We expect mathematical proofs to explain why the propositions in question are true or why certain mathematical phenomena occur in certain situations. In this paper, I reexamine explanation-building processes by taking them as problem-solvers' understanding processes and by referring to research that has analyzed the relationships between explorations, understandings, and explanations in mathematical problem-solving. I discuss some interactive features among these components during problem-solving processes by introducing some examples and referring to that research. I then use those features to offer an elaborated conception of explanation-building processes that takes into consideration local explanations, full explanations, and the direct and indirect relationships between local and full explanations.