@article{oai:nagoya.repo.nii.ac.jp:00027897,
 author = {大村, 光弘 and Ohmura, Mitsuhiro},
 journal = {IVY},
 month = {Oct},
 note = {This paper concerns the effect of the Extended Projection Principle (EPP)(Chomsky (1982: 10)), that is, the phenomenon that clauses always have subjects. The Government and Binding Theory stipulated that clauses must have subjects, since this requirement could not follow entirely from the Projection Principle (Chomsky (1981: 38)). Thus, the expletive subject must appear at least at S-structure in (1), while it does not need a θ-role. (1) It is true that John loves Mary. The problem is that the requirement at issue shares certain effects redundantly with Case theory and θ-theory. I argue, adopting the minimalist framework (Chomsky (1993, 1994)), that the stipulation that clauses must have subjects may follow from the general checking requirement on formal features. The relevant assumptions are as follows: (2) Case features, φ-features, operator features, inflectional features, etc. must be checked. (3) "Strong" features are visible at PF and "weak" features invisible at PF. These features are not legitimate objects at PF ... if a strong feature remains after Spell-Out, the derivation crashes. (Chomsky (1993: 30)) (4) The accusative Case feature of V and the nominative Case feature of Tense are strong in English. (5) An expletive may be inserted into the phrase marker Σ only if it enters into the Case checking relation with a head. In the raising constructions in (6), the italicized DPs raise overtly to [Spec, [T AgrS]]. Tense checks their Case features and then discharges its own strong Case feature. (6) a. [AgrSP T-AgrS [VP John hit Mary]]. b. [AgrSP T-AgrS was [VP hit Mary]]. c. [AgrSP T-AgrS seems to [VP John like Mary]]. The expletive insertion is also driven so that the strong Case feature of Tense may be discharged before Spell-Out. In (7), for example, the expletive DP it is inserted into [Spec, [T-AgrS]], where its Case feature is checked by Tense. (7) [AgrSP [is-T-AgrS] true [that John loves Mary]]. It is also claimed that the effect of the EPP found in the complement clauses of the ECM verbs can be handled under the split VP hypothesis (cf. Koizumi (1993)). (8) We [VP2 consider [AgrOP tAgrO [VP1 tV [AgrSP to [VP John loves Mary]]]]]. (9) We [VP2 consider [AgrOP tAgrO [VP1 tV [AgrSP to be a man in the room]]]]. Since the Case feature of V is strong in English, DP John in (8) must raise overtly to [Spec, [V AgrO]]. As for (9), DP there is inserted into [Spec, [V AgrO]] so that V may check its Case feature., 本稿は、名古屋大学英文学会第34回大会(1995年4月22日、於名古屋大学)のシンポジウムで口頭発表した原稿に加筆し修正を加えたものである。},
 pages = {91--112},
 title = {拡大投射原理について},
 volume = {28},
 year = {1995}
}