The variety of complemented lattices where the Sasaki operations form an adjoint pair
Václav Cenker, Ivan Chajda, Helmut Länger · 2024
Read The variety of complemented lattices where the Sasaki operations form an adjoint pair by Václav Cenker, Ivan Chajda, Helmut Länger (2024) 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-2412-15764.txt.
Introduction
Václav Cenker, Ivan Chajda, and Helmut Länger characterize a class of lattices tied to quantum logic in this 2024 paper. The Sasaki projection maps the subspaces of a Hilbert space onto a segment, and its paired operations were defined by two of the authors earlier. When those operations form an adjoint pair, the lattice gains special structure.
Earlier work described several classes with the property, from special lattices to semirings. Here the authors prove the complemented lattices with the property form a variety, and they state its defining identities explicitly. The variety is also congruence permutable and regular, so every ideal is the kernel of some congruence. They finish with a finite basis of ideal terms and a description of the congruence each ideal determines.
Catalog record
| CREATOR | Václav Cenker, Ivan Chajda, Helmut Länger |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 20 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-15764
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The variety of complemented lattices where the Sasaki operations form an adjoint pair Václav Cenker, Ivan Chajda, Helmut Länger Abstract The Sasaki projection was introduced as a mapping from the lattice of closed subspaces of a Hilbert space onto one of its segments. To use this projection and its dual so-called Sasaki operations were introduced by the second two authors in [ 4 ] and [ 6 ] . In [ 6 ] there are described several classes of lattices, λ \lambda -lattices and semirings where the Sasaki operations form an adjoint pair. In the present paper we prove that the class of complemented lattices with this property forms a variety and we explicitly state its defining identities. Moreover, we prove that this variety 𝒱 \mathcal{V} is congruence permutable and regular. Hence every ideal I I of some member 𝐋 \mathbf{L} of 𝒱 \mathcal{V} is a kernel of some congruence on 𝐋 \mathbf{L} . Finally, we determine a finite basis of so-called ideal terms and describe the congruence Θ I \Theta_{I} determined by the ideal I I . 0 0 footnotetext: Support of the research of the first two authors by the Czech Science Foundation (GAČR), project 24-14386L, entitled “Representation of algebraic semantics for substructural logics”, and by IGA, project PřF 2024 011, is gratefully acknowledged. AMS Subject Classification: 06C15, 06B05, 06C20 Keywords: Complemented lattice, Sasaki projection, Sasaki operations, orthomodular lattice, variety, congruence permutability, regularity, ideal, congruence kernel, ideal term, basis of ideal terms It was proved by G. Birkhoff and J. von Neumann [ 2 ] in 1936, and independently by K. Husimi [ 8 ] in 1937 that the lattice \mathcal{L}(\mathbf{H}) of closed subspaces of a Hilbert space \mathbf{H} is orthomodular. For a given element a\in\mathcal{L}(\mathbf{H}) , U. Sasaki [ 10 ] introduced the projection p_{a} of \mathcal{L}(\mathbf{H}) onto its interval [0,a] as follows: p_{a}(x)=(x\vee a^{\prime})\wedge a. The name Sasaki projection was given later by M. Nakamura [ 9 ] . Also, we can formulate its dual, i.e. the projection \overline{p_{a}}(x)=(x\wedge a)\vee a^{\prime} of \mathcal{L}(\mathbf{H}) onto its interval [a^{\prime},1] . However, we can extend these definitions by introducing the corresponding concepts in every bounded lattice (L,\vee,\wedge,{}^{\prime},0,1) equipped with a unary operation ′ . Using this framework, we define two binary operations on L . Let \mathbf{L}=(L,\vee,\wedge,{}^{\prime},0,1) be a bounded lattice with unary operation ′ . By the Sasaki operations on \mathbf{L} we mean the following term operations: x\odot y=(x\vee y^{\prime})\wedge y\quad\text{and}\quad x\to y=x^{\prime}\vee(x\wedge y). We say that they form an adjoint pair on \mathbf{L} if the following holds: (A) x\odot y\leq z if and only if x\leq y\to z . Lemma 1 . Let \mathbf{L}=(L,\vee,\wedge,{}^{\prime},0,1) be a bounded lattice with unary operation ′ . If the Sasaki operations \odot , \to form an adjoint pair in \mathbf{L} , then the unary operation ′ is a complementation. Proof. Because 1\odot a=a\leq 1 for each a\in L , by (A), we derive 1\leq a\to 1=a^{\prime}\vee(a\wedge 1)=a^{\prime}\vee a . Since 0\leq a\to 0 , we also have a^{\prime}\wedge a=(0\vee a^{\prime})\wedge a=0\odot a\leq 0 , i.e. a^{\prime}\wedge a=0 . Thus, ′ is a complementation. ∎ Hence, investigating whether the Sasaki operations in lattices with a unary operation form an adjoint pair makes sense only for lattices with complementation. At first we determine the class of such lattices. Theorem 2 . Let \mathbf{L}=(L,\vee,\wedge,{}^{\prime},0,1) be a bounded lattice with complementation. Then the following are equivalent: (a) x\odot y\leq z implies x\leq y\to