On the categorial status of 'come' in old Florentine and modern standard Italian

Unlike what happened in main questions, in old Florentine embedded questions featured no verb movement to the left periphery; as a consequence, in embedded wh-questions the subject intervened between the wh-item and the inflected verb. On the other hand, the wh-phrase could either be preceded or followed by a left-dislocated constituent; interestingly, the possibility of inserting a constituent between the wh-item and the verb was particularly frequent in embedded questions introduced by 'come'. I will argue in favor of the idea that the embedded clauses in which 'come' is followed by a topicalized constituent can in fact be analyzed as relative clauses in disguise, where 'come' is sitting not in the specifier of the Focus projection, but in the specifier of a higher functional projection, arguably Force, situated at the left of the (recursive) Topic projections. This hypothesis receives support precisely by the interpretive ambiguity that 'come' displayed in several contexts in old Florentine, where it could either be analyzed as a wh-item, that is, as a maximal projection sitting in Spec,ForceP, or as a subordinating complementizer lexicalizing the head Force°. The possibility for the corresponding item 'como' to be used as a subordinating complementizer was independently attested in other old Italian varieties, like old Paduan. In modern standard Italian 'come', when used as a complementizer, is subject to some distributional constraints which do not affect the standard complementizer 'che'. The categorial change from specifier to head within the left-periphery seems to be a relatively wide-spread phenomenon; the pervasive ambiguity of 'come' reveals that old Italian varieties featured an ongoing process of reanalysis of this element that reflects a crosslinguistically well attested diachronic tendency.