Theta-roles in DP, and an Introduction to Little n

The DP Syntax I !-roles in DP, and an introduction to little n time, we introduced the idea that the nominal • Last elements of the sentences (subjects, objects), are actually DPs, rather than NPs. • Determiners: the, a, some, every, Ø mass, Øproper, Øposs, ... we’ll continue our investigations of the internal • Today, structure of DPs. Deverbal nouns • The structure inside the DP can be as complicated as inside a clause, as it turns out. Deverbal nouns • 1) Pat broke the vase. 2) Pat’s breaking of the vase made me angry. Pat is an Agent, the vase is a Theme. 3) Pat danced. 4) Pat’s dancing startled me. 1) Pat broke the vase. • 2) Pat’s breaking of the vase startled me. 3) The bees startled me. • It seems to be possible to convert the whole clause Pat broke the vase into a “noun” (a DP). • TPs and DPs • One difference between clausal DPs and TPs is in the case realized by the arguments. • Agent is nom (from T), Theme is acc (from v) • Agent is gen, Theme looks like a PP introduced by of. So, the case assigners within a DP are different from the case assigners within a clause. Not all N’s assign !-roles. Some do, some don’t. Generally, the nouns related to a verb that assigns !-roles will assign !-roles. But something like lunch doesn’t. 1) Pat’s lunch was enormous. 2) My calling of him was unplanned. • Just as the verb break assigns !-roles, it seems as if the nominalized breaking assigns the same !-roles. The DP is in a way like a little clause. Two kinds of N 1) I called him. • What’s more, the relationship between break, Pat, and the vase seems to be the same inside the DP as it is in the clause. 2) Pat’s eating of lunch was shockingly rapid. • So, we can either find a DP with a !-role with genitive case, or we can find a possessor with genitive case, in SpecDP. Little n Ditransitive N • Consider the ditransitive verb give and the related noun gift. Just as give is responsible for three !-roles (Agent, Theme, Goal), so can gift be: 1) Pat gave an apple to Chris. 2) Pat’s gift of an apple to Chris was unexpected. • The exact same problem arises with ditransitive nouns as arose with ditransitive verbs. • Binary branching allows for just two arguments in NP. We need an additional projection for the third. Let’s try doing this just like we did for verbs… DP is like TP we suppose that DP works like TP, we can • Ifextend our theoretical machinery in an exactly analogous way. of Projections • Hierarchy D>n>N • UTAH DP daughter of nP: Agent DP daughter of NP: Theme PP daughter of N": Goal The of case • What’s the deal with this “of case” that objects in DPs get? Isn’t of a preposition? Shouldn’t of cheese in The gift of cheese to the senator was appreciated be a PP? • This of is completely meaningless, it acts like a case marker. So, we’re going to analyze it as such. Of cheese is a DP with the of case marking. Just like Pat’s is a DP with the genitive (’s) case marking. • Treating of as case allows a complete parallel between TP and DP; v has an [acc] feature, n has an [of] feature. • • • • Case in the DP In the DP, the “subject” appears with genitive case. • Cf. The subject in TP, which has nominative case, due to a [nom] feature on T. So, we say D can have a [gen*] feature. • This checks the genitive case on the subject of the DP, and forces it to move into SpecDP. In the DP, the “object” appears with the preposition of. • Cf. The object in TP, which has accusative case, due to an [acc] feature on v. So, we say that n has an [of] feature. Passive nouns • Last week, we looked at the passive construction: 1) The sandwich was eaten • Here, the Theme the sandwich becomes the subject because the strong feature of T forces it to move to SpecTP. The v does not project an Agent. Passive nouns Passive • • In the passive, v does not introduce an Agent, and does not have an [acc] feature. • T still has a [nom] feature, so it checks the [case] feature on the sandwich. • T has a [uD*] feature, so the sandwich moves to SpecTP to check it. Passive nouns If the DP has a head D like the that does not check genitive case, then there can be no Agent (nothing could check its case), and the Theme stays unmoved (its of-case checked by n). • Very similar to the passive, if an n doesn’t introduce an Agent, the Theme can move to SpecDP and surface as genitive. Case and !-roles We now predict the observation Adger makes: Either an Agent or a Theme can show up in the genitive, but only a Theme can show up with of-case. 1) Adger’s analysis of the DP is simple. 2) The DP’s analysis is simple. 3) *The analysis of Adger is simple. • This is essentially the same as the generalization that, in a clause, either an Agent or a Theme can show up with nominative case, but only a Theme can show up with accusative case. 1) I called her. 2) She tripped. • • • Back to possession Prior to today, the genitive case was associated with the possessor. So far today we’ve been looking at deverbal nouns, where genitive case goes to the subject. 3) *Her tripped. 4) *Tripped her. Possessors • Adger proposes that Possessors are introduced by a new head, Poss. • HoP: D > (Poss) > n > N Our new improved UTAH says, among other things: • • DP daughter of NP: Theme DP daughter of nP: Agent Possessors are neither of these, so possessors need to be initially Merged into a distinct place in the structure. Hungarian possessors 1) Az en kalapom the I hat ‘my hat’! ! ! 