Ascent and descent of operators concerning Jordan products

arXiv:2607.12724 · math.FA, math.OA · Submitted 2026-07-14 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Mathematics. The day in mathematics: new theorems, new proofs, and what they open.

Ada: We're looking at the paper: "Ascent and descent of operators concerning Jordan products".

Title and authors: Ada: Let's talk about the title and who wrote this paper, "Ascent and descent of operators concerning Jordan products." The authors are Xiaofei Qi, Xiaoshuo Feng, and Bing Xu Jinchuan Hou. It’s a heavy title because it immediately signals that the main tool they are using to analyze these operators is the ascent and descent of their Jordan products.

Emmy: I think the power here lies in linking these analytic properties—ascent and descent—to algebraic structure through that product operation. It suggests that this product isn't just some arbitrary way to multiply operators; it’s a key feature that reveals something deep about the operator itself.

Kolya: From a probabilistic view, I wonder if this structural characterization helps us understand the decay rates of certain systems governed by these operators? Does knowing the ascent tells us anything about how fast things stabilize or converge in some application?

Ada: Well, they are looking at very specific structures. For instance, they characterize the general preservers phi on B(X) that preserve this ascent or descent property for Jordan products. They found that these maps must either take a very specific form involving an invertible operator T and some map f, or the space X itself has to be finite-dimensional.

Emmy: That second part about finite dimensionality is important because it shows that the general machinery they build works robustly across different spaces, but when you move to infinite dimensions, you have to deal with these specific functional forms involving the operator T.

The paper's summary: Ada: So, summarizing what this paper does: they first explore how ascent and descent behave for upper triangular block matrices and certain special algebraic operators. Then they establish characterizations for rank one and rank two operators by looking at their ascent or descent values.

Emmy: The core idea is using the Jordan product to characterize general preservers phi on B(X) that keep the ascent or descent of products consistent, meaning asc of phi(A)phi(B) plus phi(B)phi(A) equals asc of AB plus BA for all operators A and B.

Kolya: That consistency condition sounds like something you’d use when you have a system where the order in which you apply two operations doesn't matter, or at least not in a way that changes the stability metric. Does this relate to any known bounds on convergence?

Ada: It connects to stability because if an operator preserves this property, it has to be related back to these specific forms involving operators T and maps f. This essentially means that the structure of these preserving maps is very constrained by the ascent or descent behavior of the Jordan product.

Emmy: They also characterize equivalence relations between rank one nilpotent operators using their ascent of the Jordan product, showing how you can distinguish between different types of rank one nilpotent operators based on this single number.

The paper's improvements: Ada: The paper suggests a few improvements to the existing ideas. For example, they provide a characterization for maps preserving the ascent or descent of Jordan products that is more concrete than just saying it has to take one of two forms. They show that if X is infinite-dimensional, those forms involving T and f are very specific descriptions of what these maps look like.

Emmy: They also refine how they characterize rank one operators by looking at the ascent of Ax tensor f plus x tensor A star f, showing a clear distinction based on whether ax or a starf is zero. This gives you a more granular way to categorize those operators than just saying "rank one."

Kolya: If we're talking about applications, does this refinement mean we can get tighter bounds in something like dynamical systems where the underlying dynamics are described by these operators? Or are these structural results mostly theoretical tools?

Ada: It seems mostly structural for now. The results help us understand the rigidity of the algebraic structure of B(X) itself. They show that if an operator isn't algebraic, there must be a rank one operator T such that the ascent of AT plus TA is at least four, which contradicts the assumption that it’s not algebraic.

Emmy: That contradiction argument is powerful because it shows how rigid the ascent and descent features are when they constrain an operator to be algebraic. It proves that these stability metrics impose strong restrictions on what kind of operators can exist in B(X).

Conclusion: Ada: So, to wrap up the paper "Ascent and descent of operators concerning Jordan products," we see that the ascent and descent of Jordan products provide a powerful way to characterize special operators like rank one and rank two ones. They also give us precise structural forms for maps that preserve these properties, linking them to invertible operators T.

Emmy: Essentially, the paper shows that if you want a map to respect how Jordan products behave in terms of ascent or descent, it has to follow very specific patterns involving linear transformations on the space X.

Kolya: For me, what this means is that if we're modeling complex interactions, these structural constraints imposed by ascent and descent might limit the possible ways those interactions can evolve over time. It suggests that not every evolution path governed by an operator is allowed; there are strong structural rules governing it.

