The V2 syntax of Medieval Romance

This chapter provides a detailed presentation of the main data and arguments which have been proposed in favour of claiming that the Medieval Romance languages were V2 systems and considers data from Old French, Old Occitan, Old Italo-Romance varieties, Old Spanish, and Old Portuguese. It provides new qualitative and quantitative evidence to show the nature of the prefield, Germanic inversion, matrix/embedded asymmetries, and the precise types of verb-first and verb-third-or-greater orders provide new evidence in favour of the V2 hypothesis. It also suggests that the diachronic emergence of a V2 grammar is entirely plausible on the basis of the available data. The main objections to the V2 account proposed in the literature are evaluated and argued to face empirical and theoretical problems.

Introduction

This chapter briefly outlines the main evidence that has been put forward in favour of the V2 hypothesis for Medieval Romance. It then gives a survey of the main developments in the study of Germanic V2, starting with seminal work in the 1970s and 1980s and outlining major empirical and theoretical developments which have taken place since. The chapter concludes by setting out the ‘Medieval Romance problem’ in need of resolution, namely the nature of microvariation or continuity between varieties, the status of V2 correlates in the null argument and clitic pronominal system and the diachronic development of V2.

Conclusion

This chapter summarizes the main findings of the book and considers their wider consequences. The Medieval Romance languages are argued to occupy important points on the V2 typology, which are linked to variable loci of V2 within the left periphery. Contrary to various claims in the literature, the analysis suggests that many Romance V2 systems are not less strict than all their Germanic counterparts. It is suggested that changes observable within the medieval period concerning V2 may account for important morphosyntactic isoglosses separating Romance varieties today, impacting the makeup of the left periphery, the null argument system, and the syntax–pragmatics mapping more generally. The chapter concludes by suggesting that the post-medieval period is an important forum for future research, with detailed research on the varieties in question from the 14th century onwards generally lacking in the literature.

Old Sardinian

This chapter presents a detailed study of the word order of Old Sardinian. The Sardinian data are of particular interest as the language has been claimed to have a form of verb-initial grammar in the small existing literature on the topic. Old Sardinian is shown to have the V-to-C movement characteristic of other Medieval Romance varieties but to lack obligatory fronting of a phrasal constituent, typical of V2 grammars. It is shown to have multiple subject positions, sensitive to the discourse status of the subject. Unusually within Romance, Old Sardinian is shown to have a VSO order in embedded clauses, with a strict adjacency between the embedded verb and the complementizer or relativizer. Overall, Old Sardinian is argued to have half of the V2 constraint, in that it has obligatory verb fronting into the left periphery, but no requirement for a phrasal constituent to also be merged.

Old Spanish

This chapter engages with recent criticism of Spanish’s status as a V2 language, suggesting that the criticism is ill-founded and arguing that it was a relatively strict V2 language, showing a prefield comparable to other V2 systems, Germanic-inversion, and direct object fronting without clitic resumption. Both verb-initial and verb-third orders are also shown to be heavily restricted in certain later Old Spanish texts when compared to other Medieval Romance V2 varieties. The chapter challenges the notion that Old Spanish was a symmetrical V2 language, presenting quantitative data that it showed word order asymmetries between matrix and embedded clauses, which are typical of asymmetric V2 systems such as Modern German and Dutch.

Old Gallo-Romance

This chapter provides a detailed account of the word order properties of Old French and Old Occitan. It shows that Old French is a descriptively stricter V2 system than Old Occitan but that both are V2 grammars, with a prefield nonspecialized for subjects, a dominant V2 order, Germanic inversion, and matrix/embedded asymmetries. However, as with Old Italo-Romance the precise makeup of the left periphery is distinct between varieties, later Old French does not license new information focus like Occitan, and both differ in their clitic pronominal and null subject properties.

Old Italo-Romance

This chapter provides a detailed account of the word order properties of Old Sicilian and Old Venetian. It shows that the two Old Italo-Romance varieties have much in common, namely a preverbal field not specialized for subjects, a dominant V2 order, two types of V2-related inversion, and matrix/embedded asymmetries. However, certain texts differ with respect to the regularity with which the verb appears in second position, the types of verb-initial and verb-third orders found, and whether new information focus can occur in the left periphery. The Tobler–Mussafia clitic system is also shown to be subject to intertextual variation.

Rethinking Medieval Romance V2

This chapter provides a comprehensive formal account of the V2 property in Medieval Romance. A number of points of continuity between varieties are highlighted, which are linked to the presence of a verb-movement trigger on a low left-peripheral head in all varieties and an Edge Feature on a low left-peripheral head in all varieties except Sardinian. The second part of the chapter argues that variation between the Romance V2 systems can be captured under the idea that there are at least two types of V2, correlated with the height of the V2 bottleneck, which can either be on Force or on Fin. The chapter then presents a map of the clausal structure of Medieval Romance and a detailed account of the evolution of Romance clausal structure in Section 2.