26.11.2012 Views

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

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

136<br />

(215) Je lui ferai parvenir votre lettre.<br />

I him.D will.make reach your letter<br />

I will make your letter reach him.<br />

(Tasmowski 1985: 281)<br />

(216) Que faire pour amadouer le patron? Lui écrire? Lui faire téléphoner?<br />

Mais par qui? Moi, je lui ferais téléphoner MARIE. Elle est fine, elle le<br />

connaît bien, elle saura quoi lui dire.<br />

What to do to s<strong>of</strong>ten <strong>the</strong> boss's heart? Write to him? Make someone<br />

phone him? But who? I would get MARY to phone him.D (lui). She is<br />

subtle, she knows him well, she'll know what to tell him.<br />

(Tasmowski 1985: 360; my emphasis)<br />

This observation might be integrated into a syntactic account by supposing that<br />

<strong>the</strong> narrowly focussed subject <strong>of</strong> (216) undergoes A'-movement <strong>and</strong> <strong>the</strong>reby no<br />

longer counts as a Specified Subject, or that <strong>the</strong> narrow focus reading coerces <strong>the</strong><br />

verb into an unaccusative-like structure. Instead, Tasmowski proposes that <strong>the</strong><br />

problem in causativizing unergatives is pragmatic. A conflict arises because both<br />

<strong>the</strong> dative clitic <strong>and</strong> <strong>the</strong> agent-causee want to be identified as <strong>the</strong> topic. There is no<br />

conflict if <strong>the</strong> clitic is not dative, (214), or <strong>the</strong> causee not an agent, (215), <strong>and</strong> it is<br />

resolved if <strong>the</strong> causee-agent is <strong>the</strong> focus ra<strong>the</strong>r than <strong>the</strong> topic, (216). 97<br />

The options seem open. On syntactic accounts, causatives are an environment<br />

where <strong>the</strong> absence <strong>of</strong> a cliticization site <strong>and</strong> a clitic climbing failure cannot suspend<br />

<strong>the</strong> Cliticization Requirement, unlike <strong>the</strong> PCC. On pragmatic accounts, <strong>the</strong><br />

PCC repair cannot respond to an extrasyntactic problem in interpretation, analogous<br />

to morphological clitic cluster problems in realization. Ano<strong>the</strong>r such<br />

interpretive problem has already been seen: Condition B in section 4.4.<br />

4.6.8 The weak PCC<br />

There may be an argument for <strong>the</strong> specificity <strong>of</strong> <strong>the</strong> PCC repair that inverses<br />

<strong>the</strong> logic hi<strong>the</strong>rto: <strong>the</strong> repair 'overapplies' on clitic clusters that should fall under<br />

<strong>the</strong> PCC but do not. Bonet (1991: 4.1.1) observes that some Romance varieties<br />

tolerate 1/2.ACC + 2/1.DAT clitic clusters while still barring 1/2.ACC + 3.DAT.<br />

She names this <strong>the</strong> weak PCC, in contrast to <strong>the</strong> strong PCC that rules out all<br />

survivre 'survive' (Kayne 1975: 252ff., Tasmowski 1985, Rooryck 1988b, Legendre 1989b: 146,<br />

L<strong>and</strong>au 2010: 3.7, Folli <strong>and</strong> Harley 2007: 213).<br />

97 Tasmowski's proposal has far greater scope than <strong>the</strong> causativization <strong>of</strong> unergatives, including<br />

cliticization failure in transitives due to <strong>the</strong> Specified Subject Condition, <strong>and</strong> <strong>the</strong> PCC created by<br />

dative causees in section 4.5. For <strong>the</strong> PCC, it fails for various reasons, such as <strong>the</strong> contrast between<br />

<strong>the</strong> PCC 1/2.ACC 3.DAT clusters <strong>and</strong> good 1/2.DAT 3.DAT clusters (Postal 1981, 1983,<br />

1984, 1990; cf. Tasmowski 1985: 362 note 7), <strong>and</strong> <strong>the</strong> non-cancellability <strong>of</strong> <strong>the</strong> PCC by dative àphrase<br />

through manipulating information structure in <strong>the</strong> manner <strong>of</strong> (216), shown in (177). Similarly<br />

<strong>the</strong> Specified Subject Condition cliticization failure in transitives like (174) seems precisely<br />

inverse to <strong>the</strong> effect predicted for locative clitics, Tasmowski (1985: 356-61).

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!