I am looking at the question "What do you think is in the box?". If a child trying to ask this question says "what do you think what is in the box?" with intonation reflecting that they are not reformulating their sentence (no pause in the middle), is their use of "what" a second time a reflection of difficulty with subordinated clause structure or something else? I appreciate the input.
This sentence involves an interrogative word what which has been extracted from its normal position at the beginning of the subordinate clause to the beginning of the matrix (main) clause.
The subordinate clause can be represented like this:
- [X ] is in the box.
- What is in the box?
The subordinate clause is embedded as the Complement of the verb know within a larger sentence:
- You think [what is in the box]?
The interrogative word is then extracted from the beginning of the subordinate clause to the beginning of the main clause. This triggers subject auxiliary inversion in the matrix (main) clause:
- What do you think [ ____ is in the box]?
In many grammars that gap in the subordinate clause is thought of as a trace, a kind of empty gap that refers back to the interrogative word, a bit like an empty pronoun:
- Whati do you think [ ___ i is in the box]?
It is common for very young speakers to plug that gap in the subordinate clause with another what (or whatever the interrogative word may be). This is taken by some linguists as evidence that a) there is a trace in that gap and b) that in the 'deep structure' (your innate mental grammar) the what at the beginning of the sentence originates from that space in the subordinate clause. According to this story the young speaker has not fully acquired yet is the correct abstract transformation in terms of deleting the original what. However, it could easily just be that the speaker is learning their language in chunks. They have added their what do you think chunk to their what is in the box chunk.