Verb-first declaratives introduced by et and the position of pro in old and middle French

Initial subordinate clauses in Old French: Syntactic variation and the clausal left periphery

“…5 As an anonymous reviewer notes, some generative accounts (e.g.,Lemieux and Dupuis, 1995;Vance, 1997) treat instances of SV in main declaratives as IPs rather than CPs; only main declaratives in which a non-subject precedes the verb are CPs. This view contrasts with analyses that view all main declaratives as CPs, regardless of the position of the subject(Adams, 1987;Benincà, 1995Benincà, , 2006Roberts, 1993;Vance, 1993;Vance et al, 2010).…”

“…Referring to an observation dating to Foulet (1928) and still widely supported, Vance (1993) notes that ''Old French allows null subjects only in environments where the subject would be postverbal if expressed'' (p. 281; see also Roberts, 1993:123 and the numerous references therein). Vanelli et al (1985) and many subsequent approaches link the ability to license pro with verb raising, which occurs in main declaratives in Old French.…”

“…In theory at least, it has been held that Old French does not tolerate descriptively V≥3 orders (see Benincà, 1995;Posner, 1996Posner, , 1997, and, as Marchello-Nizia (1995:64--65) noted, early analyses of Old French (e.g., Thurneysen, 1892) could not satisfactorily account for V≥3 orders. Nevertheless, examples of V≥3 in main declaratives represent a non-negligible proportion of the data throughout the Old French period, even in 13th-century prose works, in which, according to Vance (1993Vance ( :284, 1997 the Old French V2 grammar appears at its most pronounced.…”

On null subjects and null objects in generative grammar

“…But some recent work on null subjects in Old French shows rather convincingly that the formal licenser and the content licenser can be different heads, that is, whereas the formal licensing head is C, the content licensing head can be I, at least in some marked constructions such as et V(S) constructions (e.g. Dupuis 1989;Roberts 1993;Vance 1993). Finally, Chinese-type languages pose a serious challenge to Rizzi's theory.…”

The nature of Old Spanish verb second reconsidered

“…The status of Old Romance et is complex, but there is a general consensus in the literature that clauses beginning with et should not be grouped together with matrix declaratives where the verb is in absolute-initial position(Vance, 1993; Labelle and Hirschbühler, 2005:68;Poletto, 2005 Poletto, :227f, 2014. Grouping et-clauses with other classes of V1, as found in Elsig (2010:2565) andSitaridou (2011Sitaridou ( , 2012 may undermine the validity of their observations on V1.…”