Simplicial Interpretations of Modal Logic

• Simplicial Interpretations of Modal Logic is a framework that uses high-dimensional simplicial complexes to represent and analyze modalities in multi-agent systems.
• It generalizes traditional Kripke semantics by employing combinatorial topology, three-valued semantics, and assignment operators to manage agents’ dynamic liveness.
• The approach enables rigorous analysis of distributed knowledge, belief revision, and dynamic updates, with clear applications in distributed computing and logic.

Simplicial interpretations of modal logic establish a rigorous, high-dimensional combinatorial framework for representing and analyzing epistemic, doxastic, and dynamic modalities, especially in systems with multiple agents where agents may gain or lose “liveness.” This approach leverages the algebraic-topological structure of simplicial complexes, generalizing and connecting Kripke-style semantics for epistemic modal logic, belief operators, and distributed knowledge. Recent developments have refined these interpretations to accommodate impure and polychromatic complexes, assignment-based term-modal languages, and sophisticated bisimulation notions.

1. Simplicial Complexes as Epistemic Structures

A simplicial complex is a combinatorial structure $(C, \chi)$ with a set $V$ of vertices and $C \subseteq \mathcal{P}(V) \setminus \{\emptyset\}$ closed under non-empty subsets; every $X \in C$ is a “face.” The coloring $\chi: V \rightarrow A$ assigns each vertex an agent, requiring that on any face $X$, $\chi|_X$ is injective. Facets are maximal faces; a complex is pure if all facets have $|A|$ vertices (all agents “alive”), impure if some facets are lower-dimensional (agents “dead”) (Ditmarsch et al., 2021, Yang, 27 Nov 2025). The assignment $\ell$ decorates each vertex (or globally, each facet) with propositional variable valuations.

Simplicial models generalize Kripke models by treating possible worlds as facets and local states as colored vertices. Pure complexes correspond to fully synchronous systems or epistemic situations where all agents are present; impure complexes capture heterogeneity, failures, or varying participation.

2. Simplicial Semantics for Epistemic Modal Logic

Traditional modal operators receive new interpretations:

• Knowledge $K_a$: $(\mathcal{M}, X) \models K_a \varphi$ iff for all facets $Y$ with $a \in \chi(X \cap Y)$, $(\mathcal{M}, Y) \models \varphi$ (Ditmarsch et al., 2020). The indistinguishability relation is generated topologically.
• Three-valued Semantics: When agents are “dead” ($a \notin \chi(X)$), atomic formulas $p_a$, $K_a\varphi$, and combinations are undefined. Semantics distinguishes between true, false, and undefined, formalized via partial definability $(\mathcal{M}, X) \models \triangleright \varphi$ (Ditmarsch et al., 2021).
• Modal Axiomatization (S5$^\top$): A modified S5, with modal rules (e.g., necessitation, K-axiom) and Modus Ponens restricted to “defined” formulas. Definability is characterized inductively: propositional atoms are defined iff the agent is live, Boolean connectives preserve definability, and $K_a\varphi$ is defined if $a$ is live in some neighboring facet where $\varphi$ is defined (Ditmarsch et al., 2021).

Distributed, Mutual, and Common Knowledge extend in a natural way: $D_B\varphi$ and $C_B\varphi$ are interpreted combinatorially in terms of intersections and connectivity among facets sharing agent colors (Ditmarsch et al., 2020).

3. Impure Complexes, Term-Modal Languages, and Assignment Operators

Impure complexes introduce subtleties: certain modal formulas or atomic propositions may refer to dead agents, resulting in conceptually dubious or undefined expressions. The term-modal language $L[:$, equipped with assignment operators $[x a]$ (assign variable $x$ to agent $a$) and distributed-knowledge modalities $K_X$, was developed to syntactically preclude expressions about dead agents. This system supports:

• Simplicial and Kripke Semantics: Both first-order Kripke semantics (local epistemic models) and simplicial semantics are defined for $L[:$, with equivalence proven for a class of local epistemic models (Yang, 27 Nov 2025).
• Assignment Normal Form: Every sentence is provably equivalent to one where all assignments are pushed inside to atoms and modal operators, facilitating normal form representation and syntactic discipline (Yang, 27 Nov 2025).
• Axiomatization (LEL$[:$): Comprises normal modal axioms, assignment interaction axioms, and local-epistemic frame constraints. Completeness is established via construction of canonical quasi-models with coherent assignments.
• Expressive Power and Bisimulation: Logical equivalence coincides with bisimulation equivalence on saturated models, with bisimulation conditions formulated for both facets and Kripke worlds (Yang, 27 Nov 2025).

