Correction closes agendas, causing omissions

This proposed hypothesis asks why a person can know a relation in a source accurately yet leave it out when explaining the source to someone else. Its answer is a speaking agenda: the relations a speaker still plans to put into a message, separate from facts held privately. Accurate correction may close that unexpressed plan, leaving the next recipient without it. The test compares correction followed by reconstruction with the reverse order. An answer-free prompt identifies the relation the recipient needs without supplying it. The proposed signature is specific: opening the agenda before correction should reduce omissions; opening it afterward should not. Source knowledge, attention, memory, message length and recipient answers would distinguish this account from audience adaptation or retrieval.