Audio Visual Translation (AVT) is essential for importing and exporting moving pictures overseas. Dubbing and subtitling are the two processes involved in AVT. While dubbing is mainstream in Japanese terrestrial TV broadcasting, this was not the case at its inception. Dubbing emerged as a way to facilitate interaction between senders and audiences. This research investigated the dubbing process in NET (Nippon-Educational-Television, current TV Asahi). NET was the forerunner of AVT in Japanese TV from the 1950s to the 1970s. The translation norms of the sender were clarified by analyzing chronological changes in the form of the AVT used and the interaction between senders and audiences in Japanese TV. The study draws the following conclusions: 1) A significant degree of overlap exists between accountability and expectancy norms. The strength of expectancy norms was emphasized by Chesterman (1997) and was verified by this study. 2) Comprehension and naturalness used to be viewed as the standard in expectancy norms. However, the growing emphasis on richness and identity of expression overtook these previously held standards. 3) The norms were found to be stronger for adult programs and theater-movies than for children’s programs and television movies.