Research on minority in¯uence over the years has revealed that although sources in the numerical majority tend to be very persuasive on an immediate, public, and direct level, sources in the numerical minority tend to be resisted on this level (e.g., Moscovici, 1980, 1985a; see Wood, Lundgren, Ouellette, Busceme, & Blackstone, 1994, for a review). Early theorizing by Moscovici (1980, 1985a, 1985b) and others (e.g., Mugny & Perez, 1991) suggested that people publicly agree with majority messages and reject minority messages out of hand to avoid aligning themselves with deviant groups. That is, public acceptance and rejection of majority and minority messages has been thought to occur with very little issue-relevant thought. Importantly, though, both majority persuasion and minority resistance have also been argued to involve more thoughtful processes. For example, Erb and Bohner (2001) have argued that the majority or minority status of the source of a persuasive message can bias message recipients' thinking about the attitude issue in one direction or another. Speci®cally, majority sources are postulated to produce more favourable thoughts (i.e., pro-arguments) in response to persuasive messages, whereas minority sources are posited to produce more unfavourable thoughts (i.e., counterarguments). These thoughts, in turn, foster relative persuasion and resistance.