Abstract:Modal logic becomes action logic by adding programs as in propositional dynamic logic or the μ– calculus. Modal languages can be s en as decidable fragments of first-order logic that admit a natural bisimulation, and hence enjoy a good model theory. Recently, much stronger 'guarded fragments' of first order logic have been identified that enjoy the same pleasant features. The latter can serve as richer action languages as well. We will develop the logic of guarded fragments as a form of process theory. ln particular, moving from sequential to parallel process operations correlates with moving to first-order fragments that are close to, or perhaps just over the decidable–undecidable fence. 1 The modal dynamics of actions We will start by reviewing the basics. Standard poly modal logic is a decidable fragment of the first-order logic of process graphs (labeled transition systems, Kripke models). It can be characterized semantically as consisting, up to logical equivalence, of those first order formulas which are invariant for bisimulation. Propositional dynamic logic turns this into an explicit action language by treating propositions and programs on a par, adding a syntactic component of regular programs, including tests for all propositions. Again, this system is decidable, its propositions are invariant for bisimulation, while its programs are what may be called 'safe for bisimulation'. (Roughly speaking, transition relations for all programs enjoy automatic zig-zag over any existing bisimulation). To obtain the full power of fixed-point operations over all syntactically positive predicate transformers, however, one must move to the modal μ-calculus. Again, the latter system is decidable, and it consists of all bisimulation-invariant statements in a first-order logic with fixed-point operators over process graphs. (This convenient paraphrase of a recent semantic characterization is equivalent to the version involving monadic second-order logic.) This line of logics runs into clear limitations, as it does not handle joint or parallel action. But read on. 2 From modal to guarded logics Modal logic behaves much like a miniature of first order logic in its main system properties (effective axiomatizability, interpolation, preservation results). The mechanism that drives this strong similarity is essentially the following meta-equation: ML : FOL = bisimulation : potential isomorphism We will unpack this terse, but meaning-laden tatement somewhat in the tutorial. Of course, modal

Program Constructions that are Safe for Bisimulation

Modal Logic and Invariance.

UvA-DARE ( Digital Academic Repository ) Expressiveness modulo bisimilarity : a coalgebraic perspective

3 1 Exploring Logical Dynamics : The Main Lines

The Safe Lambda Calculus

Bisimulation and expressivity for conditional belief, degrees of belief, and safe belief

Exploring logical dynamics

Possible Worlds Semantics: A Research Program That Cannot Fail?

Process Operations In Extended Dynamic Logic

Learning to Prove Safety over Parameterised Concurrent Systems (Full Version)

Structural Invariants for Parametric Verification of Systems with Almost Linear Architectures

Modal Logic, Transition Systems and Processes

Compositional Safety Verification with Max-SMT

Better Bounded Bisimulation Contractions (Preprint)

Proving Cutoff Bounds for Safety Properties in First-Order Logic

A Topos View of Blockchain Consensus Protocols

Extending and Implementing the Stable Model Semantics

Modality, Bisimulation and Interpolation in Infinitary Logic

Bisimulations for probabilistic linear lambda calculi

Semantics of Sets of Programs

Effect of the second-generation antihistamine, fexofenadine, on cough reflex sensitivity and pulmonary function.