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
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
22<br />
several boats(M).PL.NOM<br />
(Icel<strong>and</strong>ic, Sigurðsson 1991: 355f.)<br />
This proposal posits uninterpretable requirements behind phi-agreement <strong>and</strong><br />
Case that need have no interpretable consequences. In Chomsky (1995), it was not<br />
so. <strong>Phi</strong>-agreement <strong>and</strong> Case were also established through uninterpretable phi<strong>features</strong><br />
on T/v <strong>and</strong> Case <strong>features</strong> on nominals. However, <strong>the</strong> interpretable phi<strong>features</strong><br />
<strong>of</strong> a nominal were moved to T/v to license <strong>the</strong> deletion <strong>of</strong> <strong>the</strong> uninterpretable<br />
ones. One factor in <strong>the</strong> reconstrual <strong>of</strong> <strong>the</strong> mechanism through uninterpretable<br />
<strong>features</strong> alone is <strong>the</strong> absence <strong>of</strong> detectable consequences to agreement, as<br />
in (19) (Chomsky 2000a: 119, 146 note 71). <strong>Phi</strong>-agreement <strong>and</strong> Case are interpretively<br />
inert, unless accompanied by movement <strong>of</strong> interpretable material.<br />
Chomsky's work lays bare <strong>the</strong> ontological strangeness <strong>of</strong> such uninterpretable<br />
<strong>features</strong> <strong>and</strong> dependencies <strong>of</strong> syntax in <strong>the</strong> Minimalist Program (e.g. Chomsky<br />
2000ab). The program proposes that <strong>the</strong> content <strong>of</strong> lexical items is motivated by<br />
<strong>the</strong>ir role in <strong>the</strong> syntax-external systems <strong>of</strong> interpretation at LF <strong>and</strong> realization at<br />
PF, <strong>the</strong> Interpretability Condition (27)a, <strong>and</strong> that syntax arranges <strong>the</strong>m into structures<br />
to meet <strong>the</strong> legibility requirements <strong>of</strong> <strong>the</strong>se systems, Full Interpretation<br />
(27)b, without introducing <strong>features</strong> <strong>of</strong> its own, (27)c (see section 5.4).<br />
(27) a. Full Interpretation: Objects submitted by syntax to external systems<br />
must be fully legible to <strong>the</strong>m, both <strong>features</strong> <strong>and</strong> <strong>the</strong>ir arrangements.<br />
b. Interpretability Condition: Lexical items have no <strong>features</strong> o<strong>the</strong>r than<br />
those interpreted at <strong>the</strong> interfaces, properties <strong>of</strong> sound <strong>and</strong> meaning.<br />
c. Inclusiveness Condition: Syntax does not introduce new <strong>features</strong> like<br />
indices or deletion marks.<br />
Uninterpretable <strong>features</strong> are a purely syntax-internal device beyond <strong>the</strong>se proposals.<br />
<strong>Phi</strong>-agreement <strong>and</strong> Case assignment do not occur to meet interpretive<br />
needs, nor need <strong>the</strong>y be realized (as on nouns <strong>and</strong> past tense verbs in English). Yet<br />
<strong>the</strong>y appear to be syntactic dependencies. Their character is that <strong>of</strong> phrasestructurally<br />
unbounded dependencies subject to syntactic constraints like locality,<br />
illustrated in <strong>the</strong> remote agreement in (28) (<strong>and</strong> (19), (21)). They are also part <strong>of</strong><br />
<strong>the</strong> licensing conditions on syntactic structures. In (29), <strong>the</strong> need <strong>of</strong> DPs to bear<br />
Case rules out five <strong>and</strong> seven when ano<strong>the</strong>r DP, her, is closer to <strong>the</strong> matrix Case<br />
assigner <strong>and</strong> absorbs <strong>the</strong> only available Case (see section 5.4, 5.9).<br />
(28) a. There seemi/*seemsj to have been several studentsi awarded a bookj.<br />
b. There seemsi/*seemj to have been one studenti awarded several booksj.<br />
c. There seemsi/*seemj [to <strong>the</strong> watchersj] to have been a fishi caught.<br />
d. In <strong>the</strong> lakej werei/*wasj caught several fishi.<br />
(29) a. They showed her that five <strong>and</strong> seven are primes.<br />
b. They showed (*her) five <strong>and</strong> seven to be primes.<br />
c. Five <strong>and</strong> seven were shown (*her) t to be primes.