Why saying an explanation aloud makes you believe it more

Someone offers a tentative theory in a meeting to see how it sounds. By the end of the hour they are defending it. Nobody supplied new evidence in between. The theory acquired a supporter during the act of being spoken, and the supporter was the speaker.

What everyone sees

This is read as stubbornness, an unwillingness to lose face. Face plays a part, and the effect is present in conditions where nobody is watching and nothing is at stake reputationally, including private note-taking. The act of producing the material, rather than the presence of witnesses, is doing part of the work.

What is actually happening

Slamecka and Graf established the generation effect: material a person produces themselves is remembered better than the identical material read passively, so a self-spoken claim is simply more available afterwards than one merely heard. Higgins and Rholes added the attitudinal half, finding that participants who described a person to a listener with known views later remembered and evaluated that person in line with the version they had produced, an effect summarised as saying is believing. Combined, they describe a ratchet. Articulating a position makes it more retrievable and shifts the speaker’s own attitude towards it, and both changes happen without any new information entering the room.

Why it stays hidden

The ratchet hides because the speaker cannot separate their memory of the argument from their assessment of it. The version they generated is the one that comes to mind fastest, ease of retrieval feels like evidence of soundness, and the whole loop presents itself as having thought about the matter more carefully than the people who stayed quiet.