A new representation of finite Hoops using a new type of product of structures

Michal Botur · 2025

Read A new representation of finite Hoops using a new type of product of structures by Michal Botur (2025) 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-2511-06730.txt.

Introduction

Michal Botur's paper in algebra defines a new kind of product for hoops, which are algebraic structures used in the study of many-valued logic. The construction works for finite hoops, meaning hoops with only finitely many elements. It lets any finite hoop be split into two simpler parts: one of its filters and the matching quotient structure built from that filter.

Botur also shows that this new product behaves well under regrouping, a property like associativity for plain multiplication. As a consequence, every finite hoop can be built as a product of finite MV-chains, which are among the best understood building blocks in this area. The result gives mathematicians a cleaner way to describe the structure of finite hoops.

Catalog record

CREATORMichal Botur
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE47 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-06730

Browse the full library · About the Tower of Babel

Full text

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

A new representation of finite Hoops using a new type of product of structures Thanks: Author acknowledge the support by GAČR 24-14386L and by IGA, project PřF 2024 011 and 2025 008. Michal Botur Abstract . We define a new type of product of hoops which, in the case of finite hoops, can decompose an arbitrary hoop 𝐀 \mathbf{A} into a filter F F and the corresponding homomorphic image 𝐀 / F \mathbf{A}/F . Since our new product is associative up to isomorphism, we can show that every finite hoop is representable as a product of finite MV-chains. 2010 Mathematics Subject Classification Primary 06D35, Secondary 03B50 1. Motivation Hoops were introduced in an unpublished manuscript by Büchi and Owens, grounded in the work of Bosbach on partially ordered monoids [ 3 ] . Hoops have been thoroughly studied and we have a very good understanding of their structure (see for example [ 2 , 6 , 4 ] ). This paper provides a new representation of finite hoops, whose general idea grew out of investigations of various possibilities of wreath and semidirect products of residuated structures. These ideas, in their initial form, occurred to me during my internship with Constantine Tsinakis, and it was this eminent scholar who gave the initial impetus to this research. I was fortunate to be able to work with Constantine, and receive support from him. This article is a token of respect and gratitude. Another motivation and inspiration for the notions introduced in the main part was the deep theory of semidirect and wreath products of semigroups, which culminates in the famous Krohn-Rhodes theorem [ 1 ] . A semigroup action \mathcal{A}=(X,\mathbf{A},\bullet) is a structure where \mathbf{A}=(A;\cdot,1) is a semigroup, X is a set and \bullet\colon X\times A\longrightarrow X is a map, called the right action , which satisfies x\bullet(a\cdot b)=(x\bullet a)\bullet b , for all x\in X and a,b\in A . For any semigroup action \mathcal{B}=(X,\mathbf{B},\bullet) and any semigroup \mathbf{A}=(A;\cdot) we can define the following semigroup \mathbf{A}\wr\mathcal{B}=(A^{X}\times B,\cdot) where (\overline{a}_{1},b_{1})\cdot(\overline{a}_{2},b_{2})=(\overline{a}_{1}\cdot(\overline{a}_{2}\asterisk b_{1}),b_{1}\cdot b_{2}), (\overline{a}_{2}\asterisk b_{1})(x)=\overline{a}_{2}(x\bullet b_{1}) for all (\overline{a}_{1},b_{1}),(\overline{a}_{2},b_{2})\in A^{X}\times B and any x\in X . The semigroup \mathbf{A}\wr\mathcal{B} is called a wreath product of \mathbf{A} and \mathcal{B} . There is a certain asymmetry in this notion of wreath product, since \mathbf{A} does not need to act on anything. But wreath product can be defined for two semigroup actions as well. If \mathcal{A}=(X,\mathbf{A},\bullet) and \mathcal{B}=(Y,\mathbf{B},\bullet) are semigroup actions then we define \mathcal{A}\wr\mathcal{B}=(X\times Y,\mathbf{A}\wr\mathcal{B},\bullet) such that (x,y)\bullet(\overline{a},b)=(x\bullet\overline{a}(y),y\bullet b) for any (x,y)\in X\times Y and (\overline{a},b)\in A^{Y}\times B . Defined this way, wreath product can be seen as an operation in the class of all semigroup actions. The first (and at first sight perhaps surprising) result is associativity of such wreath product. More precisely, if \mathcal{A}=(X,\mathbf{A},\bullet) , \mathcal{B}=(Y,\mathbf{B},\bullet) and \mathcal{C}=(Z,\mathbf{C},\bullet) are semigroup actions then \mathcal{A}\wr(\mathcal{B}\wr\mathcal{C})\cong(\mathcal{A}\wr\mathcal{B})\wr\mathcal{C} where the isomorphism ( \cong ) naturally