Theories of visual and diagrammatic reasoning typically seek to account for either the formal semantics of diagrams, or for the advantages which visual representations hold for the reasoner over other forms of representation. Regrettably, almost no theory ex.ists which accounts for both of these issues together, nor how they affect one another. We do not attempt to provide such an account here. We do, however, seek to layout a larger context than is generally used for examining the processes of using diagrams in reasoning or communication. A context in which detailed studies of sub-problems, in particular what it is that makes diagrams effective, may be embedded. Accounts of the embedding of typical text-based logics in the computational processes of reasoners and communicators are relatively well developed from several decades of research. Analogies between the sentential and the graphical cases are quite revealing about both similarities and differences. This paper examines carefully these analogies, the decomposition they provide of subproblems for analysing diagrammatic reasoning, and the effectiveness thereof.