We are essentially creatures of habit. If you have any question about this, just transfer your wristwatch to the other hand. Even though it should not make the slightest difference on which hand it is worn, you will soon find that you are very aware of the change.
This is the context. Let's consider this part:
It should not make the slightest difference on which hand it is worn.
An ELL user has asked about the function of the it in this sentence, and I'd like to answer their question. As far as I know, this phenomenon is called extraposition. The clause on which hand it is worn has been moved to the end of the sentence, with the dummy it taking its place as the subject. To show how this phenomenon works, I'd like to use this sentence:
On which hand it is worn should not make the slightest difference.
I'm not a native speaker. I asked about my paraphrase on a couple of ELL websites, and everyone who answered said it was grammatical. However, here is an excerpt from The Cambridge Grammar of the English Language, Chapter 12, 6.1d:
 FUSED RELATIVE INTEGRATED RELATIVE i a. What she referred to was Riga. b. The city which she referred to was Riga. ii a. *To what she referred was Riga. b. The city to which she referred was Riga.
When the relativised element is complement of a preposition the fused construction requires that the preposition be stranded, as in [ia]: it cannot be fronted along with its complement, as it can in the integrated relative construction [iib]. The difference in grammaticality here reflects the fact that which she referred to is a clause while what she referred to is an NP. Fronting the preposition in the integrated construction places it at the beginning of the clause, while fronting it in the fused construction places it before the NP. The deviance of [iia] is thus comparable to that of *To the city which she referred was Riga. In the integrated case the antecedent city and the relative pronoun which are distinct and the preposition can come between them, but in the fused case the antecedent and relative pronoun are not distinct and hence there is no place for a fronted preposition to occupy.
Example [iia] has the same structure as my paraphrase above — both have a preposition before a relative pronoun — yet [iia] is ungrammatical.
I asked about [iia] and got mixed answers: some said it was grammatical, some said it wasn't. I didn't ask why because I didn't expect those people to be linguists or teachers.
- Is the clause in my example also a fused relative?
- Why is my example correct while the CGEL one is not?