pith. sign in

arxiv: 2508.06934 · v2 · submitted 2025-08-09 · 🧮 math.RA

Affine subspaces of units in simple algebras

Cl\'ement de Seguins Pazzis This is my paper

Pith reviewed 2026-05-19 00:38 UTC · model grok-4.3

classification 🧮 math.RA
keywords simple algebrasaffine subspacesunitscomposition algebrasquadratic formsHermitian formsoperator spacesAtkinson's theorem
0
0 comments X

The pith

In a simple algebra over a field, the largest affine subspace consisting entirely of units has a dimension fixed by the algebra's structure and known form classifications.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper determines the greatest possible dimension of an F-affine subspace lying inside the multiplicative group of units of a simple algebra A over a field F, under a mild cardinality assumption on F. This is recast as the problem of finding the largest linear subspace S such that every element x in S has x minus the identity as a unit. The maximal cases are described up to conjugation and reduce to the classification of associative composition algebras over F. When the characteristic is not 2, the analysis further relies on the known classifications of nonisotropic quadratic forms over F and nonisotropic Hermitian forms over quadratic and quaternionic extensions of F. The work also develops the theory of intransitive operator spaces, yielding a generalization of the dual form of Atkinson's theorem on primitive spaces of bounded-rank matrices.

Core claim

Under a mild cardinality assumption on the field F, the greatest possible dimension for an F-affine subspace of a simple algebra A contained in the group of units A^× is determined, and all such maximal subspaces are described up to conjugation. The problem reduces to the classification of associative composition algebras over F. Over fields of characteristic other than 2, the first problem is essentially reduced to the classification of nonisotropic quadratic forms over F and of nonisotropic Hermitian forms over quadratic and quaternionic extensions of F. These results are intimately connected with the problem of intransitive operator spaces between finite-dimensional vector spaces over the

What carries the argument

the reduction of maximal affine unit subspaces to associative composition algebras over F together with nonisotropic quadratic and Hermitian forms over F and its extensions

If this is right

  • The maximal dimension is achieved precisely by subspaces arising from standard constructions based on composition algebras.
  • In characteristic not 2 the dimension is governed by the possible dimensions of the associated nonisotropic forms.
  • All maximal affine unit subspaces are conjugate to one of a short list of model spaces built from those forms.
  • The dual version of Atkinson's theorem holds for primitive spaces of bounded rank matrices over division rings in greater generality.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The classification supplies explicit dimension formulas once the Witt index or rank of the relevant forms is known for a given algebra.
  • The same reduction technique may be tested on explicit low-dimensional examples such as matrix rings over quaternion algebras.
  • The operator-space results suggest a route to bounding the size of large sets of units in algebras that are not necessarily simple.

Load-bearing premise

The base field F satisfies a mild cardinality assumption that rules out small-field pathologies, and prior classifications of quadratic and Hermitian forms over extensions of F are taken as given.

What would settle it

An explicit simple algebra over a sufficiently large field F together with an affine subspace of units whose dimension exceeds the value predicted by the classification, or a maximal subspace that is not conjugate to any of the listed types.

read the original abstract

Let $A$ be a simple algebra over a field $F$. Under a mild cardinality assumption on $F$, we determine the greatest possible dimension for an $F$-affine subspace of $A$ that is included in the group of units $A^\times$, and we describe the spaces that have the greatest possible dimension. This is equivalent to the problem of determining the greatest possible dimension for an $F$-linear subspace $S$ of $A$ in which $x-1_A$ is a unit for all $x \in S$, and we elucidate the structure of these linear subspaces up to conjugation when their dimension reaches the greatest possible one. These classifications involve the associative composition algebras over $F$. Over fields of characteristic other than $2$, the first problem is essentially reduced to the classification of nonisotropic quadratic forms over $F$ and of nonisotropic Hermitian forms over quadratic and quaternionic extensions of $F$. These results are intimately connected with the problem of intransitive operator spaces between finite-dimensional vector spaces over division rings, which we study in depth: in particular, we generalize a dual version of Atkinson's theorem on primitive spaces of bounded rank matrices.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript determines, under a mild cardinality assumption on the base field F, the maximal dimension of an F-affine subspace of a simple F-algebra A that lies entirely inside the group of units A^×, and classifies all such maximal subspaces up to conjugation. Equivalently, it identifies maximal F-linear subspaces S ⊆ A such that x − 1_A is a unit for every x ∈ S. The classification reduces the problem to associative composition algebras over F; in characteristic not 2 it further reduces to the classification of nonisotropic quadratic forms over F and nonisotropic Hermitian forms over quadratic and quaternionic extensions of F. The work also develops the theory of intransitive operator spaces between finite-dimensional vector spaces over division rings and proves a dual version of Atkinson’s theorem on primitive spaces of bounded-rank matrices.

