In my dissertation I devise a number of new puzzles about belief reports which, unlike Frege's and Kripke's classical puzzles, cannot be solved by appealing to the notion of a mode of presentation. A solution to both classical and new puzzles can be achieved by adopting a new Russellian account of belief reports, which I present in this dissertation. The new account involves two psychological devices (which will be proved not to be modes of presentation): belief subsystems and cognitive coordination. The former device originates in Donald Davidson's idea of explaining away cases of apparent irrationality of a subject by partitioning her mind into semi-autonomous compartments or subsystems. The latter device takes its cue from Kit Fine's notion of coordination and deals with the ability (or inability) of a subject to recognize two occurrences of an object within Russellian propositions as occurrences of the same object.