You Don't Believe This Is The Title
Moore's Paradox and its relation to the Surprise Exam Paradox, the Knowability Paradox, the Toxin Problem and Newcomb’s Problem
Max van den Broek

Abstract:
This thesis concerns Moore’s paradox and its relations to other paradoxes and problems. In particular, it concerns the relations between Moore’s paradox and the surprise exam paradox, the knowability paradox, the Toxin problem, Newcomb’s problem and multiple problems that are formulated for the first time in this thesis. The main claim defended in this thesis is that these problems are similar insofar as they all involve Moore sentences.
To defend this claim we develop a formal account of Moore sentences. Briefly, we state that a sentence φ is a Moore sentence for an agent A if (i) φ contains a modal operator $\box_A$ , (ii) φ is satisfiable with respect to a class of models C suitable for modeling $\box_A$ and (iii) $\box_A φ$ is unsatisfiable with respect to C. This definition has multiple upshots compared to previous definitions. Among these are that it adequately captures the intuition that doxastic Moore sentences can be true but cannot be believed. Further, the definition is formulated so that it is not limited to doxastic Moore sentences but applies to Moore sentences concerning other propositional attitudes as well.
Using our formal account of Moore sentences we discuss under what assumptions it can be said that a Moore sentence is involved in the surprise exam paradox, the knowability paradox, the Toxin problem and Newcomb’s problem. This discussion produces several interesting results. We resolve a debate about which sentence, precisely, is a Moore sentence in the surprise exam paradox. We also reconsider the consequences of the knowability paradox for the anti-realist verification thesis. Further, we introduce new variations of the Toxin problem and Newcomb’s problem, as well as two paradoxes concerning intention and desire that are similar to the knowability paradox, and suggest how all these problems may be solved.