Significance. If the central claims are verified, the paper supplies a sharp dimension bound together with an explicit description of all maximal affine unit subspaces, directly linking them to the well-studied theory of composition algebras and nonisotropic forms. The generalization of the dual Atkinson theorem and the systematic treatment of intransitive operator spaces constitute a concrete advance in the study of linear spaces of matrices with controlled rank or primitivity properties. The reduction to existing classifications of forms is a methodological strength provided the applicability to general central simple algebras is fully justified.

major comments (2)
  1. [main classification theorem and reduction paragraph] The reduction to nonisotropic Hermitian forms over quadratic and quaternionic extensions of F (stated in the abstract and carried out in the main classification theorem): for a general central simple algebra A ≅ M_n(D) with D a division algebra equipped with an involution, the relevant forms are Hermitian with respect to that involution on D. It is not shown that the cited literature on forms over field extensions or split quaternion algebras exhausts the non-split cases; if the classifications invoked do not cover arbitrary involutions on division algebras, then either the dimension bound fails to be sharp or some conjugacy classes of maximal subspaces are omitted.
  2. [section on intransitive operator spaces and dual Atkinson theorem] The correspondence between intransitive operator spaces and the affine subspaces of units (developed in the section on operator spaces and used to prove the up-to-conjugation description): the dual Atkinson theorem is invoked to classify the maximal spaces, but the explicit bijection or reduction step that translates the operator-space data back into the original affine-subspace problem in A is not load-bearingly verified for the general simple-algebra case.
minor comments (2)
  1. [abstract and notation section] Clarify the precise meaning of “quaternionic extensions of F” in the abstract and main statements—whether this includes non-split quaternion algebras or only commutative quadratic extensions.
  2. [references and reduction steps] Add explicit cross-references to the precise theorems in the cited literature on Hermitian forms that are being applied to the non-split division-algebra setting.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful and constructive report. The major comments identify areas where additional explicitness would strengthen the presentation of our reductions and correspondences. We address each point below and indicate the revisions planned for the next version of the manuscript.

read point-by-point responses

  1. Referee: [main classification theorem and reduction paragraph] The reduction to nonisotropic Hermitian forms over quadratic and quaternionic extensions of F (stated in the abstract and carried out in the main classification theorem): for a general central simple algebra A ≅ M_n(D) with D a division algebra equipped with an involution, the relevant forms are Hermitian with respect to that involution on D. It is not shown that the cited literature on forms over field extensions or split quaternion algebras exhausts the non-split cases; if the classifications invoked do not cover arbitrary involutions on division algebras, then either the dimension bound fails to be sharp or some conjugacy classes of maximal subspaces are omitted.

    Authors: We appreciate this observation on the scope of the cited classifications. Our reduction proceeds by first passing from a general central simple algebra to the associative composition algebras over F (quadratic extensions or quaternion algebras, split or otherwise). The Hermitian forms are taken with respect to the standard involution on these algebras. The existing literature on nonisotropic Hermitian forms does extend to the non-split setting through the general theory of forms over division algebras with involution. Nevertheless, to make this applicability fully transparent, we will insert a short clarifying paragraph immediately following the statement of the main classification theorem. This paragraph will recall how the general theory covers arbitrary involutions on quaternion division algebras and confirm that the dimension bound remains sharp with no conjugacy classes omitted. revision: yes

  2. Referee: [section on intransitive operator spaces and dual Atkinson theorem] The correspondence between intransitive operator spaces and the affine subspaces of units (developed in the section on operator spaces and used to prove the up-to-conjugation description): the dual Atkinson theorem is invoked to classify the maximal spaces, but the explicit bijection or reduction step that translates the operator-space data back into the original affine-subspace problem in A is not load-bearingly verified for the general simple-algebra case.