4. Simplicial Models of Belief and Polychromatic Complexes

The extension of simplicial semantics to doxastic modalities utilizes polychromatic simplicial complexes, where faces may admit multiple vertices of the same color (i.e., agent). This framework naturally induces plausibility preorders:

• Plausibility Order from Multiplicity: For agent $a$, $m_a(X)$ is the number of $a$-colored vertices in world (facet) $X$, defining $X \leq_a Y$ iff $m_a(X) \leq m_a(Y)$ (Cachin et al., 12 Jan 2026).
• Safe Belief and Plain Belief: Safe belief $[\unrhd]_a\varphi$ requires $\varphi$ in all worlds at least as plausible as $X$ for $a$, while plain belief $B_a\varphi$ demands $\varphi$ hold in all minimal $a$-plausible worlds indistinguishable from $X$. Safe belief has S4.2 axiom properties; $$B_a\varphi \equiv \langle\unrhd\rangle_a[\unrhd]_a\varphi$$ (Cachin et al., 12 Jan 2026).
• Epistemic vs. Doxastic: Knowledge operators rely on topological indistinguishability as in the epistemic case, while belief modalities depend on the additional structure from agent multiplicities, which is induced rather than postulated.
• Limitations: Properness fails in polychromatic complexes; it is generally not possible to separate all worlds using knowledge modalities alone. Unlike knowledge, belief operators are not preserved by simplicial morphisms (Cachin et al., 12 Jan 2026).

5. Equivalence and Correspondence with Kripke Semantics

Simplicial models and Kripke frames are closely related:

• Transformations: The $\sigma$-transform converts Kripke models into simplicial complexes by encoding equivalence classes as colored vertices and facets; the $\kappa$-transform yields Kripke models from simplicial complexes by taking facets as states and defining accessibility via shared colors (Ditmarsch et al., 2021, Ditmarsch et al., 2020).
• Local Epistemic Models: In impure settings, each agent’s accessibility relation becomes a partial equivalence, defined only where the agent is alive. Correspondence is proven: every formula is valid in the simplicial model iff it is valid in the associated Kripke model, and vice versa (Ditmarsch et al., 2021, Yang, 27 Nov 2025).
• Expressive Equivalence: Both perspectives validate the same formulas, and modal equivalence coincides with bisimilarity on finite complexes (Ditmarsch et al., 2020, Yang, 27 Nov 2025).

Model Type	Key Features	Logical Correspondence
Pure simplicial	All agents alive in every facet	Classical S5 modal logic
Impure simplicial	Facet-dependent agent “liveness”	Three-valued S5$^\top$ (Ditmarsch et al., 2021)
Polychromatic simplicial	Multiple local states per agent (belief)	S4.2-style doxastic logic (Cachin et al., 12 Jan 2026)
Kripke (local epistemic)	Partial equivalence accessibility	Bisimilar to simplicial models

6. Bisimulation, Dynamics, and Applications

Bisimulation for simplicial complexes generalizes the classical definition: relations between facets must preserve valuation and permit zig-zag witnessing for adjacency along agent colors. On finite complexes, bisimulation aligns with logical equivalence in the modal language (Ditmarsch et al., 2020, Yang, 27 Nov 2025).

Dynamic Epistemic Logic is realized via simplicial action models (complexes of “actions” with pre- and postconditions); product update constructs new complexes whose topology encodes informational changes. This approach extends to dynamic belief revision and distributed knowledge (Ditmarsch et al., 2020).

The geometric/topological nature of simplicial semantics enables the use of combinatorial topology results in distributed computing, notably for characterizing task solvability via invariants like connectivity or holes (Ditmarsch et al., 2020).

7. Illustrative Examples and Implications

Explicit examples show the semantic subtleties of impure and polychromatic simplicial models. For instance, a formula like $K_a p_c$ can be true at a facet $X$ where $c$ is dead, provided that in all adjacent facets where $a$ is alive and $p_c$ is defined and true, but $p_c$ itself is undefined at $X$ (Ditmarsch et al., 2021). Assignment operators systematically handle reference to agents' local states, avoiding ill-defined expressions (Yang, 27 Nov 2025). Polychromatic models illustrate how an agent's belief depends on the plausibility induced by color-multiplicities, with safe and plain belief referents distinguished even for the same agent and facet (Cachin et al., 12 Jan 2026).

A plausible implication is that simplicial semantics offers a unified, geometrically meaningful foundation for multi-agent modal logic, accommodating epistemic, doxastic, and dynamic modalities, and providing strong correspondences with established Kripke-style semantics. The approach is particularly advantageous in distributed computing, where process failures and system topology are best modeled combinatorially.