Monoids generated by projections

Matthew Fayers · 2026

Read Monoids generated by projections by Matthew Fayers (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-28500.txt.

Introduction

Fayers defines semireflection monoids, sets of linear maps built from reflections that fix a codimension-one subspace. When the building blocks are non-invertible projections, the result is a projection monoid. He develops the basic theory and works through key examples. Full classifications follow for projection monoids on the complex plane and on real three-space. A short section then opens the affine version of the story.

Catalog record

CREATORMatthew Fayers
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE115951
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.28500

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Monoids generated by projections Matthew Fayers Abstract We define and explore semireflection monoids on a finite-dimensional vector space. These are monoids generated by semireflections : linear maps fixing a subspace of codimension 1 1 . We mostly focus on the case of projection monoids (where the generating semireflections are non-invertible). After exploring some general theory, we give some important examples, and give classification results for projection monoids on ℂ 2 \mathbb{C}^{2} and ℝ 3 \mathbb{R}^{3} . We then briefly introduce affine projection monoids. 2020 Mathematics subject classification: 15A99, 20F55, 20M99 1 Introduction The theory of finite reflection groups (both real and complex) is one of the cornerstones of twentieth-century mathematics, with applications in numerous areas, and continues to inspire a great deal of research. We begin by summarising the essentials. Throughout this paper we let V be a finite-dimensional vector space over a field \mathbb{F} . We define a reflection on V to be an invertible linear map r:V\to V whose fixed-point space has codimension 1 (or equivalently, for which the space \left\{\left.(r-1)v\ \right|\ \smash{v\in V}\right\} has dimension 1 ). A reflection group on V is a subgroup of \operatorname{GL}(V) generated by reflections. We are interested primarily in finite reflection groups. Obviously in a finite reflection group each reflection r has finite order, so has eigenvalues 1,\dots,1,\zeta for a root of unity \zeta\in\mathbb{F} . Moreover, if \operatorname{char}(\mathbb{F})=0 then \zeta\neq 1 , so that r is diagonalisable. (What we have called a reflection is often called a pseudoreflection , with the term ‘reflection’ being reserved for the case where \zeta=-1 .) A finite reflection group G on V is irreducible if there is no subspace 0<W<V which is preserved by G . It is completely reducible if for every W_{1}\leqslant V which is preserved by G there is another subspace W_{2} which is preserved by G such that V=W_{1}\oplus W_{2} . In this case G decomposes as a direct product G_{1}\times G_{2} , where G_{i} is the pointwise stabiliser of W_{3-i} , and acts as a completely reducible reflection group on W_{i} . Iterating this decomposition shows that for any completely reducible finite reflection group G on V there is a direct sum decomposition V=W_{1}\oplus\dots\oplus W_{r} and a corresponding direct product decomposition G=G_{1}\times\dots\times G_{r} , where G_{i} acts as an irreducible reflection group on V_{i} and fixes V_{j} pointwise for each j\neq i . It is essentially a consequence of Maschke’s Theorem that if \operatorname{char}(\mathbb{F})=0 then every finite reflection group is completely reducible, so to classify finite reflection groups it suffices to classify irreducibles. Over \mathbb{R} this was done by Coxeter [ C1 ] , who showed [ C2 ] that the irreducible finite real reflection groups are precisely the finite Coxeter groups. Over \mathbb{C} the irreducible finite reflection groups were classified by Shephard and Todd [ ST ] ; these consist of the infinite family of groups G(m,p,n) , together with 34 exceptional groups G_{4},\dots,G_{37} . Finite reflection groups in prime characteristic have been considered in a series of papers by Zaleskii and Serezhkin [ ZS1 , ZS2 , ZS3 ] . Now we introduce the generalisations of reflection groups that we study in this paper. Suppose V is a finite-dimensional vector space over \mathbb{F} . We define a semireflection to be a (not necessarily invertible) linear map r:V\to V whose fixed-point space has codimension 1 . So a semireflection is either a reflection or a projection onto a subspace of dimension n-1 . For the rest of this paper, we use the term ‘projection’ to mean a projection onto a subspace of codimension 1 ; or in other words, an idempotent linear map of nullity 1 . We define a semireflection monoid on V to be a submonoid of \operatorname{End}_{\mathbb{F}}(V) generated by semireflections. A semireflection monoid is a projection monoid if it is generated by projections. We will mainly restrict attention to projection monoids in this paper; we are particularly interested in finding finite projection monoids. We will concentrate on irreducibility and classification, leaving the study of the algebraic structure of these monoids to future work. In the next section we will explore some basic theory of semireflection monoids and projection monoids, particularly relating to irreducibility and complete reducibility. In Section 3 we introduce two important examples and give come of their properties. In Sections 4 and 5 we classify irreducible projection monoids on \mathbb{C}^{2} and on \mathbb{R}^{3} , and in the final section we introduce affine projection monoids. 2 Semireflection monoids 2.1 Some more definitions Here we give some more basic definitions relating to semireflection monoids. First suppose M,N are semireflection monoids on vector spaces V,W respectively. Then we say that M and N are equivalent if there is a bijective linear map f:V\to W such that N=\left\{\left.fmf^{-1}\ \right|\ \smash{m\in M}\right\} . If r:V\to V is a semireflection, then so is the dual map r^{\ast}:V^{\ast}\to V^{\ast} . So if M is a semireflection monoid on V , then there is a dual semireflection monoid M^{\ast} on V^{\ast} . Unlike in the case of finite reflection groups, a finite semireflection monoid need not be equivalent (or even isomorphic an an abstract monoid) to its dual. Now we show how to decompose a semireflection monoid into a ‘reflection part’ and a ‘projection part’. Suppose M is a semireflection monoid. Then we define M_{0} to be the submonoid generated by all the projections in M , and M_{1} the submonoid generated by the reflections in M . Then M_{1} is simply the set of invertible elements of M and is a reflection group. Moreover, M_{0}\cap M_{1}=\{1\} , M_{0}M_{1}=M , and M_{1} normalises M_{0} . (In many cases M_{0} comprises all the non-invertible elements of M , but this is not true in general.) Given a projection monoid M_{0} , we define its normalising reflection group to be the group generated by all the reflections that normalise M_{0} . Now to construct a finite semireflection monoid, we can proceed in stages: \diamond construct a finite projection monoid M_{0} ; \diamond let G be the normalising reflection group of M_{0} ; \diamond let M_{1} be a finite subgroup of G , and let M be the semireflection monoid M_{0}M_{1} . In this paper we will mainly be concerned with the first step, though for many of the projection monoids that we construct we will find the normalising reflection group. We remark that one can consider instead projection semigroups instead of projection monoids, with no practical difference; a projection semigroup is the same as the corresponding projection monoid but with the identity element removed. The choice is largely a matter of taste; our choice to work with monoids simplifies some statements (e.g. the expression M_{0}M_{1} above would need to be M_{0}M_{1}\cup M_{1} when working with semigroups, and references to direct products below would need modification) while complicating others (in Section 3 below we repeatedly make reference to the non-invertible elements of a projection monoid, which are simply the elements of the corresponding projection semigroup). 2.2 Irreducibility and complete reducibility Suppose M is a projection monoid on V . We say that M is reducible if there is a subspace 0<W<V which is mapped to itself by all elements of M . We call such

Read the complete text