Abstract
Whereas our discussion of V2 in OFr in previous chapters required extensive comparison with the modern Germanic languages, the topic of the present chapter invites comparison primarily with the modern Romance languages. Just as German and Dutch were the first V2 languages to be studied in detail in the generative framework, Spanish and Italian were the first null subject languages to receive widespread attention from generativists (cf. e.g. Perlmutter (1971), Taraldsen (1978), Burzio (1981, 1986), Rizzi (1982)). In these languages, it appears that the presence of six distinct endings on the verb forms for the three singular and three plural persons substitutes entirely for the subject pronoun, much as in their parent language, Latin. If a pronoun appears, it receives a marked interpretation: it is usually stressed or emphatic in some way.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Notes
The tendency to avoid subject pronouns unless a tonic form is needed (expressed as part of the “Avoid Pronoun Principle” in (Chomsky (1981)) in Spanish and Italian is not absolute. In Italian, the syncretous first, second, and third person singular forms of the subjunctive are partially differentiated by the use of overt unemphatic to in the second person (Renzi and Vanelli (1983)). A more general use of atonic overt pronouns is found in many Spanish dialects (cf. Cameron (1992), Toribio (1993)).
There are, however, differences among adverbs with respect to the frequency with which they trigger inversion of the Sp rather than its omission (cf. Offord (1971)).
Roberts (1993:124) also points out that the verb governing pro may be in Agr° rather than in C°, but he does so for different reasons. His double Agr projection, first proposed in Cardinaletti and Roberts (1991), allows pro to be governed by a finite verb even in embedded CV clauses such as those discussed in section 5.3. As the discussion in 5.3. shows, however, the double Agr proposal cannot account for the et V clauses discussed here.
For Dupuis (1989), the licensing head and the identifying head of pro can be distinct: when the verb does not move to C°, a [+wh] C° can license pro under directional (i.e. rightward) government. There is no directional requirement on the identification of pro by Agr; hence referential pro is possible in the initial position of a [+wh] embedded clause where it is licensed by C° under directional government and identified to the left by Agr°. Since I consider that these types of clauses are only grammatical in Early OFr (and in Middle French — see Chapter 6), I do not want to include this context in my account of 13th century pro. Roberts (1993) also rejects Dupuis’ claim for 13th c. French (p. 135) but suggests that a split system (not entirely identical to Dupuis’) may have been active in the transitional period of Middle French (p. 182). For further discussion of Roberts’ account of MidF, see Chapter 6.
Old French could be considered to be “morphologically uniform” if the first person singular form is taken to consist of a root plus a zero affix. Jaeggli and Safir explicitly reject the possibility of zero affixes, however, since to do so would permit a language such as English to be considered morphologically uniform. For a further suggestion, see Roberts (1993:125–128).
Rohrbacher does not attempt to characterize the distributional properties of pro in OFr. However, his reliance on the principle of Economy of Projection suggests a wày in which his system could be reinterpreted in such a way as to produce something resembling the OFr case. Accept for a moment two notions that we have presumed are untrue: (1) that the OFr paradigm is weak rather than strong and (2) that OFr is an IP-V2 language as hypothesized by Lemieux and Dupuis (1995). Under this scenario, AgrP is projected but the verb in its head is not sufficient to license the projection. The effect is that Spec, AgrP must be overtly filled, but nothing in the theory forces it to be filled by a subject. If Spec, AgrP is an A’ position in symmetric V2 languages, then the fronting of a topic to Spec, AgrP could fulfill the conditions of licensing just as well as a subject, and pro would be possible just in case Spec, AgrP were filled by a non-subject constituent.
Adams (1988) suggests that complex subordinators like devant que (cf. (49)) behave like wh-phrases because they are contentful. This strategy will not accommodate (49).
Note, however, that Roberts (1993: 137) proposes a postverbal account of pro in these contexts that does not suffer from this drawback. His double Agr structure (cf. Cardinaletti and Roberts (1991)), which allows a verb in Agrl° to license pro in the spec of Agr2P, is presumed to disappear from the language by the 13th century. Examples (48–50) would thus be considered archaic by Roberts. As we find other reasons to reject the double Agr account (cf. Chapter 4), we do not follow this lead.
I have not found embedded SpCV clauses parallel to (61)-(64), although Moignet (1965: 99) cites an example of impersonal it in this construction in a less formulaic context: (i) et se it de cele journee vous remenbrast (Mort Artu 119, 102) and if it of this day to-you might-remember ‘and if you were to remember this day’ It is possible, however, to find overt pronouns with these expressions in the Queste outside of SpCV: (ii) mes si tost com nos cuiderons qu’il en soit lex et mestiers, but so soon as we will-believe that-it ’en’ be fitting and necessary nos l’i envoierons.“ (Q 3, 18) we him-there will-send ’but as soon as we think it’s fitting and necessary, we will send him there.’
Offord (1971) studied CVSp and CV clauses in four 14th century texts and found that certain initial adverbs triggered inversion, as opposed to omission, more frequently than others. Offord did not offer an explanation for this observation. Although I have not gone over his data systematically, I suspect that his class of inversion-triggering adverbs would correlate closely with those that are most likely to evoke deictic linking of the discourse.
Dekydtspotter (1995), in a semantic framework, considers such clauses to be CP’s.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1997 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Vance, B.S. (1997). Null Subjects in Old French. In: Syntactic Change in Medieval French. Studies in Natural Language and Linguistic Theory, vol 41. Springer, Dordrecht. https://doi.org/10.1007/978-94-015-8843-0_5
Download citation
DOI: https://doi.org/10.1007/978-94-015-8843-0_5
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-4886-8
Online ISBN: 978-94-015-8843-0
eBook Packages: Springer Book Archive