    Authors: We thank the referee for noting the need for a more explicit verification. The manuscript constructs the correspondence by associating an affine subspace of units in A to an intransitive operator space via left and right multiplications, under which the unit condition translates into a primitivity condition. The dual Atkinson theorem is then applied to obtain the classification up to conjugation. To render this reduction load-bearing in the general case, we will add a dedicated lemma in the operator-spaces section that spells out the bijection step by step, including the translation of maximality and conjugacy from the operator-space setting back to affine subspaces of A^×. This will make the argument fully rigorous for arbitrary simple algebras. revision: yes

Circularity Check

0 steps flagged

No significant circularity; central result is a reduction to independent external classifications

full rationale

The paper's derivation reduces the maximal dimension and conjugacy classes of affine subspaces of units in a simple algebra A to the known classification of associative composition algebras together with nonisotropic quadratic and Hermitian forms over F and its extensions. This reduction is explicitly stated as depending on prior literature under a cardinality assumption on F to avoid small-field cases. No self-definitional loops, fitted inputs renamed as predictions, or load-bearing self-citations appear in the abstract or described chain; the classifications invoked are treated as external mathematical results rather than derived within the paper. The argument is therefore self-contained once those independent results are granted, yielding a normal non-circular finding for a classification theorem in algebra.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on the standard definition of simple algebras, the group of units, and the external classification theorems for quadratic forms and composition algebras; no new free parameters or invented entities are introduced in the abstract.

axioms (2)
  • domain assumption A is a finite-dimensional simple algebra over a field F
    Stated at the beginning of the abstract as the setup for the problem.
  • domain assumption F satisfies a mild cardinality assumption
    Invoked to guarantee the classification holds and to exclude small-field counterexamples.

pith-pipeline@v0.9.0 · 5735 in / 1431 out tokens · 55466 ms · 2026-05-19T00:38:24.255256+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Spaces of matrices with few eigenvalues (II)

    math.RA 2026-05 unverdicted novelty 7.0

    Maximum dimensions are determined for spaces of n-by-n matrices with at most two eigenvalues or one nonzero eigenvalue over fields of characteristic 2.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages · cited by 1 Pith paper

  1. [1]

    Atkinson, S

    M.D. Atkinson, S. Lloyd, Primitive spaces of matrices of bounded rank. J. Austral. Math. Soc. (Ser. A) 30 (1980) 473–482

    work page 1980

  2. [2]

    Atkinson, Primitive spaces of matrices of bounded rank II

    M.D. Atkinson, Primitive spaces of matrices of bounded rank II. J. Austral. Math. Soc. (Ser. A) 34 (1983) 306–315

    work page 1983

  3. [3]

    Doboviˇ sek, Simultaneous symmetrization

    M. Doboviˇ sek, Simultaneous symmetrization. Linear Algebra Appl. 383 (2004) 107–112. 79

    work page 2004

  4. [4]

    Eisenbud, J

    D. Eisenbud, J. Harris, Vector spaces of matrices of low rank. Adv. Math. 37 (1988) 135–155

    work page 1988

  5. [5]

    Farenick, D.A.F

    D. Farenick, D.A.F. Pidkowich, The spectral theorem in quaternions. Linear Algebra Appl. 371 (2003) 75–102

    work page 2003

  6. [6]

    Flanders, On spaces of linear transformations with bounded rank

    H. Flanders, On spaces of linear transformations with bounded rank. J. Lond. Math. Soc. 37 (1962) 10–16

    work page 1962

  7. [7]

    Gerstenhaber, On nilalgebras and linear varieties of nilpotent matrices (I)

    M. Gerstenhaber, On nilalgebras and linear varieties of nilpotent matrices (I). Amer. J. Math. 80 (1958) 614–622

    work page 1958

  8. [8]

    M.-A. Knus, A. Merkurjev, M. Rost, J.-P. Tignol, The Book of Involutions, Colloquium Publications, vol. 44, AMS, 1998

    work page 1998

  9. [9]

    Mathes, M

    B. Mathes, M. Omladiˇ c, H. Radjavi, Linear spaces of nilpotent matrices. Linear Algebra Appl. 149 (1991) 215–225

    work page 1991

  10. [10]

    Meshulam, On two extremal matrix problems

    R. Meshulam, On two extremal matrix problems. Linear Algebra Appl. 114-115 (1989) 261–271

    work page 1989

  11. [11]

    Motzkin, O

    T.S. Motzkin, O. Taussky, Pairs of matrices with property L II. Trans. Amer. Math. Soc. 80 (1955) 387–401

    work page 1955

  12. [12]

    Quinlan, Spaces of matrices without non-zero eigenvalues in their field of definition, and a question of Szechtman

    R. Quinlan, Spaces of matrices without non-zero eigenvalues in their field of definition, and a question of Szechtman. Linear Algebra Appl. 434 (2011) 1580–1587

    work page 2011

  13. [13]

