The "Saying is Believing" Effect

In psychology, the saying-is-believing effect says that we tend to remember and believe what we say to other people, even if we told them what we thought they wanted to hear (instead of the full truth about what we really think). If we adjust our message to fit our audience, and then remember that version as the truth about something, it's no wonder that there can be such strong agreement between people who interact with each other a lot, and such deep, divisive, disagreement between those who rarely interact. Just think about the "conservative" and "liberal" ways of looking at the world!

Here are a few experiments in support of the saying-is-believing effect:

In an experiment by Higgins (1992) (replicated many times, as reviewed by Echterhoff et al. 2005), subjects are given behavioral descriptions of someone and told to describe this person to an audience so that the audience would be able to identify him. Subjects are told the audience is familiar with this person already, and, in different experimental conditions, that they either like or dislike the person.

The results are that subjects tended to tailor their descriptions of people to fit their perception of their audiences' expectations. Furthermore, their subsequent memories of the person described were more consistent with their audience-tuned message than the original description they were given.

This effect disappears, however, when they are led to believe the audience didn't identify the person on the basis of their tailored description. The message-tuning remains, but the saying-is-believing effect also disappears when the audience is an "out" group rather than an "in" group.

The common factor that seems to modulate the saying-is-believing effect is a sense of epistemic trust in the audience; The speaker has to trust the knowledge-forming credentials of their audience.

There is an especially interesting case of this for the acculturation of new immigrants. An experiment by Kosic et al. (2004) shows that, if immigrants are positively disposed toward their initial reference group of the new culture, they will tend to acculturate in proportion with their need for epistemic certainty in their beliefs (that is, for their need to have their own beliefs confirmed by people they trust). Otherwise, they will resist acculturation proportional to this need.

Here are references to the experiments:

Higgins ET. 1992. Achieving "shared reality" in the communication game: a social action that creates meaning. J. Lang. Soc. Psychology. 11:107-31

Echterhoff G, Higgins ET, Grolls S. 2005. Audience-tuning effects on memory: the role of shared reality. J. Personal. Soc. Psychol. 89:257-79

Kosic A, Kruglanksi AW, Pierro A, Mannetti L. 2004. Social cognition of immigrants' acculturation: effects of the need for closure and the referece group at entry. J. Personal. Soc. Psychol. 86:796-813

All of these experiments are summarized in the psychology review article:

E. Tory Higgins and Thane S. Pittman. Motives of the Human Animal: Comprehending, Managing, and Sharing Inner States. Annu. Rev. Psychol. 2008. 59:361-85