3) A te kalapod ! 2) A Mari kalapja the Mary hat ‘Mary’s hat’ Adjectives • Adjectives are to nouns as adverbs are to verbs. So what would the structure be for Pat’s complete destruction of the sidewalk? Or the silly idea? Or the pencil on the desk? • In Pat completey destroyed the sidewalk, we adjoin completely to vP. The subject moves to SpecTP. • In the same way, we adjoin complete to nP, and Pat moves to SpecDP. the you hat ‘your hat’ 4) Marinak a kalapja Mary the hat ‘Mary’s hat’ • Assuming that the DP in Hungarian has the basic structure we’ve been discussing, what is the structure of this kind of possessive construction? • How about that (person?) agreement on ‘hat’? Adjuncts The Italian DP • In Italian, in many cases, there is simply an option (stylistically governed) as to whether you say The Gianni or just Gianni: 1) Gianni mi ha telefonato. Gianni me has telephoned ‘Gianni called me up.’ 2) Il Gianni mi ha telefonato. the Gianni me has telephoned ‘Gianni called me up.’ The Italian DP However, there is a difference with respect to the order of adjectives and the noun depending on which one you use. 1)L’ antica Roma the ancient Rome ‘Ancient Rome’ • 5)*E’venuto vecchio Cameresi came older Cameresi 3)Roma antica Rome ancient 6)E’venuto Camersi vecchio. came Cameresi older Generalization: If there’s a determiner, the noun follows the adjective. If there isn’t the noun precedes the adjective. We can apply the same analysis to the order nouns and adjectives as we did to the order of adverbs and verbs. • 4) E’venuto il vecchio Cameresi. came the older Cameresi 2)*Antica Roma ancient Rome The Italian DP • Recall that in French, verbs precede adverbs, but in English, verbs follow adverbs. We conclude that in French, v moves to T. In Italian, when the noun precedes the adjective it has moved over it, to D. The generalization is that this happens except if D is already filled. • L’ antica Roma the ancient Rome • Roma antica!! ! *Antica Roma Rome ancient! ! ancient Rome Parameters • Languages differ on whether n moves to D, yielding some languages where nouns precede adjectives, and some languages where nouns follow adjectives. • • What governs whether n moves to D is the strength of an uninterpretable feature checked on D or n by the other. One such feature is [unum:]. • • • Italian: [unum:*] is strong on null determiners. English: [unum:] is weak, even on null determiners. • [Øindef Happy students] poured forth from the classroom. harisat ha-oyev ’et ha-’ir destruction the-enemy OM the-city ‘The enemy’s destruction of the city’ • tipul ha-Siltonot ba-ba’aya treatment the-authorities in-the-problem ‘The authorities’ treatment of the problem’ Construct state. What seems to be happening here? Again, parametric variation. • [gen] feature of D is weak in Hebrew, strong (when there) in English. But [unum:] feature is strong in Hebrew. • Rather like VSO languages, where v moves to T (like in French, unlike in English), but the subject doesn’t move to SpecTP (the [uD] feature of T is weak). Where does the byphrase attach? Adverb tests can give us a hint… • • • • • • • • •[ Il mio Gianni] ha finalmente telefonato. the my G. has finally called ‘My Gianni has finally called.’ DP • *[ •[ The sandwich was eaten by Pat today at noon The sandwich was eaten by Pat at noon today The sandwich was eaten today _ by Pat _ at noon The sandwich was eaten at noon _ by Pat _ today The dishes were washed by Pat _ poorly _ yesterday The dishes were washed poorly by Pat yesterday The sandwich was eaten by Pat _ sloppily _ at noon The sandwich was eaten sloppily by Pat at noon Conclusion? DP Mio DP Some Hebrew • • • Likewise, languages differ on whether v moves to T, yielding some languages (e.g., French) where verbs precede adverbs, and some languages (e.g., English) where verbs follow adverbs. More Italian, same point Gianni] ha finalmente telefonato. Gianni mio] ha finalmente telefonato. Ditransitive passives • Consider again Pat gave Chris books. • Chris was given books. • *Books were given Chris. • Pat gave books to Chris. • Books were given to Chris. • *Chris was given books to. Japanese Numeral Quantifiers • • • • • • Gakusei ga hon o 4-satu katta students nom book acc 4-cl bought ‘The students bought four books.’ ?*Gakusei ga hon o 4-nin katta students nom book acc 4-cl bought Gakusei ga 4-nin hon o katta students nom 4-cl book acc bought ‘Four students bought books.’ Gakusei ga kyoo 3-nin kita students nom today 3-cl came ‘Three students came today.’ Hon o Taroo ga 2-satu katta books acc T nom 2-cl bought ‘Books, Taroo bought two.’ Yuube, kuruma ga doroboo ni 2-dai nusum-are-ta last night cars nom thief by 2-cl steal-pass-past ‘Last night, two cars were stolen by a thief.’!(Miyagawa 1989) Italian ne-cliticization • Maria ha visto Gianni. Maria lo ha visto. M has seen G. M him has seen. • Gianni trascorrerà tre settimane a Milano. G spend.fut3sg 3 weeks in M • Gianni ne trascorrerà tre (*ne) a Milano. G of-them spend.fut3sg 3 in M. • Alcuni {persone/*ne} trascorreranno tre settimane a Milano some people/of-them spend.fut3pl 3 weeks in M. • • • • Telefoneranno tre persone domani *Ne telefoneranno tre domani Ne arriveranno tre domani Ne furono arrestati molti.