    Rand´ e, C

    B. Rand´ e, C. de Seguins Pazzis, The linear preservers of real diagonalisable matrices. Linear Algebra Appl. 435-6 (2011) 1257–1266

    work page 2011

  14. [14]

    Scharlau, Quadratic and Hermitian Forms, Grundlehren der mathema- tische Wissenschaften, vol

    W. Scharlau, Quadratic and Hermitian Forms, Grundlehren der mathema- tische Wissenschaften, vol. 270, Springer-Verlag, Berlin, 1985

    work page 1985

  15. [15]

    de Seguins Pazzis, On the matrices of given rank in a large subspace

    C. de Seguins Pazzis, On the matrices of given rank in a large subspace. Linear Algebra Appl. 435-1 (2011) 147–151

    work page 2011

  16. [16]

    de Seguins Pazzis, Large affine spaces of matrices with rank bounded below

    C. de Seguins Pazzis, Large affine spaces of matrices with rank bounded below. Linear Algebra Appl. 437-2 (2012) 499–518

    work page 2012

  17. [17]

    de Seguins Pazzis, On Gerstenhaber’s theorem for spaces of nilpotent matrices over a skew field

    C. de Seguins Pazzis, On Gerstenhaber’s theorem for spaces of nilpotent matrices over a skew field. Linear Algebra Appl. 438-11 (2013) 4426–4438. 80

    work page 2013

  18. [18]

    de Seguins Pazzis, Large affine spaces of non-singular matrices

    C. de Seguins Pazzis, Large affine spaces of non-singular matrices. Trans. Amer. Math. Soc. 365 (2013) 2569–2596

    work page 2013

  19. [19]

    de Seguins Pazzis, From primitive spaces of bounded rank matrices to a generalized Gerstenhaber theorem

    C. de Seguins Pazzis, From primitive spaces of bounded rank matrices to a generalized Gerstenhaber theorem. Quart. J. Math. 65-2 (2014) 319–325

    work page 2014

  20. [20]

    de Seguins Pazzis, Local linear dependence seen through duality I

    C. de Seguins Pazzis, Local linear dependence seen through duality I. J. Pure Appl. Algebra 219 (2015) 2144–2188

    work page 2015

  21. [21]

    de Seguins Pazzis, Local linear dependence seen through duality II

    C. de Seguins Pazzis, Local linear dependence seen through duality II. Lin- ear Algebra Appl. 462 (2014) 133–185

    work page 2014

  22. [22]

    de Seguins Pazzis, Primitive spaces of matrices with upper rank two over the field with two elements

    C. de Seguins Pazzis, Primitive spaces of matrices with upper rank two over the field with two elements. Linear Multilinear Algebra64 (2016) 1321–1353

    work page 2016

  23. [23]

    de Seguins Pazzis, From trivial spectrum subspaces to spaces of diago- nalisable real matrices

    C. de Seguins Pazzis, From trivial spectrum subspaces to spaces of diago- nalisable real matrices. Math. Proc. R. Ir. Acad. 118-A (2018) 5–8

    work page 2018

  24. [24]

    de Seguins Pazzis, The structured Gerstenhaber problem I

    C. de Seguins Pazzis, The structured Gerstenhaber problem I. Linear Alge- bra Appl. 567 (2019) 263–298

    work page 2019

  25. [25]

    de Seguins Pazzis, Spaces of triangularizable matrices

    C. de Seguins Pazzis, Spaces of triangularizable matrices. Acta Sci. Math. (Szeged), in press. https://doi.org/10.1007/s44146-025-00190-w

    work page doi:10.1007/s44146-025-00190-w

  26. [26]

    Serezhkin, Linear transformations preserving nilpotency (in Russian)

    V.N. Serezhkin, Linear transformations preserving nilpotency (in Russian). Izv. Akad. Nauk BSSR, Ser. Fiz.-Mat. Nauk 125 (1985) 46–50

    work page 1985

  27. [27]

    Shafarevich, Basic Algebraic Geometry 1 (3rd edition) Springer, Hei- delberg, New York, Dordrecht, London (2013)

    I.R. Shafarevich, Basic Algebraic Geometry 1 (3rd edition) Springer, Hei- delberg, New York, Dordrecht, London (2013)

    work page 2013

  28. [28]

    Waterhouse, Self-adjoint operators and formally real fields

    W. Waterhouse, Self-adjoint operators and formally real fields. Duke Math. J. 43 (1976) 237–243. 81

    work page 1976