Ada: That’s a good way to put it. We've seen how the paper uses these analytic properties to define algebraic boundaries, and how they constrain the set of maps that can exist within B(X). This work opens up new avenues for understanding the deep structure of operator algebras.

XIAOFEI QI, XIAOSHUO FENG, BING XU JINCHUAN HOU

Shanxi University · Taiyuan University of Technology

math.FA, math.OA

Submitted: 2026-07-14

Updated: 2026-11-18

Comments: 39 Pages

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 73/100

The gist: Let B(X) be the algebra of all bounded linear operators on a real or complex Banach space X with dim X ≥ 3…

Key concepts

Ascent (asc(T))
The ascent of an operator measures how many times you need to apply it before the kernel stabilizes. It is defined as the smallest integer n such that ker(T^n) equals ker(T^(n+1)). For rank-one operators, this value is often 1 or 2.
Descent (desc(T))
The descent of an operator measures how many times you need to apply it before the range stabilizes. It is defined as the smallest integer n such that ran(T^n) equals ran(T^(n+1)). For rank-one operators, this value is also typically 1 or 2.
Jordan Product Ascent
This concept relates to the ascent of the Jordan product of two rank-one operators. The paper shows that for these products, the ascent can be 1, 2, or 3. This property is used to characterize operators like A=0 and A in F¹I.
Map Preserving Ascent/Descent
This concept studies functions phi: B(X) → B(X) that keep the ascent or descent values of Jordan products unchanged. The theorem provides specific forms for such maps, showing they are related to invertible operators T and specific functional forms f.

Terminology

Summary

Let B(X) be the algebra of all bounded linear operators on a real or complex Banach space X with dim X ≥ 3, and this paper explores characterizations for certain special operators by the ascent (descent) of Jordan products. p. 0

Where the paper says it

Let B(X ) be the algebra of all bounded linear operators on a real or complex Banach space X with dim X ≥ 3. In this paper, we first explore the ascent (descent) of upper triangular block operator matrices and certain special algebraic operators, and then establish characterizations for the ascent (descent) of rank-one and rank-two operators.

Fundamental Definitions and Properties

The ascent of an operator T is defined as asc(T) = inf<n ∈ N ∪ 0: ker(T n) = ker(T(n+1)) and the descent is defined as desc(T) = inf<n ∈ N ∪ 0: ran(T n) = ran(T(n+1)). Key properties include that for a rank one operator A, asc(A) = desc(A) ∈ 1, 2, and moreover A ∈ N1(X) ⇔ asc(A) = 2 ⇔ desc(A) = 2. p. 1

Where the paper says it

(6) If A is a k-nilpotent operator, then asc(A) = desc(A) = k. (7) If A = x ⊗ f is of rank one, then asc(A) = desc(A) ∈ {1, 2}; and moreover, A ∈ N1(X ) ⇔ asc(A) = 2 ⇔ desc(A) = 2.

Furthermore, for any rank-one idempotent operators A1, A2 in N1(X), the equivalence of their similarity is characterized by the ascent of their Jordan product. p. 0
Where the paper says it

Based on these results, we characterize features for some special operators by the ascent (descent) of Jordan products. As an application, we give the structure of all maps with range containing all bounded operators of rank at most three preserving the ascent (descent) of operator Jordan product on B(X ).

Ascent and Descent of Block Operators

The ascent and descent properties are studied for upper triangular block operator matrices S = AB0C! in B(X ⊕Y). Proposition 2. p. 1

Where the paper says it

The following well-known properties concerning ascent and decent are useful. Proposition 1.

1 states that if asc(A) = n, then asc(S) ≥ n, and if C = 0, then n ≤ asc(S) ≤ n + 1. p. 0
Where the paper says it

For any operator T ∈ B(X ), ker(T) and ran(T) stand for the kernel and the range of T, respectively. Recall that the ascent and the descent of T are defined respectively by [21] asc(T) = inf{n ∈ N ∪ {0}: ker(T n ) = ker(T n+1)} and desc(T) = inf{n ∈ N ∪ {0}: ran(T n ) = ran(T n+1)}.

Similarly, for the descent of upper triangular operator matrices S = AB0C! p. 3
Where the paper says it

It follows that asc(S) ≤ n + 1. □ Similarly, for the descent of upper triangular operator matirx, we have Proposition 2.

