| Abstract: | Past research demonstrates that children learn from a previously accurate speaker rather than from a previously inaccurate one. This study shows that children do not necessarily treat a previously inaccurate speaker as unreliable. Rather, they appropriately excuse past inaccuracy arising from the speaker's limited information access. Children ( N = 67) aged 3, 4, and 5 years aimed to identify a hidden toy in collaboration with a puppet as informant. When the puppet had previously been inaccurate despite having full information, children tended to ignore what they were told and guess for themselves: They treated the puppet as unreliable in the longer term. However, children more frequently believed a currently well-informed puppet whose past inaccuracies arose legitimately from inadequate information access. |
