I would like to sketch the outlines of a theory of INFL. This is the set of features assigned to the clause as a whole, including tense and agreement (with various arguments), and which are relevant to determining the scope of a domain such as tensed (or finite) clause which may be relevant to the binding and/or bounding theory. On the present view, these features (which are fundamentally assigned to the clause node itself) are then realized by a rule of special cliticization which assigns them either directly to the verb (considered as the head of S), or to another special clitic position (tipically, second position; recall that location with respect to the head is also a possible position for special clitics).
- Anderson (2004), a pag.42-43