What's in a manner of speaking? Children's sensitivity to partner-specific referential precedents

Do young children form "referential pacts"? If a person has referred to an object with a certain term (e.g., the horse), will children expect this person to use this term in the future but allow others to use a different expression (e.g., the pony)? One hundred twenty-eight children between 3 and 5 years old co-operated with an experimenter (E1) to move toys to new locations on a shelf. E1 established referential terms for all toys in a warm-up game. Then, either the original partner, E1, or a new partner, E2, played a second game with the same toys. In this game, the experimenters referred to toys using either their original terms from the warm-up game or new terms. Children were slower to react to new terms than old, and this difference in reaction times was greater in the original partner condition (but only on the first trial). Children sometimes protested at the use of new terms, doing so regardless of their interlocutor's identity. We contrast these findings with those for adults and discuss their implications for the debate regarding the nature of referential pacts.