Multiple aspects of the external argument ArgE need better understanding. First, no principle of UG explains why ArgE must stay structurally outside any lexical verbal projection. This fact is argued to result from the USM requiring an isomorphic mapping between semantics and structure while UG itself cannot guarantee such a result via X'-theory. The solution is iconicity of independence, which matches ArgE’s conceptual “independent existence” (Dowty 1991) from an event with its structural separation from the projection of the event-denoting V. Second, the grammatical properties of ArgE, especially given the iconicity account, must be compared with those of oblique arguments, eventually leading to a theory of the morphology–syntax interface which allows a uniform account of several types of cross-linguistic fact. Third, regarding word order, moved constituents exhibit the earlier-iff-structurally-higher correlation while in situ constituents don’t, with ArgE typologically in both groups. This property, together with the unique word orders produced by linear iconicity in previous chapters, prompts the hypothesis that linearization results from computational cost and the 2nd law of thermodynamics, which further identifies a new locality phenomenon: the functional domain island.

Chapter 6 aims at accounting for modification by temporal adverbials. The binary approach allows them to operate on different levels of tense structure above S_0. Crucial is the new insight that the relation between modifier A and modified B is not to be expressed as location of A in B but rather as the intersection of A and B. This insight and the flexibility of (deictic) adverbial modification offers the possibility of presenting a structural solution to the Present Perfect puzzle. It also sheds light on the nature of the in/for-test: it explains why terminative sentences with a for-adverbial introduced at a higher level may be regarded as well-formed as opposed to those which require a token interpretation.