Are Bundles Good Deals for First-Order Modal Logic?

Mo Liu,Anantha Padmanabha,R. Ramanujam,Yanjing Wang
DOI: https://doi.org/10.1016/j.ic.2023.105062
IF: 1.24
2023-01-01
Information and Computation
Abstract:Bundled products are often offered as good deals to customers. In first-order modal logic (FOML), when we bundle quantifiers and modalities together (as in ∃ x □, ◇ ∀ x, etc.), we get new logical operators whose combinations produce interesting fragments of FOML without any restriction on the arity of predicates, the number of variables, or the modal scope. It is well-known that finding decidable fragments of FOML is hard, so we may ask: do bundled fragments that exploit the distinct expressivity of FOML constitute good deals in balancing the expressivity and complexity? There are a few positive earlier results towards identifying the decidable ones. In this paper, we map the terrain of bundled fragments of FOML in (un)decidability, and in the cases without a definite answer yet, we show that they lack the finite model property. Moreover, whether the class of models considered has constant domains (across states/worlds) or increasing domains presents another layer of complexity. We also present the loosely bundled fragment , which generalizes the bundles and yet retains decidability (over increasing domain models).
What problem does this paper attempt to address?