Argues that in the final phase of the eight months of US-Japan talks leading to the attack on Pearl Harbor, serious mistranslations in the US decoding system were a significant factor in the misunderstandings which grew between the two sides.