, Proposition 2. p. 1
Where the paper says it

The following well-known properties concerning ascent and decent are useful. Proposition 1.

2 states that if desc(A) = n, then desc(S) ≥ n, and if C = 0, then n ≤ desc(S) ≤ n + 1. p. 0
Where the paper says it

For any operator T ∈ B(X ), ker(T) and ran(T) stand for the kernel and the range of T, respectively. Recall that the ascent and the descent of T are defined respectively by [21] asc(T) = inf{n ∈ N ∪ {0}: ker(T n ) = ker(T n+1)} and desc(T) = inf{n ∈ N ∪ {0}: ran(T n ) = ran(T n+1)}.

Characterization of Rank-One and Rank-Two Operators

For operators with rank two, Proposition 3.1 establishes that asc(A) ∈ 1, 2, or 3. p. 9

Where the paper says it

Thus, combining the foregoing discussions, the result concerning the ascent holds. Symmetrically, an analogous argument establishes that the proposition is also valid for the descent.

Specifically, asc(A) = 1 if and only if rank(A 2) = 2. p. 1
Where the paper says it

Specifically, we first explore the ascent (descent) of upper triangular block operator matrices and certain special algebraic operators (Section 2). Subsequently, we establish characterizations for the ascent (descent) of rank-one and rank-two operators (Section 3).

The ascent of the Jordan product of two rank-one operators A = x ⊗ f and B = y ⊗ g is characterized by Proposition 3. p. 0
Where the paper says it

In this paper, we first explore the ascent (descent) of upper triangular block operator matrices and certain special algebraic operators, and then establish characterizations for the ascent (descent) of rank-one and rank-two operators. Based on these results, we characterize features for some special operators by the ascent (descent) of Jordan products.

4, which states that asc(AB + BA) ∈ 1, 2, or 3. p. 0
Where the paper says it

For any operator T ∈ B(X ), ker(T) and ran(T) stand for the kernel and the range of T, respectively. Recall that the ascent and the descent of T are defined respectively by [21] asc(T) = inf{n ∈ N ∪ {0}: ker(T n ) = ker(T n+1)} and desc(T) = inf{n ∈ N ∪ {0}: ran(T n ) = ran(T n+1)}.

Characterization of Operators via Jordan Products

The paper provides characterizations for operators based on the ascent (descent) of Jordan products. Proposition 4. p. 1

Where the paper says it

The following well-known properties concerning ascent and decent are useful. Proposition 1.

p. 1
Where the paper says it

The following well-known properties concerning ascent and decent are useful. Proposition 1.

1 states that A = 0 if and only if asc(AN + NA) = 1 for all N ∈ N1(X). p. 0
Where the paper says it

For any operator T ∈ B(X ), ker(T) and ran(T) stand for the kernel and the range of T, respectively. Recall that the ascent and the descent of T are defined respectively by [21] asc(T) = inf{n ∈ N ∪ {0}: ker(T n ) = ker(T n+1)} and desc(T) = inf{n ∈ N ∪ {0}: ran(T n ) = ran(T n+1)}.

Similarly, A ∈ F1I is equivalent to asc(AN + NA) = 2 for all N ∈ N1(X). p. 3
Where the paper says it

It follows that asc(S) ≤ n + 1. □ Similarly, for the descent of upper triangular operator matirx, we have Proposition 2.

Characterization of Maps Preserving Jordan Products

Theorem 5.1 provides a characterization for maps ϕ: B(X) → B(X) with range containing F3(X) that preserve the ascent (descent) of Jordan products. p. 0

Where the paper says it

Based on these results, we characterize features for some special operators by the ascent (descent) of Jordan products. As an application, we give the structure of all maps with range containing all bounded operators of rank at most three preserving the ascent (descent) of operator Jordan product on B(X ).

The map satisfies this if and only if one of two conditions holds: either there exists an invertible bounded linear or conjugate-linear operator T: X → X and a map f: B(X) → F1 such that ϕ(A) = f(A)T A T−1 for all A ∈ B(X), or the space X is finite-dimensional, in which case there exist an invertible matrix T ∈ Mn(F), an automorphism τ: F → F and a map f: Mn(F) → F1 such that either ϕ(A) = f(A)T[τ (aij)]T−1 or ϕ(A) = f(A)T[τ (aij)]trT−1 for all A = [aij] ∈ Mn(F).

