Against the identity of complementizers and (demonstrative) pronouns

In this work I will provide direct evidence against the identity of relative pronouns and complementizers/subordinators – proposed in recent works (Kayne 2010b) within the generative paradigm – with the aid of (diachronic and synchronic) data from Akkadian (an extinct Semitic language of Mesopotamia), Sogdian (an extinct Middle Iranian language), Germanic languages, Eastern and Western Iranian languages and Creoles. I will also show that the mismatch of relative (demonstrative) pronoun and complementizers does not weaken a proposal of a unified syntactic structure underlying the two clause-linkage phenomena of complementation and relativization. I will try to demonstrate that subordinate clauses (relative and complement clauses) are headed by light nouns/pro-nouns acting as "bridges of features" between matrix and dependent clauses. In particu-lar the simultaneous presence of demonstratives / relative pronouns and complementizers signalling clauses' edges in many languages is the primary evidence that (light) nom-inal elements are possibly required to trigger phenomena of clause linkage.