[Paper Review] Topological Modal Logics with Difference Modality

Kudinov Andrey|arXiv (Cornell University)|Nov 25, 2010

Logic, Reasoning, and Knowledge8 references3 citations

TL;DR

This paper introduces a topological modal logic extended with a difference modality $[\neq]$, interpreted as 'true everywhere except at the current point', enabling expressivity for previously undefinable topological properties such as $T_1$ separation, density-in-itself, and connectedness. The key contribution is proving completeness theorems for three logics—$\mathbf{S4D}$, $\mathbf{S4DS}$, and $\mathbf{S4DT_1S}$—with respect to increasingly specific classes of topological spaces, including all topological spaces, dense-in-itself spaces, and zero-dimensional $T_1$ metric spaces.

ABSTRACT

We consider propositional modal logic with two modal operators $\Box$ and $\D$. In topological semantics $\Box$ is interpreted as an interior operator and $\D$ as difference. We show that some important topological properties are expressible in this language. In addition, we present a few logics and proofs of f.m.p. and of completeness theorems.

Motivation & Objective

• To extend basic topological modal logic with the difference modality $[\neq]$ to increase its expressive power for spatial properties.
• To investigate whether topological properties like $T_1$ separation, density-in-itself, and connectedness become definable in this enriched language.
• To establish completeness theorems for new logics built on $\mathbf{S4}$ with axioms for $[\neq]$ and additional constraints.
• To clarify the logical characterization of important topological spaces such as $\mathbb{R}^n$ and zero-dimensional $T_1$ spaces.
• To determine the extent to which the difference modality enables the expression of the universal modality and spatial connectivity.

Proposed method

• Introduces a bimodal language with $\Box$ (interior operator) and $[\neq]$ (difference modality), where $[\neq]\phi$ means $\phi$ holds at all points except the current one.
• Defines three logics: $\mathbf{S4D}$, $\mathbf{S4DS}$, and $\mathbf{S4DT_1S}$, each extending $\mathbf{K_2}$ with specific axioms including $B_D$, $4^{-}_D$, $T_\Box$, $4_\Box$, $D_\Box$, $DS$, and $AT_1$.
• Uses topological models where formulas are evaluated over points in a topological space with a valuation assigning sets to propositional letters.
• Employs continuous, surjective, and closed p-morphisms (cd-p-morphisms) from topological spaces to Kripke frames to prove completeness via the canonical model method.
• Applies unravelling and decomposition techniques to reduce complex frames to simpler ones (e.g., generated subframes or clusters), leveraging zero-dimensionality and clopen sets.
• Demonstrates that $[\forall]\phi \equiv [\neq]\phi \land \phi$ expresses the universal modality, enabling quantification over all points.

Experimental results

Research questions

• RQ1 Can the difference modality $[\neq]$ express topological properties such as $T_1$ separation and density-in-itself in topological semantics?
• RQ2 Is $\mathbf{S4D}$, the logic with $\Box$ and $[\neq]$, complete with respect to all topological spaces?
• RQ3 Can $\mathbf{S4DS}$ and $\mathbf{S4DT_1S}$ be shown complete with respect to dense-in-itself and zero-dimensional $T_1$ metric spaces, respectively?
• RQ4 Does the addition of $[\neq]$ allow the expression of connectedness and distinguish $\mathbb{R}$ from $\mathbb{R}^n$ for $n \geq 2$?
• RQ5 What is the D-logic of $\mathbb{R}$, and is $\mathbf{S4D} + AT_1$ complete with respect to all $T_1$ spaces?

Key findings

• The logic $\mathbf{S4D}$ is complete with respect to all topological spaces, as proven via a canonical model construction using cd-p-morphisms.
• The logic $\mathbf{S4DS}$ is complete with respect to all dense-in-itself topological spaces, showing that the difference modality enables definability of density-in-itself.
• The logic $\mathbf{S4DT_1S}$ is complete with respect to all zero-dimensional, dense-in-itself, metric $T_1$ spaces, establishing a strong completeness result for a class of spaces including $\mathbb{R}^n$ for $n \geq 2$.
• The difference modality allows the expression of the universal modality via $[\forall]\phi \equiv [\neq]\phi \land \phi$, enhancing the language's expressive power.
• The axiom $(AE_1)$, which involves $[\neq]$ and $\Box$, helps distinguish $\mathbb{R}$ from $\mathbb{R}^n$ for $n \geq 2$, suggesting that $\mathbf{S4DT_1S} + (AE_1) + \text{``connectedness"}$ is complete over $\mathbb{R}^n$.
• The paper leaves open the question of whether $\mathbf{S4D} + AT_1$ is complete with respect to all $T_1$ spaces, and the D-logic of $\mathbb{R}$ remains unknown.