In adult native English, (non-echoic) wh-questions require movement of a whexpression to the front of the relevant interrogative clause. In long-distance questions, wh-movement can move a wh-expression out of a lower into a higher clause (as in What dress do you think she will wear?). UG principles require wh-movement to involve a local copying operation in which wh-expressions move in a local (successive-cyclic, one-clause-at-a time) fashion to the front of each clause containing them, and movement involves placing a copy of the moved constituent at the front of the relevant clause and then deleting the original. In this paper, we report on elicitation and grammaticality judgment tasks designed to explore the range of structures produced (and judged grammatical) by elementary Japanese Learners of English (= JLE) in contexts where native English speakers use Long-Distance Questions (LDQ). Our findings show that alongside target-like LDQ_s, the learners in our study produced a range of other types of structure, some of which provide evidence for wh-movement being a local operation (including wh-splitting structures such as What do you think color she likes? where the wh-word what is split from the noun color that it modifies, and the noun color is stranded at the front of the complement clause), and some of which provide evidence for wh-movement being a copying operation (including wh-doubling structures such as What do you think what color she likes? which have a copy of the wh-word what at the front of both the matrix clause and complement clause). Examination of the syntax of the structures produced by the learners in our study leads us to the conclusion that such sentences provide empirical evidence that wh-movement involves a local copying operation in the L2 grammars of the learners in our study, and hence that principles of Universal Grammar constrain the syntactic representations formed by L2 learners.