Application to Rank-One Nilpotent Operators

The paper also characterizes the equivalence relation ∼ between rank-one nilpotent operators by the ascent of their Jordan product. Proposition 4.5 shows that A1 ∼ A2 if and only if there exists a rank-one nilpotent operator B ∈ N1(X) linearly independent of both A1 and A2, such that for every T ∈ B(X), asc(T Ai + AiT) ∈ 2, 3 implies asc(T B + BT) ∈ 2, 3. p. 0

Where the paper says it

Every rank one operator can be written as A = x ⊗ f, defined by (x ⊗ f)y = f(y)x for any y ∈ X , for some vector x ∈ X and bounded linear functional f ∈ X ∗ . Clearly, A = x ⊗ f is nilpotent if and only if A2 = 0, and in turn, if and only if f(x) = ⟨x, f⟩ = 0.

This result is crucial for understanding the structure of rank-one nilpotent operators. p. 0
Where the paper says it

Clearly, A = x ⊗ f is nilpotent if and only if A2 = 0, and in turn, if and only if f(x) = ⟨x, f⟩ = 0. Denote by F1(X ) the set of all rank one operators on X .

Characterization of Maps Preserving Ascent/Descent

The final part characterizes maps preserving the ascent (descent) of Jordan products. Theorem 5.1 concludes that the map satisfies this if and only if it has one of the forms described in Claim 6, which involves either an invertible operator T: X → X and a map f: N1(X) → F1 such that ϕ(N) = f(N)T NT−1 or X is reflexive, leading to a form involving T: X∗ → X and a map f: N1(X) → F1 such that ϕ(N) = f(N)T N∗T−1, depending on the dimension of X. p. 1

Where the paper says it

Assume that A, B are any two operator algebras and ⋄ is some product of elements A, B in A. We say that a map ϕ : A → B preserves the ascent (decent) of A⋄B if ϕ satisfies asc(A⋄B) = asc(ϕ(A)⋄ϕ(B)) (desc(A⋄B) = desc(ϕ(A)⋄ϕ(B))) for all A, B ∈ A.

Pruning redundant reasoning tokens

The paper proves that if A is not an algebraic operator, then there exists a rank-one operator T such that asc(AT + TA) ≥ 4, leading to a contradiction. This argument is used to establish the algebraic nature of operators. p. 0

Where the paper says it

Let B(X ) be the algebra of all bounded linear operators on a real or complex Banach space X with dim X ≥ 3. In this paper, we first explore the ascent (descent) of upper triangular block operator matrices and certain special algebraic operators, and then establish characterizations for the ascent (descent) of rank-one and rank-two operators.

Where the gains come from

The proof that A ∈ F1I requires showing that if there exists a nonzero vector x0 such that x0 and Ax0 are linearly independent, then by choosing a suitable rank-one nilpotent operator N = x0 ⊗ f0, one obtains asc(AN + NA) ∈ 2 or 3, contradicting the assumption. This contradiction is central to proving A = 0. p. 4

Where the paper says it

Otherwise, as Ta 0 0 −2a ! 2 =   3a 2 −3a 3 0 −4a 3a 2 0 0 0 4a 2   one gets 3a 2 − aλ1 + λ2 = 0, −4a + 2λ1 = 0 and 4a 2 − 2aλ1 + λ2 = 0, which implies a = 0, a contradiction.

The proof of Theorem 5.1 relies on showing that if the condition holds, then ϕ(I) ∈ F1I by considering asc(S2) = asc(ϕ(S2)). p. 6

Where the paper says it

ASCENT AND DESCENT OF BOUNDED LINEAR OPERATORS 7 Proof of Proposition 2. We will construct suitable T such that asc(AT + T A) ̸= asc(AT + T A + ωT) by considering the following two cases.

This implies that all injective operators belong to the range of ϕ. p. 1
Where the paper says it

(2) asc(A) = asc(T −1AT) and desc(A) = desc(T −1AT) hold for all invertible operators T ∈ B(X ). (3) A is injective ⇔ asc(A) = 0 ⇔ asc(A2 ) = 0; A is surjective ⇔ desc(A) = 0 ⇔ desc(A2 ) = 0.

