Phi-features and the Modular Architecture of - UMR 7023 - CNRS
Phi-features and the Modular Architecture of - UMR 7023 - CNRS
Phi-features and the Modular Architecture of - UMR 7023 - CNRS
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
210<br />
You will be invited. (si + acc. object)<br />
(327) a. proarb + TNOM [vP ti [v [VP V O.NOMj]]]<br />
person<br />
(Italian, Burzio 1986: 49)<br />
nominative., number, *person si + nom. object<br />
b. proarb + TNOM [vP ti [vACC [VP V O.ACCj]]]<br />
person accusative si + acc. object<br />
The same scenario plays out for <strong>the</strong> objects <strong>of</strong> structurally poor nonfinite<br />
clauses under passives <strong>and</strong> unaccusatives. In (328), <strong>the</strong> unaccusative täytyä 'must'<br />
embeds an infinitive whose subject raises to or is controlled by <strong>the</strong> matrix oblique<br />
subject minun. The nominative <strong>of</strong> <strong>the</strong> matrix TNOM passes across minun to <strong>the</strong> object<br />
<strong>of</strong> <strong>the</strong> infinitive, which has no Agree/Case system <strong>of</strong> its own. If <strong>the</strong> object is<br />
[+person], <strong>the</strong> o<strong>the</strong>rwise unavailable accusative again steps in. 158<br />
(328) Minu-ni täyty-y ti nähdä kirje-et / se / *se-n / sin-ut.<br />
me-GEN must-3SG see-INF letter-PL.N/A it.N *it-A you-A<br />
I must see <strong>the</strong> letters/it/him/you.<br />
(Finnish)<br />
It is possible to stack such infinitives as in (329) to induce indefinite distance<br />
between <strong>the</strong> nominative assigner, intervener, <strong>and</strong> object: see Timberlake (1975),<br />
Vainikka (1989), Kiparsky (2001), <strong>and</strong> esp. Vainikka <strong>and</strong> Brattico (2009). 159<br />
158 The oblique argument seems attributable in some cases to <strong>the</strong> matrix predicate (Kiparsky<br />
2001: 334, followed in Rezac 2007), in o<strong>the</strong>rs it is clearly <strong>the</strong> genitive or o<strong>the</strong>r oblique subject <strong>of</strong><br />
<strong>the</strong> infinitive (Laitinen <strong>and</strong> Vilkuna 1993: 3.1, Maling 1993: 54 note 8, Koskinen 1999, 2000;<br />
Vainikka 1989: 303f., Brattico <strong>and</strong> Vainikka 2009: ex. 30). The oblique can be silent but syntactically<br />
active progeneric, distinct from proarb (Rezac 2007: 132 note 20, Holmberg forthc). It is detectable<br />
by preventing 3PL nominative from agreeing even when fronted, Lehmät pitää tuoda<br />
kotiin 'The cows(NOM) must.3SG bring (=be brought) home' (Laitinen <strong>and</strong> Vilkuna 1993; in<br />
<strong>the</strong>ir dialectal Lehmät (for Lehmien) pitää tulla 'cows.NOM (for cows.GEN) must.3SG come', cf.<br />
Kiparsky 2001: 359, <strong>the</strong>re might be an expletive pro). There is no satisfactory <strong>the</strong>ory <strong>of</strong> which<br />
infinitives are transparent, although it <strong>of</strong>ten goes with genitive/possessive subject (Kiparsky<br />
2001: 356f., but Vainikka 1989: 303f., Brattico <strong>and</strong> Vainikka 2009). Also to be addressed are independent<br />
subjectless infinitives in subject, adjunct, <strong>and</strong> noun complement positions that allow<br />
or require nominative objects (Hakulinen <strong>and</strong> Karlsson 1975: 343f., Taraldsen 1986, Sakuma<br />
1998, 1999, Hakulinen et al. 2004: §937-§940, Vainikka <strong>and</strong> Brattico 2009; cf. Maling 1993: 70<br />
note 19). The logic would suggest that <strong>the</strong>y have TNOM <strong>and</strong> an oblique subject. This seems confirmed<br />
by North Russian <strong>and</strong> Lithuanian parallels (Lavine 2000: 265).<br />
159 Finnish nonagreeing nominative/accusative may recurse across multiple DPs through transparent<br />
infinitives: Maija pyysi Jukan lukemaan kirjan 'Maija-NOM asked Jukka-ACC to.read<br />
book-ACC' vs. Pyydä Jukka lukemaan kirja 'Ask Jukka-NOM to.read book-NOM' (Vainikka<br />
1989: 267f.; Kiparsky 2001: 356; Hakulinen et al. 2004: §936, Reime 1993: 102 note 9, <strong>and</strong> esp.<br />
Brattico 2009: 90, but perhaps distinct from <strong>the</strong> partitive recursion discussed <strong>the</strong>re, over greater<br />
distances <strong>and</strong> with alternatives analyses). It suggests that a locus may Agree with multiple DPs<br />
until valued, so <strong>the</strong> nonagreeing low nominative is a Case assigned without valuation. If so, section<br />
5.2 (note 120) requires more to be said about why <strong>the</strong> PCC occurs with such a nominative.