Abstract
This paper deals with Japanese LF-movement and variable binding by a wh-expression (WH). The main theme is exactly to what position a WH moves at LF. It seems that this theme has not been dealt with in great detail, and it is grossly assumed that a WH is either adjoined to COMP or S′ or is moved under COMP. This seems to me to be a gratuitous assumption, which is perhaps rooted in the thinking that nothing significant would come out by dwelling on the issue. However, it will be shown that the exact position of a moved WH has significant relevance to the problem of variable binding.