z for all x,y,z\in L , (b) \mathbf{L} satisfies the identity x\vee y^{\prime}\approx y^{\prime}\vee\big((x\vee y^{\prime})\wedge y\big) , (c) x^{\prime}\leq y implies y=x^{\prime}\vee(y\wedge x) . Furthermore, the following are equivalent: (d) x\leq y\to z implies x\odot y\leq z for all x,y,z\in L , (e) \mathbf{L} satisfies the identity x\wedge y\approx x\wedge\big((x\wedge y)\vee x^{\prime}\big) , (f) x\leq y implies x=(y^{\prime}\vee x)\wedge y . Proof. (a) \Rightarrow (c): Assume x^{\prime}\leq y for x,y\in L . Then, y\odot x=(y\vee x^{\prime})\wedge x=y\wedge x . According to (a) we obtain y\leq x\to(y\wedge x)=x^{\prime}\vee(y\wedge x)\leq y\vee(y\wedge x)=y and hence y=x^{\prime}\vee(y\wedge x) . (c) \Rightarrow (b): Because y^{\prime}\leq x\vee y^{\prime} , we apply (c) to get x\vee y^{\prime}=y^{\prime}\vee\big((x\vee y^{\prime})\wedge y) . (b) \Rightarrow (a): Assume x\odot y\leq z . Using (b) we derive \displaystyle x \displaystyle\leq x\vee y^{\prime}=y^{\prime}\vee\big((x\vee y^{\prime})\wedge y\big)=y^{\prime}\vee\Big(\big((x\vee y^{\prime})\wedge y\big)\wedge y\Big)=y^{\prime}\vee\big((x\odot y)\wedge y\big)\leq \displaystyle\leq y^{\prime}\vee(z\wedge y)=y\to z. Similarly, we prove the remaining. (d) \Rightarrow (f): Let x\leq y . Then y\to x=y^{\prime}\vee(y\wedge x)=y^{\prime}\vee x . By (d) we have (y^{\prime}\vee x)\odot y\leq x whence (y^{\prime}\vee x)\wedge y=\big((y^{\prime}\vee x)\big)\vee y^{\prime})\wedge y=(y^{\prime}\vee x)\odot y\leq x. Conversely, from the assumption and fact that x\leq y^{\prime}\vee x we conclude x\leq(y^{\prime}\vee x)\wedge y . Consequently, x=(y^{\prime}\vee x)\wedge y . (f) \Rightarrow (e): Because x\wedge y\leq x , we apply (f) to get x\wedge y=x\wedge\big((x\wedge y)\vee x^{\prime}\big) . (e) \Rightarrow (d): Assume x\leq y\to z . Then, using (e) we obtain \displaystyle x\odot y \displaystyle=(x\vee y^{\prime})\wedge y\leq\big((y\to z)\vee y^{\prime}\big)\wedge y=\Big(\big(y^{\prime}\vee(y\wedge z)\big)\vee y^{\prime}\Big)\wedge y= \displaystyle=\big(y^{\prime}\vee(y\wedge z)\big)\wedge y=y\wedge z\leq z. ∎ Corollary 3 . The class of bounded lattices with a unary operation \,{}^{\prime} where the Sasaki operations form an adjoint pair is a variety determined by the identities of complemented lattices and the identities (b) and (c) from Theorem 2 . Let us denote this variety by \mathcal{V} . Remark 4 . It is evident that the aforementioned variety \mathcal{V} includes the variety of orthomodular lattices, but does not coincide with it. For instance, the modular lattice \mathbf{M}_{3} depicted in Figure 1 [formula] [formula] [formula] [formula] [formula] Fig. 1 Modular lattice [formula] with complementation defined by the table \begin{array}[]{l|ccccc}x&0&a&b&c&1\\ \hline\cr x^{\prime}&1&b&c&a&0\end{array} belongs to \mathcal{V} but it is not an orthomodular lattice since the complementation is not an involution since (a^{\prime})^{\prime}=b^{\prime}=c\neq a . However, the complementation of \mathbf{M}_{3} is antitone. Moreover, \mathcal{V} contains also non-modular lattices since not every orthomodular lattice is modular. Further, \mathcal{V} includes the variety of complemented modular lattices since within these lattices we have \displaystyle y^{\prime}\vee\big(y\wedge(x\vee y^{\prime})\big) \displaystyle\approx(y^{\prime}\vee y)\wedge(x\vee y^{\prime})\approx x\vee y^{\prime}, \displaystyle\big((x\wedge y)\vee x^{\prime}\big)\wedge x \displaystyle\approx(x\wedge y)\vee(x^{\prime}\wedge x)\approx x\wedge y. Finally, \mathcal{V} is contained in the variety of dually weakly orthomodular lattices, and every member of \mathcal{V} whose complementation is both surjective and an involution is a weakly orthomodular lattice. For more details on ( dually ) weakly orthomodular lattices cf. [ 5 ] . Example 5 . The modular lattice visualized in Figure 2 [formula] [formula] [formula] [formula] [formula] [formula] [formula] [formula] [formula] [formula] Fig. 2 Member of [formula] with complementation defined by the table \begin{array}[]{l|cccccccccc}x&0&a&b&c&d&e&f&g&h&1\\ \hline\cr x^{\prime}&1&g&h&f&e&d&c&a&b&0\end{array} is a member of \mathcal{V} . Observe that the complementation of this lattice is an involution but not antitone since a\leq f , but f^{\prime}=c\not\leq g=a^{\prime} . In what follows we describe some important congruence properties of the variety \mathcal{V} . Recall the following facts concerning congruence properties. Let \mathbf{A}=(A,F) be an algebra and \BCon\mathbf{A}=(\Con\mathbf{A},\subseteq) denote its congruence lattice. Then \mathbf{A} is called (i) congruence permutable if \Theta\circ\Phi=\Phi\circ\Theta for all \Theta,\Phi\in\Con\mathbf{A} , (ii) congruence distributive if \BCon\mathbf{A} is distributive, (iii) arithmetical if it is both congruence permutable and congruence distributive, (iv) regular if every congruence on \mathbf{A} is determined by each of its classes, i.e. if a\in A , \Theta,\Phi\in\Con\mathbf{A} and [a]\Theta=[a]\Phi together imply \Theta=\Phi , see e.g. [ 3 ] and [ 7 ] . A variety is said to have