Now that all phrases had to have the same basic (XP, X-bar, X) structure, something had to be done about the Operator (Op) node.

It was decided that the Op node projected into a Complementizer Phrase (CP). So we can imagine that the tense element starts off in the head of the IP and moves to the head of the CP in order to make the sentence into a question sentence.

This kind of movement, referred to as move alpha (move α), became the most important part of Chomsky’s theory of syntax. Move α meant that you could pretty much move anything anywhere! However, there were rules (or principles) that limited this movement. These principles were then used to explain what kind of sentences were grammatical or not grammatical.

Think of a sentence like this:
1. She bought something.
In 1 we can imagine that there is a tense element in the head of IP after some kind of invisible movement operation as we can see above. What about the following sentence?
2. What did she buy?
If we use movement to explain 2, we could say that the tense element has moved to the head of CP in order to make a question sentence. So maybe we could imagine that the question word what in 2 starts out as in 3:
3. She bought what?
This would be similar to a non-question sentence like 4:
4. She bought something.
So maybe we can imagine the question word what has moved from the normal position as object of the verb buy to a position before the head of CP.

So it seems that we have a nice, neat solution if we imagine that what moves to the position indicated above. It is hanging off to the top left of CP. The subject she is also hanging off to the left of the IP in the sentence above. What should we name this position?

Chomsky, N. (1986). Barriers. Massachusetts: MIT Press.