means the existence of a pair of bijections between sets together with a monoid isomorphism, preserving actions. Associativity is an important property guiding the approach taken in this work, so let me analyse it in a little more detail. First of all, for sets on which the semigroups act, the associativity is the usual associativity of direct product of sets. For semigroups, their wreath product corresponds to the cascade product of automata. The second “semigroup factor” forms a kind of skeleton of the wreath product, and each element x of the skeleton induces an automorphism on the first semigroup. Denoting this automorphism, in a category-theory fashion, by -\asterisk x , we see that in the wreath product any such x is replaced by the “ -\asterisk x transformed” copy of the semigroup. This is illustrated in Figure 1 . Associativity, as illustrated in Figure 2 , states that such an cascade expansion is independent on the order in which it is performed. The Krohn-Rhodes theorem (and all of the subsequent theory) is central to our motivation for introducing new products for hoops. Namely, Krohn-Rhodes theorem shows that all monoid actions can in some sense be decomposed into simple groups and three element flip-flop monoids (for more details we refer the reader to [ 1 ] ). Similarly we show that any finite hoop is a “new type” product of MV-chains. We want to emphasize that the introduction of this new product for hoops is not meant to be a mere syntactic analogy with the semigroup case. Rather, we look for a “product-like” operation, which would be at the very least associative in a natural way, and we wish to proceed towards a representation of hoops by way of some result that would say: every finite hoop is constructible as a new product of certain new-product-irreducible ones . Figure 1. Idea of a wreath product of semigroup actions. Figure 2. Associativity of a wreath product. For the sake of completeness, we mention that the possibility of a semidirect product similar to the one considered in this paper is also dealt with in [ 5 , 8 ] . However, neither of these articles focuses on associativity, which is our main motivation. 2. Introduction A commutative ordered monoid (A,\cdot,1,\leq) is ordered naturally if the order relation satisfies : a\leq b if and only if exists c\in A such that b\cdot c=a (for all a,b\in A ). A hoop can be defined as naturally ordered commutative monoid possessing the binary operation \rightarrow such that the adjointnes property a\cdot b\leq c\text{ if and only if }a\leq b\rightarrow c holds for all a,b,c . The usual equivalent definition that follows is more algebraic but less clear. Definition 1 . A hoop is the algebra \mathbf{A}=(A;\cdot,\rightarrow,1) of the type \langle 2,2,0\rangle , where (A;\cdot,1) is a commutative monoid and satisfying the identities (H1) x\rightarrow x=1 , (H2) (x\cdot y)\rightarrow z=x\rightarrow(y\rightarrow z), (H3) x\cdot(x\rightarrow y)=y\cdot(y\rightarrow x) . If \mathbf{A}=(A;\cdot,\rightarrow,1) is a hoop then we introduce the relation \leq on A by x\leq y if 1=x\rightarrow y (for any x,y\in A ). It is well known that (A;\leq) is an ordered set with the top element 1 such that there exists an infimum x\wedge y=x\cdot(x\rightarrow y) for any x,y\in A . This property is known as divisibility ; in presence of the other defining identities it is equivalent to (H3). Lemma 2 ( [ 6 ] ) . If \mathbf{A}=(A;\cdot,\rightarrow,1) is a hoop and x,y,z\in A then it satisfies (i) if x\leq y then x\cdot z\leq y\cdot z, z\rightarrow x\leq z\rightarrow y and y\rightarrow z\leq x\rightarrow y, (ii) x\cdot y\leq z if and only if x\leq y\rightarrow z, (iii) x\rightarrow(y\wedge z)=(x\rightarrow y)\wedge(x\rightarrow z), (iv) if the supremum y\vee z exists then also (x\cdot y)\vee(x\cdot z) exists and x\cdot(y\vee z)=(x\cdot y)\vee(x\cdot z), (v) if the supremum x\vee y exists then (x\vee y)\rightarrow z=(x\rightarrow z)\wedge(y\rightarrow z). Lemma 3 . If \mathbf{A}=(A;\cdot,\rightarrow,1) is a hoop then x\cdot(y\wedge z)=(x\cdot y)\wedge(x\cdot z) holds for any x,y,z\in A. Proof. It is clear that x\cdot(x\rightarrow(x\cdot z))\leq x\cdot z and z\leq x\rightarrow(x\cdot z) implies the inequality (x\rightarrow(x\cdot z))\rightarrow y\leq z\rightarrow y. Consequently we obtain \displaystyle(x\cdot y)\wedge(x\cdot z) \displaystyle= \displaystyle x\cdot y\cdot((x\cdot y)\rightarrow(x\cdot z)) \displaystyle= \displaystyle x\cdot y\cdot(y\rightarrow(x\rightarrow(x\cdot z))) \displaystyle= \displaystyle x\cdot(x\rightarrow(x\cdot z))\cdot((x\rightarrow(x\cdot z))\rightarrow y) \displaystyle\leq \displaystyle x\cdot z\cdot(z\rightarrow y) \displaystyle= \displaystyle x\cdot(y\wedge z). The converse inequality is clear. ∎ The next

Read the complete text