Monotonicity in Intensional Contexts: Weakening and
Pragmatic Effects under Modals and Attitudes
Jialiang Yan

Abstract:
Modality holds a central position in the fields of formal semantics and modal logic. This dissertation delves into epistemic and desiderative modalities, focusing on addressing various monotonicity puzzles in modal contexts.
Two primary topics are investigated throughout the dissertation. The first entails a systematic and unified approach to explain puzzles related to monotonicity in intensional contexts, which manifest as empirical phenomena in natural language. By these puzzles, monotonic inferences under modal and attitude verbs appear infelicitous, thereby challenging the monotonic semantics of modalities. This dissertation suggests that these puzzles arise from pragmatic effects triggered by the typical underspecified nature of conclusions reached by monotonic reasoning. In Chapter 4, a comprehensive analysis and a uniform account for this issue is provided.
Another focus of this dissertation is the study of modality. The thesis employs epistemic and desiderative modalities as examples to examine whether and how the puzzles related to monotonicity influence the semantic and logical properties of the modalities. Firstly, the interactions between knowledge, beliefs, and epistemic possibilities (expressed by the English modal verb might) are systematically investigated. In Chapter 2, a Bilateral State-based Epistemic Logic (BSEL) is proposed, exploring the concepts of knowing and believing an epistemic possibility. Specifically, it investigates the phenomenon of epistemic might and how its non-classical properties interact with the monotonicity of know and believe.
In Chapter 3, we broaden our discussion to include situations involving multiple agents, aiming to investigate how epistemic modals interact in multi-agent contexts. This extension is motivated by the argument that epistemic might is perspective-sensitive. Consequently, when it comes to claims involving epistemic might, different agents may evaluate them differently, thereby requiring the inclusion of agency as a factor in their interpretation.
Chapter 5 shifts its focus towards desiderative modality, proposing a novel logic of desire that incorporates causal inference. This proposal combines the betterness model of preference logic with the causal model of causal inference, resulting in a desire-causality model. Furthermore, a complete logic is developed for this model.
 In this dissertation, the treatment of monotonicity and modality is not conducted in isolation, but rather from the standpoint of their interaction. The investigations offer insights into the relationship between logic and language, semantics and pragmatics. These findings contribute to our understanding of these interrelated domains.