The proof that A is an algebraic operator relies on contradiction by considering a suitable operator T such that asc(AT + TA) ≥ n + 1, where n = sup T ∈ Fn(X) asc(AT + TA). This demonstrates the rigidity of the ascent/descent features in determining algebraic structure. p. 2

Where the paper says it

ASCENT AND DESCENT OF BOUNDED LINEAR OPERATORS 3 other related results, the reader can see [12, 13, 14, 17, 18, 22] and the references therein. These results reveal that the ascent as well as decent have some rigid features that determine the algebraic structure of the operator algebra B(X ).

The proof that A is an algebraic operator relies on contradiction by considering a suitable operator T such that asc(AT + TA) ≥ 4, derived from the structure of A. This shows how the ascent/descent features constrain the algebraic structure of B(X). p. 2

Where the paper says it

ASCENT AND DESCENT OF BOUNDED LINEAR OPERATORS 3 other related results, the reader can see [12, 13, 14, 17, 18, 22] and the references therein. These results reveal that the ascent as well as decent have some rigid features that determine the algebraic structure of the operator algebra B(X ).

p. 2
Where the paper says it

ASCENT AND DESCENT OF BOUNDED LINEAR OPERATORS 3 other related results, the reader can see [12, 13, 14, 17, 18, 22] and the references therein. These results reveal that the ascent as well as decent have some rigid features that determine the algebraic structure of the operator algebra B(X ).

The proof that A is an algebraic operator relies on contradiction by considering a suitable operator T such that asc(AT + TA) ≥ 4, derived from the structure of A. This shows how the ascent/descent features constrain the algebraic structure of B(X).

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved analysis of operator ascent/descent using block operator matrices allows for precise characterization of ascent/descent properties based on diagonal blocks, specifically showing that if asc(A) = n, then for a block upper triangular matrix S = A B 0 C!, asc(S) >= n and if C = 0, then n <= asc(S) <= n + 1.

  2. Enhanced capability to characterize the general preserver map phi on B(X) preserving the ascent (descent) of operator Jordan products by providing concrete forms for phi when X is infinite-dimensional, such as phi(A) = f(A)T AT-1 where T is an invertible bounded linear or conjugate-linear operator.

  3. Ability to determine if an operator A is a zero operator by checking the condition asc(AN + NA) = 1 for all N in N1(X), which leads to the conclusion that If A != 0, then there exists a nonzero vector x0 such that x0 and Ax0 are linearly independent.

  4. Precise classification of rank-one operators and their Jordan products is possible by evaluating the ascent of Ax ⊗ f + x ⊗ A∗f, distinguishing between cases where asc = 1 (if Ax != 0 and A∗f != 0) and asc = 2 (if either Ax = 0 or A∗f = 0).

  5. Characterization of rank-one nilpotent operators is improved by establishing the equivalence that N1 and N2 are linearly dependent if and only if phi(N1) and phi(N2) are linearly dependent, which is a key tool in verifying the properties of Jordan product preservers.

  6. Verification of operator algebraic properties can be done by checking the ascent of AT + TA for all injective operators T, where the result implies that A /∈ FI and A is an algebraic operator.

  7. Determination of whether an operator A is in F1(X) + FI relies on checking if there exists a nonzero vector x0 such that x0 and Ax0 are linearly independent, which leads to a contradiction if the ascent condition holds.

  8. Detailed structural analysis for rank-two operators allows for the determination of their ascent based on the similarity transformation of A, resulting in specific matrix forms like B1, B2, etc., where asc(B1) = 1 and asc(B2) = 2.

  9. Characterization of maps preserving the ascent/descent of Jordan products can be restricted to cases where X is finite-dimensional by providing concrete forms for phi(A), such as phi(A) = f(A)T[tau (aij)]T-1 or phi(A) = f(A)T[tau (aij)]trT-1.

Abstract

Let B(X) be the algebra of all bounded linear operators on a Banach space X over the real or complex field F with X at least 3. In this paper, we give several properties of the ascent (descent) of operators or Jordan products of operators in B(X) and use these properties to characterize certain operators such as algebraic operators, rank-one, rank-two operators, rank-one nilpotent operators as well as some relations between operators in terms of ascent (descent) of Jordan products. As an application, we give a characterization, up to a multiplier of nonlinear functional f: B(X) to F*, of ring isomorphisms of B(X) in the case X= infinity and Jordan isomorphisms of B(X) in the case X< infinity via the maps with range containing all operators of rank at most three that preserve the ascent of Jordan products of operators. A slightly different characterization in term of descent is also obtained.

Related papers