I could not find syntactic arguments to support the existence of a separate T(ense) category inside the tree for the sentence “John rarely spends the weekend with his family” As well as syntactic arguments supporting the claim that T(ense) is a syntactic head of the traditional clausal category S. If someone please can simplify the arguments for me because this is only my first semester in syntax.
The usual argument is that the present-tense marking only ever appears once in the phrase. If you put another verb before the main one, that one takes the tense marking instead:
- John spends […]
- John can spend […]
- John does spend […]
And even once you start stacking up modals and auxiliaries, it's only the leftmost one that gets the present-tense marking:
- John is spending […]
- John has been spending […]
- John could have been spending […]
You can explain this by saying the present-tense marking isn't actually an innate part of the verb, but is a separate syntactic element. If this element exists further to the left, that also explains why it's always the leftmost verb in these chains that gets the marking.
Similar experiments suggest that the best place for this hypothetical element is outside the VP (or vP if you like) but inside the CP, which points to a structure like this.
(Why do we posit a CP in these sentences? Well, similar arguments. We posit an invisible element that causes certain things to happen, and figure out where that invisible element would have to appear, and in the end it all works out.)
Now, is this the only way to explain these phenomena? Absolutely not. But it's the one that's caught on in generativism, as jlawler points out in the comments. And that's why it's the one that's taught in all the syntax courses.