Mixing Words and Pictures: Mixed Evidence for Common Conceptual Representations