadowing or mixing) in locally convex spaces.

In summary, the paper successfully generalizes the theory of expansivity to locally convex spaces, providing rigorous characterizations for weighted shifts and revealing that average expansivity permits a much wider range of dynamical behaviors, including chaos and transitivity, which are forbidden for uniform expansivity.