On the basis of Abe's (2016) proposal that in the case of an A-chain, any member can be the target for pronunciation, I argue that Japanese passives may have the bottom copies of the produced A-chains pronounced. When the passivized subjects appear to occupy their original θ-positions, it is standardly claimed that the relevant word order is derived by scrambling other material sentence-initially. Contrary to this claim, I demonstrate that there are cases, including those of what Miyagawa (1989) calls the causative-passive construction, where the passivized subjects actually occupy their original θ-positions. I then provide evidence that those passivized subjects undergo “covert” A-movement, hence giving support to my bottom copy pronunciation analysis.